<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[oppr - Medium]]></title>
        <description><![CDATA[Applying our solutions for a real-world impact. - Medium]]></description>
        <link>https://medium.com/oppr?source=rss----9de3d09b6cbe---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>oppr - Medium</title>
            <link>https://medium.com/oppr?source=rss----9de3d09b6cbe---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 21:18:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/oppr" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Medical Imaging Downloader for CornerstoneJS and Orthanc]]></title>
            <link>https://medium.com/oppr/medical-imaging-downloader-for-cornerstonejs-and-orthanc-d08c3a508d9b?source=rss----9de3d09b6cbe---4</link>
            <guid isPermaLink="false">https://medium.com/p/d08c3a508d9b</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[downloader]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[medical-imaging]]></category>
            <category><![CDATA[artficial-intelligence]]></category>
            <dc:creator><![CDATA[Francisco Maria Calisto]]></dc:creator>
            <pubDate>Tue, 22 Oct 2019 09:16:23 GMT</pubDate>
            <atom:updated>2020-01-21T13:29:11.693Z</atom:updated>
            <content:encoded><![CDATA[<p>An advanced medical application solution comprises several routines and advanced image processing, as well as reading complexity. This article discloses a possible <a href="https://github.com/opprDev/medical-image-downloader">solution for a Medical Imaging Downloader</a> within the use of both <a href="https://cornerstonejs.org/">CornerstoneJS</a> and <a href="https://www.orthanc-server.com/">Orthanc</a> [1, 2] technologies.</p><figure><img alt="Medical Imaging Downloader for CornerstoneJS and Orthanc" src="https://cdn-images-1.medium.com/max/1024/1*v5tY9_9-ZDC4DFBpKo9YRw.png" /><figcaption>The clinical workflow support is provided by systems such as <a href="https://cornerstonejs.org/">CornerstoneJS</a> and <a href="https://www.orthanc-server.com/">Orthanc</a> to enable role-based processing of several patient cases.</figcaption></figure><h3>Introduction</h3><p>In the last article named “<a href="https://medium.com/oppr/using-cornerstonejs-and-orthanc-to-support-deep-learning-projects-c9675819c33a"><em>Using CornerstoneJS and Orthanc to Support Deep Learning Projects</em></a>”, we explain the importance of tools such as <a href="https://cornerstonejs.org/">CornerstoneJS</a> [3, 4] and <a href="https://www.orthanc-server.com/">Orthanc</a> [1, 2] to support our <a href="https://en.wikipedia.org/wiki/Artificial_intelligence"><strong>Artificial Intelligence (AI)</strong></a> algorithms. Briefly, these tools are of chief importance to our projects, allowing us to empower our clinicians to provide us with <em>datasets</em> by interacting with those tools [8]. In the end, the <em>datasets</em> will feed our AI models, providing us with the autonomous outputs that clinicians need [5, 6, 7]. The use of these tools will allow the application of <a href="https://en.wikipedia.org/wiki/Machine_learning"><strong>Machine Learning (ML)</strong></a> to the <a href="https://en.wikipedia.org/wiki/Medical_imaging"><strong>Medical Imaging (MI)</strong></a> field. This is addressed in our “<a href="https://medium.com/oppr/applying-new-paradigms-in-human-computer-interaction-to-health-informatics-86d6a7046130"><em>Applying New Paradigms in Human-Computer Interaction to Health Informatics</em></a>” article. The following description will be a technical, yet comprehensive piece of information for any kind of reader.</p><figure><img alt="Cloud-Based Solution" src="https://cdn-images-1.medium.com/max/300/1*HFHyqWhSvm0B9MIk6bVniQ.png" /><figcaption>System setup and architecture as a cloud-based solution. Communications between the <a href="https://cornerstonejs.org/">CornerstoneJS</a> and <a href="https://www.orthanc-server.com/">Orthanc</a>. [<a href="https://www.researchgate.net/publication/336699969_Medical_Imaging_Downloader_Cloud-Based_Solution_Flow">DOI: 10.13140/RG.2.2.28460.41605</a>]</figcaption></figure><h3><strong>Solution</strong></h3><p>A cloud-based solution like the presented architecture is showing a simplified way to retrieve and download sets of medical images. In our solution, a cloud-based platform to the proposed architecture was implemented and deployed as a <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/index.html">Demo</a> to serve the purpose of this article. The <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a> is presented on our <a href="https://github.com/opprDev">GitHub</a> page. Our <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a> is licensed over the MIT License. For more details and information, follow the <a href="https://github.com/opprDev/medical-image-downloader/wiki">wiki page of the repository</a>.</p><h3>Instructions</h3><p>From our <a href="https://github.com/opprDev/medical-image-downloader">repository</a>, first of all, you need to clone our <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a>. Then you need to follow the seral stages of our instructions to achieve both article and repository goals. The instructions are simple to follow, however, if you have any question feel free to ask in the comment section of this article or open a <a href="https://github.com/opprDev/medical-image-downloader/issues/new">new GitHub issue</a> so we can help you with that.</p><h4>Clone</h4><p>To clone the hereby repository, follow the guidelines below. It is easy as that.</p><p><strong>1.1.</strong> Please clone the repository by typing the command:</p><pre>git clone <a href="https://github.com/opprDev/medical-image-downloader.git">https://github.com/opprDev/medical-image-downloader.git</a></pre><p><strong>1.2.</strong> Get inside of the repository directory:</p><pre>cd medical-image-downloader/</pre><p><strong>1.3.</strong> To install and run of the source code, follow the next steps;</p><h4>Install</h4><p>The installation guidelines are as follows. Please, be sure that you follow them correctly.</p><p><strong>2.1.</strong> Run the following command to install the <a href="https://pip.pypa.io/en/stable/user_guide/">library</a> set using <a href="https://pypi.org/project/pip/">pip</a>:</p><pre>pip3 install -r requirements.txt</pre><p><strong>2.2.</strong> Follow the next step;</p><h4>Run</h4><p>The running guidelines are as follows. Please, be sure that you follow it correctly.</p><p><strong>3.1.</strong> Run the sample using the following command:</p><pre>python3 src/core/main.py</pre><p><strong>3.2.</strong> Enjoy our source code!</p><h3>Dependencies</h3><p>Our downloader was made in <a href="https://www.python.org/">Python</a>, which is easy to learn while being the go-to language for fast prototyping. For this job, we mainly used <a href="https://pandas.pydata.org/">pandas</a>, <a href="https://docs.python.org/3/library/json.html">json</a> and <a href="https://docs.python.org/3/library/urllib.html">urllib</a> libraries. We used <a href="https://pandas.pydata.org/">pandas</a> to manipulate our data processing and analysis. To read the several <a href="https://www.json.org/">JSON</a> files that manage <a href="https://cornerstonejs.org/">CornerstoneJS</a> the connection with <a href="https://www.orthanc-server.com/">Orthanc</a>, we used the <a href="https://docs.python.org/3/library/json.html">json</a> library. Finally, to fetch the connection between URLs (Uniform Resource Locators), we used the <a href="https://docs.python.org/3/library/urllib.html">urllib</a> library.</p><h3>Source Code</h3><p>As we stated in our <a href="https://medium.com/oppr/using-cornerstonejs-and-orthanc-to-support-deep-learning-projects-c9675819c33a">previous article</a>, on a client-server communication, the client-side of the <a href="https://cornerstonejs.org/">CornerstoneJS</a> searches for <strong>studies</strong> (cases) or <strong>series</strong> (digital slides) resources (see <a href="https://miro.medium.com/max/3816/1*k6_XGp32djI2zf0yTbNoyg.png">Figure</a>) while providing query parameters to filter DICOM objects based on given attribute values. To read the set of values, the <a href="https://docs.python.org/3/library/urllib.html">urllib</a> library fetches the URL, so that we can interpret the medical image structures. The schemas are in the <a href="https://www.json.org/">JSON</a> format. Therefore, we used the <a href="https://docs.python.org/3/library/json.html">json</a> library to interpret the schemes.</p><h4>Folder Structure</h4><p>Our project will have the following folders:</p><pre>src/<br>├── core/<br>├── methods/<br>├── notebooks/<br>├── tests/<br>└── variables/</pre><p>The full <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a> is inside the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src"><strong><em>src/</em></strong></a> folder. The folder has several other sub-folders organizing our code in sections. The first section is the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/core"><strong><em>core/</em></strong></a> sub-folder where we have the <a href="https://github.com/opprDev/medical-image-downloader/blob/master/src/core/main.py"><strong><em>main.py</em></strong></a> file. The <em>main.py</em> file is the principal file, working as coordinator. The <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/methods"><strong><em>methods/</em></strong></a> sub-folder is where we have the set of methods to download the medical images, or to support the process. As we are using <a href="https://jupyter.org/">Jupyter Notebook</a>, we created the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/notebooks"><strong><em>notebooks/</em></strong></a> sub-folder to place them in. Each project must have its own set of tests so that we can guarantee the validity of our <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a>. In the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/tests"><strong><em>tests/</em></strong></a> sub-folder, we placed our tests for the project and in the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/variables"><strong><em>variables/</em></strong></a> sub-folder, we have our used variables and constants.</p><h4>Methods</h4><p>Inside the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src"><strong><em>src/</em></strong></a> folder and inside the <a href="https://github.com/opprDev/medical-image-downloader/tree/master/src/methods"><strong><em>methods/</em></strong></a> sub-folder, we have the <a href="https://github.com/opprDev/medical-image-downloader/blob/master/src/methods/accessing.py"><strong><em>accessing.py</em></strong></a> file. This file has the main method to download the medical images that are shown in the <a href="https://cornerstonejs.org/">CornerstoneJS</a> viewer and stored on an <a href="https://www.orthanc-server.com/">Orthanc</a> server.</p><p>But first, let us explain how to achieve each functional goal. The first part is to request the URL by using the <a href="https://docs.python.org/3/library/urllib.html">urllib</a> library.</p><pre>dataStudyList = urllib.request.urlopen(mainServer).read()</pre><p>Secondly, we need to interpret the data (see <a href="https://miro.medium.com/max/3816/1*k6_XGp32djI2zf0yTbNoyg.png">Figure</a>) coming from the <em>request</em>. For that, we used the <a href="https://docs.python.org/3/library/json.html">json</a> library which will read the answer coming from the <a href="https://www.orthanc-server.com/">Orthanc</a> server, loading it to a <a href="https://www.w3schools.com/js/js_object_definition.asp">JavaScript object</a>.</p><pre>outputStudyList = json.loads(dataStudyList)</pre><p>Now, we are on the way to trigger the schemas of the medical images. By looking at our <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/index.html">Demo</a> and <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Prototype Cornerstone</a> examples, there are two important endpoints: (1) the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studyList.json"><strong><em>studyList.json</em></strong></a> file; and (2) the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/"><strong><em>studies/</em></strong></a> folder. The <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studyList.json"><strong><em>studyList.json</em></strong></a> file is where we have the patients’ list. While the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/"><strong><em>studies/</em></strong></a> folder is where we have the details for each patient.</p><p>As an example of the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studyList.json"><strong><em>studyList.json</em></strong></a> file, we have the following:</p><pre>{<br>    &quot;studyList&quot;: [<br>        {<br>            &quot;patientName&quot;: &quot;MR/BRAIN/GRASE/1024&quot;,<br>            &quot;patientId&quot;: &quot;7&quot;,<br>            &quot;internalId&quot;: &quot;8f14e4&quot;,<br>            &quot;studyDate&quot;: &quot;19950330&quot;,<br>            &quot;modality&quot;: &quot;BRAIN&quot;,<br>            &quot;studyDescription&quot;: &quot;BRAIN&quot;,<br>            &quot;numImages&quot;: 1,<br>            &quot;studyId&quot;: &quot;7&quot;<br>        },<br>        {<br>            &quot;patientName&quot;: &quot;Anonymized&quot;,<br>            &quot;patientId&quot;: &quot;0&quot;,<br>            &quot;internalId&quot;: &quot;cfcd20&quot;,<br>            &quot;studyDate&quot;: &quot;20070101&quot;,<br>            &quot;modality&quot;: &quot;Knee (R)&quot;,<br>            &quot;studyDescription&quot;: &quot;Knee (R)&quot;,<br>            &quot;numImages&quot;: 71,<br>            &quot;studyId&quot;: &quot;0&quot;<br>        },<br>        {<br>            &quot;patientName&quot;: &quot;Case1&quot;,<br>            &quot;patientId&quot;: &quot;Case1&quot;,<br>            &quot;internalId&quot;: &quot;a294f5&quot;,<br>            &quot;studyDate&quot;: &quot;20080408&quot;,<br>            &quot;modality&quot;: &quot;BREAST IMAGING TOMOSYNTHESIS&quot;,<br>            &quot;studyDescription&quot;: &quot;BREAST IMAGING TOMOSYNTHESIS&quot;,<br>            &quot;numImages&quot;: 6,<br>            &quot;studyId&quot;: &quot;Case1&quot;<br>        }<br>    ]<br>}</pre><p>As an example of the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/"><strong><em>studies/</em></strong></a> folder and more specifically of a patient case from <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/Case1.json"><strong><em>Case1.json</em></strong></a> file, we have the following:</p><pre>{<br>    &quot;patientName&quot;: &quot;Case1&quot;,<br>    &quot;patientId&quot;: &quot;Case1&quot;,<br>    &quot;internalId&quot;: &quot;a294f5&quot;,<br>    &quot;studyDate&quot;: &quot;20080408&quot;,<br>    &quot;modality&quot;: &quot;BREAST IMAGING TOMOSYNTHESIS&quot;,<br>    &quot;studyDescription&quot;: &quot;BREAST IMAGING TOMOSYNTHESIS&quot;,<br>    &quot;numImages&quot;: 1,<br>    &quot;studyId&quot;: &quot;dec7949e-a99467d6-305d0d64-424bfd75-8a1579ab&quot;,<br>    &quot;seriesList&quot;: [<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;72100000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 29,<br>                    &quot;imageId&quot;: &quot;dd03786b-6cc667d1-a65fa0e9-34101126-6f82ca59/file&quot;<br>                }<br>            ]<br>        },<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;72100000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 15,<br>                    &quot;imageId&quot;: &quot;ae407aac-af2c66a3-5801a6c9-ffff4ae9-06496953/file&quot;<br>                }<br>            ]<br>        },<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;73100000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 16,<br>                    &quot;imageId&quot;: &quot;062ce5a8-0d015abb-73f65ae3-f95d55e8-9cbdfcb6/file&quot;<br>                }<br>            ]<br>        },<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;73100000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 30,<br>                    &quot;imageId&quot;: &quot;6f6f0dc7-db7a6fa0-f3e46a80-8bcda478-194bcf11/file&quot;<br>                }<br>            ]<br>        },<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;73200000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 30,<br>                    &quot;imageId&quot;: &quot;e0551a96-ccf93766-a2d57ddf-361c8462-46b8b196/file&quot;<br>                }<br>            ]<br>        },<br>        {<br>            &quot;seriesDescription&quot;: &quot;MG&quot;,<br>            &quot;seriesNumber&quot;: &quot;73200000&quot;,<br>            &quot;instanceList&quot;: [<br>                {<br>                    &quot;instanceNumber&quot;: 16,<br>                    &quot;imageId&quot;: &quot;88653231-d6fba822-1af594af-e6f5b30d-dd1ab7f2/file&quot;<br>                }<br>            ]<br>        }<br>    ]<br>}</pre><p>To better understand the schema structure it is highly important to read our <a href="https://medium.com/oppr/using-cornerstonejs-and-orthanc-to-support-deep-learning-projects-c9675819c33a">last article</a>, as mentioned previously. Unless you are already familiar with this <a href="https://en.wikipedia.org/wiki/Picture_archiving_and_communication_system">PACS</a> structure. To describe the above structures, in short, a <em>seriesList</em> array (see <a href="https://miro.medium.com/max/3816/1*k6_XGp32djI2zf0yTbNoyg.png">Figure</a>) has a list of <em>series</em>. Each one of the series has a <em>seriesDescription</em>, a <em>seriesNumber</em> and a list of <em>instanceList</em> as arrays with several <em>instances</em> of images.</p><p>Now, for each patient, we need to read the respective information of the object. For that, we just go through the <em>studyList</em> of the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studyList.json"><strong><em>studyList.json</em></strong></a> file and extract the <em>patientId</em> to retrieve patient <em>study</em> on the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/"><strong><em>studies/</em></strong></a> folder. In our example, we use the <em>“patientName”: “</em><strong><em>Case1</em></strong><em>”</em> to retrieve the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/Case1.json"><strong><em>Case1.json</em></strong></a> file.</p><pre>for ptnt in range(len(outputStudyList[&#39;studyList&#39;])):<br>    patientIdToCompare = outputStudyList[&#39;studyList&#39;][ptnt][&#39;patientId&#39;]<br>    pntFileOnServer = lnk004 + patientIdToCompare + ext003<br>    dataStudies = urllib.request.urlopen(pntFileOnServer).read()<br>    outputStudies = json.loads(dataStudies)</pre><p>From there, we can read the information inside the <a href="http://breastscreening.isr.tecnico.ulisboa.pt:8982/src/studies/Case1.json"><strong><em>Case1.json</em></strong></a> file, also as a <a href="https://www.w3schools.com/js/js_object_definition.asp">JavaScript object</a>. Now, we want to parse the <em>seriesList</em>, passing through the <em>instanceList</em> to get the <em>imageId</em>. From the <em>imageId</em>, we can retrieve and download each image.</p><pre>for study in range(len(outputStudies)):<br>      seriesList = outputStudies[&#39;seriesList&#39;]<br>      for serie in range(len(seriesList)):<br>        seriesNumber = seriesList[serie][&#39;seriesNumber&#39;]<br>        instanceList = seriesList[serie][&#39;instanceList&#39;]<br>        for instance in range(len(instanceList)):<br>          image_counter = image_counter + 1<br>          imageId = instanceList[instance][&#39;imageId&#39;]<br>          dcmUrl = dicomServer + imageId<br>          dcmFileName = folderToSave + str(image_counter) + ext002<br>          urllib.request.urlretrieve(dcmUrl, dcmFileName)</pre><p>Final <em>method</em> <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a> of downloading the medical images:</p><pre>def dwnldMainServImgStorOnDicomServ(folderToSave, mainServer, dicomServer):<br>  &#39;&#39;&#39;<br>  Downloading all medical images from your main server<br>  that are stored on a DICOM server.<br>  &#39;&#39;&#39;<br>  image_counter = 10000000<br>  dataStudyList = urllib.request.urlopen(mainServer).read()<br>  outputStudyList = json.loads(dataStudyList)<br>  for ptnt in range(len(outputStudyList[&#39;studyList&#39;])):<br>    print(c010)<br>    patientIdToCompare = outputStudyList[&#39;studyList&#39;][ptnt][&#39;patientId&#39;]<br>    pntFileOnServer = lnk004 + patientIdToCompare + ext003<br>    dataStudies = urllib.request.urlopen(pntFileOnServer).read()<br>    outputStudies = json.loads(dataStudies)<br>    print(outputStudies)<br>    for study in range(len(outputStudies)):<br>      seriesList = outputStudies[&#39;seriesList&#39;]<br>      for serie in range(len(seriesList)):<br>        seriesNumber = seriesList[serie][&#39;seriesNumber&#39;]<br>        instanceList = seriesList[serie][&#39;instanceList&#39;]<br>        for instance in range(len(instanceList)):<br>          image_counter = image_counter + 1<br>          imageId = instanceList[instance][&#39;imageId&#39;]<br>          print(msg004, imageId, msg005, seriesNumber)<br>          dcmUrl = dicomServer + imageId<br>          dcmFileName = folderToSave + str(image_counter) + ext002<br>          urllib.request.urlretrieve(dcmUrl, dcmFileName)<br>    print(c010)</pre><h3>Conclusions</h3><p>In this article, we propose a solution to retrieve and download sets of medical images that are shown on a <a href="https://cornerstonejs.org/">CornerstoneJS</a> viewer and stored on an <a href="https://www.orthanc-server.com/">Orthanc</a> server. We provide a clear and simple <a href="https://github.com/opprDev/medical-image-downloader"><strong><em>source code</em></strong></a> well described and documented in this article, as well as on the <a href="https://github.com/opprDev/medical-image-downloader/wiki">wiki page of the repository</a>. Furthermore, we show how this solution enables the application of <a href="https://en.wikipedia.org/wiki/Machine_learning"><strong>ML</strong></a> methods to the <a href="https://en.wikipedia.org/wiki/Medical_imaging"><strong>MI</strong></a> field and how to promote such kind of research projects.</p><h3>Acknowledgments</h3><p>This post is supported by the <a href="https://github.com/MIMBCD-UI/meta/wiki">case studies</a> of <a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a>, <a href="https://mida-project.github.io/">MIDA</a>, and <a href="https://breastscreening.github.io/">BreastScreening</a> projects at <a href="http://tecnico.ulisboa.pt/">IST</a> from <a href="https://www.ulisboa.pt/">ULisboa</a>. The three projects are strongly sponsored by <a href="https://www.fct.pt/">FCT</a>, a <a href="https://www.portugal.gov.pt/">Portuguese</a> public agency that promotes science, technology, and innovation, in all scientific domains. The <a href="https://breastscreening.github.io/">BreastScreening</a> project is an <a href="https://cs.adelaide.edu.au/~dlmmbia/">ARC Discovery Project (DP140102794)</a> in collaboration with <a href="http://tecnico.ulisboa.pt/">IST</a>, <a href="http://www.adelaide.edu.au/">Adelaide</a>, and <a href="https://www.uq.edu.au/">UQueensland</a>. The genesis of this post was a research work between <a href="http://welcome.isr.tecnico.ulisboa.pt/">ISR-Lisboa</a> and <a href="https://iti.larsys.pt/">ITI</a>, both associated laboratories of <a href="http://larsys.pt/">LARSyS</a>. From these institutions, I would like to convey a special thanks to Professor <a href="http://users.isr.ist.utl.pt/~jan/">Jacinto C. Nascimento</a> and Professor <a href="http://nunonunes.info/">Nuno Nunes</a> for advising me during my research work. Last but not least, I would like to thank several important people of this noble organization called <a href="https://oppr.io/">oppr</a>. A special thanks to <a href="https://www.linkedin.com/in/gustavo-passos-de-gouveia-a8300764/">Gustavo Passos de Gouveia</a>, <a href="https://github.com/bruno-oliveira">Bruno Oliveira</a>, <a href="https://medium.com/@joaobernardomadeiracampos">João Campos</a>, <a href="https://www.linkedin.com/in/carlotagalvaodemelo/">Carlota Galvão de Melo</a> and <a href="https://github.com/jrcruz">João Cruz</a> for reviewing this article giving me great inputs. Last but not least, a special thanks to <a href="https://www.linkedin.com/in/chafey/">Chris Hafey</a>, the propelling person of <a href="https://cornerstonejs.org/">CornerstoneJS</a>, who also developed the <a href="https://github.com/chafey/cornerstoneDemo">cornerstoneDemo</a>. Not forgetting the three supporters of the <a href="https://cornerstonejs.org/">CornerstoneJS</a> library, <a href="https://www.linkedin.com/in/alois-dreyfus">Aloïs Dreyfus</a>, <a href="http://dannyrb.com/">Danny Brown</a>, and <a href="https://www.npmjs.com/~swederik">Erik Ziegler</a>. We also would like to give a special thanks to <a href="https://www.npmjs.com/~swederik">Erik Ziegler</a> who supports several <a href="https://groups.google.com/forum/#!forum/cornerstone-platform">issues</a> during this path. In the end, a great thank to all the Orthanc project team, but especially to <a href="https://www.linkedin.com/in/jodogne/">Sébastien Jodogne</a>.</p><h3>Supporters</h3><p>Our organization is a non-profit organization. However, we have many expenses across our activity. From infrastructure to service expenses, we need some money, as well as help, to support our team and projects. For the expenses, we created several channels that will mitigate this problem. First of all, you can support us by being one of our <a href="https://patreon.com/oppr">Patreons</a>. Second, you can support us on the <a href="https://opencollective.com/oppr">Open Collective</a> page. Thirdly, you can <a href="https://ko-fi.com/opprGroup">buy one coffee</a> (or more) for us. Fourth, you can also support us on our <a href="https://liberapay.com/oppr">Liberapay</a> page. Last but not least, you can directly support us on <a href="https://paypal.me/oppr">PayPal</a>. On the other hand, we also need help in the development of our projects. Therefore, if you have the knowledge we welcome you to support our projects (<em>e.g.</em>, <a href="https://github.com/opprGroup/meta">GitHub</a>). Just follow our channels and repositories.</p><h3>Social</h3><p>We are on the top social media networks. The links are as follows. Also, you can chat with us. Just follow us to know about our work and news.</p><p>List of our social media networks and chats channels:</p><p><a href="https://facebook.com/opprGroup">Facebook</a> | <a href="https://twitter.com/opprGroup">Twitter</a> | <a href="https://www.linkedin.com/company/oppr">LinkedIn</a> | <a href="https://gitter.im/opprTeam">Gitter</a> | <a href="https://www.youtube.com/user/callitech">YouTube</a></p><h3>References</h3><p>[1] Jodogne, S., Bernard, C., Devillers, M., Lenaerts, E. and Coucke, P., 2013, April. Orthanc-A lightweight, restful DICOM server for healthcare and medical research. In 2013 IEEE 10th International Symposium on Biomedical Imaging (pp. 190–193). IEEE.</p><p>[2] Jodogne, S., 2018. The Orthanc Ecosystem for Medical Imaging. Journal of digital imaging, 31(3), pp.341–352.</p><p>[3] Hostetter, J., Khanna, N. and Mandell, J.C., 2018. Integration of a zero-footprint cloud-based picture archiving and communication system with customizable forms for radiology research and education. Academic Radiology, 25(6), pp.811–818.</p><p>[4] Sedghi, A., Hamidi, S., Mehrtash, A., Ziegler, E., Tempany, C., Pieper, S., Kapur, T. and Mousavi, P., 2019, March. Tesseract-medical imaging: an open-source browser-based platform for artificial intelligence deployment in medical imaging. In Medical Imaging 2019: Image-Guided Procedures, Robotic Interventions, and Modeling (Vol. 10951, p. 109511R). International Society for Optics and Photonics.</p><p>[5] Calisto, F.M., Lencastre, H., Nunes, N.J. and Nascimento, J.C., Medical Imaging Diagnosis Assistant: AI-Assisted Radiomics Framework User Validation.</p><p>[6] Calisto, F.M., Lencastre, H., Nunes, N.J. and Nascimento, J.C., Medical Imaging Diagnosis Assistant: AI-Assisted Radiomics Framework User Validation.</p><p>[7] Calisto, F.M., Miraldo, P., Nunes, N. and Nascimento, J.C., BreastScreening: A Multimodality Diagnostic Assistant.</p><p>[8] Calisto, F.M., Ferreira, A., Nascimento, J.C. and Gonçalves, D., 2017, October. Towards Touch-Based Medical Image Diagnosis Annotation. In Proceedings of the 2017 ACM International Conference on Interactive Surfaces and Spaces (pp. 390–395). ACM.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d08c3a508d9b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/oppr/medical-imaging-downloader-for-cornerstonejs-and-orthanc-d08c3a508d9b">Medical Imaging Downloader for CornerstoneJS and Orthanc</a> was originally published in <a href="https://medium.com/oppr">oppr</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using CornerstoneJS and Orthanc to Support Deep Learning Projects]]></title>
            <link>https://medium.com/oppr/using-cornerstonejs-and-orthanc-to-support-deep-learning-projects-c9675819c33a?source=rss----9de3d09b6cbe---4</link>
            <guid isPermaLink="false">https://medium.com/p/c9675819c33a</guid>
            <category><![CDATA[humancomputer-interaction]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[health-informatics]]></category>
            <category><![CDATA[medical-imaging]]></category>
            <category><![CDATA[deep-learning]]></category>
            <dc:creator><![CDATA[Francisco Maria Calisto]]></dc:creator>
            <pubDate>Mon, 24 Jun 2019 19:38:51 GMT</pubDate>
            <atom:updated>2019-10-21T09:41:07.378Z</atom:updated>
            <content:encoded><![CDATA[<p>Implementing and improving tools to support research about the automated analysis of medical images. Our projects feature several datasets and tools that will be important for our Artificial Intelligence novel methods across Medical Imaging applicability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*Q8Db2atQhgBmFKsNmCjOYQ.png" /><figcaption>Source: <a href="https://www.gomolldesign.com/">Gomoll Research + Design</a> of the <a href="https://www.gomolldesign.com/ge-healthcare">GE Healthcare User Interface (UI) consistency efforts</a> post.</figcaption></figure><h3>Introduction</h3><p>Artificial Intelligence (AI) has the potential to alter many application domains fundamentally [5, 6, 7]. One prominent example is clinical radiology as <a href="https://www.radiomics.io/">radiomics</a>. The literature hypothesizes that <a href="https://medium.com/zylapp/review-of-deep-learning-algorithms-for-image-classification-5fdbca4a05e2">Deep Learning Algorithms (DLA)</a> will profoundly affect the clinical workflow [8]. Indeed, on an early story titled as “<a href="https://medium.com/oppr/applying-new-paradigms-in-human-computer-interaction-to-health-informatics-86d6a7046130">Applying New Paradigms in Human-Computer Interaction to Health Informatics</a>”, we already underline the new paradigms of Health Informatics (HI) and AI applicabilities.</p><p>In our projects, the final main goal is to develop an improved <a href="https://link.springer.com/chapter/10.1007/978-1-4471-4474-8_22">Clinical Decision Support System (CDSS)</a> [9, 10], as well as several <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> and <a href="https://github.com/BreastScreening/meta/wiki/Tools">tools</a> supporting the main goal. To do this, three principal projects were created: (1) <a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a>; (2) <a href="https://mida-project.github.io/">MIDA</a>; and (3) <a href="https://breastscreening.github.io/">BreastScreening</a>. First of all, the <a href="https://www.researchgate.net/project/Medical-Imaging-Multimodality-Breast-Cancer-Diagnosis-User-Interface">Medical Imaging Multimodality Breast Cancer Diagnosis User Interface</a> (<a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a>) project aims to develop several systems (<a href="https://github.com/BreastScreening/meta/wiki/Prototypes">prototypes</a>), as Proofs-Of-Concept for breast cancer diagnosis. With this, we improve the final system <a href="https://www.researchgate.net/publication/331687177_Medical_Imaging_Multimodality_Breast_Cancer_Diagnosis_User_Interface">User Interface (UI) and Information Visualization (InfoVis)</a> with novel interaction techniques and visualizations. The second project, that we call <a href="https://www.researchgate.net/project/Medical-Imaging-Diagnosis-Assistant">Medical Imaging Diagnosis Assistant</a> (<a href="https://mida-project.github.io/">MIDA</a>), aims to develop an assistant that will help the automation of cancer diagnosis. Finally, the last project is called <a href="https://breastscreening.github.io/">BreastScreening</a>. This project aims to use <a href="http://libccv.org/doc/doc-convnet/">Deep ConvNet</a> models pre-trained from our <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> on medical image analysis applications, and the use of <a href="http://libccv.org/doc/doc-convnet/">Deep ConvNet</a> to analyze unregistered medical images.</p><p>In this story, we will describe our projects and developed tools which will allow several Deep Learning (DL) techniques regarding the Medical Imaging (MI) domain. We <a href="https://github.com/opprDev/medical-image-downloader">developed a tool that allows to retrieve all used medical images</a> provided on a <a href="https://www.sciencedirect.com/science/article/pii/S0009926005002199">DICOM viewer</a> and available from a <a href="https://ieeexplore.ieee.org/abstract/document/6556444/">standalone DICOM server</a>. <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Our DICOM viewer prototype</a> was developed thanks to the <a href="https://cornerstonejs.org/">CornerstoneJS</a> library [3, 4] and is highly based on their <a href="https://github.com/chafey/cornerstoneDemo">CornerstoneDemo</a> version, developed by <a href="https://github.com/chafey/cornerstoneDemo">Chris Hafey</a> and improved by us. On the other hand, for our projects, we need a <strong>lightweight</strong> and <strong>standalone</strong> solution to store our medical images. That said, we used the <a href="https://www.orthanc-server.com/">Orthanc</a> servers [1, 2] providing us a <strong>RESTful API</strong>.</p><h3>Medical Imaging Platforms</h3><p>In our example, we are using the <a href="https://www.orthanc-server.com/">Orthanc</a> server [1, 2] version for <a href="https://www.orthanc-server.com/static.php?page=download-mac">OS X</a>, but you can use <a href="https://www.orthanc-server.com/download.php">other versions</a>. The result will be the same. But first, we will introduce and explain how does it work.</p><h4>How does a DICOM server work?</h4><p>A DICOM server, like Orthanc its built-in <a href="https://book.orthanc-server.com/users/rest.html">RESTful API</a>, that can be used to drive it from external applications, for instance, the <a href="https://cornerstonejs.org/">CornerstoneJS</a> set of tools and originated platforms. The Orthanc uses several DICOM tags from the stored medical images in the <a href="https://www.json.org/">JSON</a> file format. The structures for the stored DICOM resources are fourfold <a href="https://book.orthanc-server.com/dicom-guide.html#dicom-identifiers">identifiers</a>: (1) Patient; (2) Study; (3) Series; and (4) Instance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*I-NQnDWNDqf_x0Vx1WYCLw.png" /><figcaption>This diagram shows that a given patient benefits from a set of medical imaging studies. Each study is made from a set of series. Each series is, in turn, a set of instances. [<a href="https://www.researchgate.net/publication/333984812_Medical_Imaging_Patient_Structure">DOI: 10.13140/RG.2.2.19014.32321</a>]</figcaption></figure><p>The medical information encoded by a DICOM file is called a <em>dataset</em> and takes the form of an associative array. Each value can itself be a list of <em>datasets</em>, leading to a hierarchical data structure that is much like a JSON file. In the DICOM terminology, each key is called a DICOM tag. The list of the standard DICOM tags is normalized by an official dictionary. For improved readability, it is also common to name these DICOM tags (<em>e.g.</em>, “PatientName” or “StudyDescription”). The standard associates each DICOM tag with a data type, that is known as its value representation.</p><h4>How does a DICOM viewer work?</h4><p>As said before, we are using a set of <a href="https://github.com/cornerstonejs/cornerstoneTools">tools</a> from the <a href="https://cornerstonejs.org/">CornerstoneJS</a> library. From those <a href="https://github.com/cornerstonejs/cornerstoneTools">tools</a>, we created the <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Cornerstone Prototype</a> which is a platform that will be of chief importance to view and generate various sets of data regarding our Medical Imaging (MI) projects. The data, in the format of <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a>, will be provided to our <a href="https://medium.com/zylapp/review-of-deep-learning-algorithms-for-image-classification-5fdbca4a05e2">Deep Learning Algorithms (DLA)</a>. However, in this post, we will not focus on it for now.</p><p><a href="https://cornerstonejs.org/">CornerstoneJS</a> is a lightweight <a href="https://www.javascript.com/">JavaScript</a> library for displaying medical images. The library is not restricted to a particular pixel container or transport mechanism and does not define an interaction paradigm. As such, it may be adapted to any image format using an extensible image loader class and can utilize present and future transport protocols and user interaction techniques. It is why we chose this library and created the <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Cornerstone Prototype</a> as a Proof-Of-Concept tool to potential generate our <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a><em>.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WYM789SmbrizAqX2o4GV3A.png" /><figcaption>Image retrieval from the Orthanc Servers to the DICOM viewer. In this figure, the DICOM viewer is supported by both <a href="https://cornerstonejs.org/">CornerstoneJS</a> library and <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Cornerstone Prototype</a>. Now, Radiologists can annotate each lesion providing this data to the Deep Learning Algorithms. [<a href="https://www.researchgate.net/publication/333984796_Medical_Imaging_Cornerstone_Prototype_User_Workflow">DOI: 10.13140/RG.2.2.27402.93126</a>]</figcaption></figure><p>The present <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Cornerstone Prototype</a> extends the native <strong>RESTful API</strong> of Orthanc with a reference implementation of the DICOM standard. The client searches for <strong>studies</strong> (cases) or <strong>series</strong> (digital slides) resources while providing query parameters to filter DICOM objects based on given attribute values. The server responds with resource representations for each matched object in JSON format according to the DICOM JSON model. The client retrieves image metadata resources from the server for each image (resolution level) belonging to the matched study or series through a request.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k6_XGp32djI2zf0yTbNoyg.png" /><figcaption>Sample diagram of the DICOM meta <em>tag</em> structures essential for retrieving data. Further information can be found <a href="https://github.com/MIMBCD-UI/prototype-cornerstone/wiki/Image-Processing">here</a> on our project repository. The CornerstoneJS link each tag to show the right image on the viewer. [<a href="https://www.researchgate.net/publication/333984856_Medical_Imaging_Meta_Tag_Structures_Sample_Diagram">DOI: 10.13140/RG.2.2.34113.81762</a>]</figcaption></figure><p>As we can see in the above image, medical images come with lots of non-pixel-wise metadata, such as the pixel spacing of the image, the <strong>Patient ID</strong>, or the scan acquisition date. With DICOM file, this information is stored within the file header and can be read and parsed, as well as passed around the application. However, it is common to provide metadata independently from the transmission of pixel data from the server to the client since this can considerably improve performance. To handle these scenarios, our prototype provides infrastructure for the definition and usage of <em>Metadata Providers</em>. <em>Metadata Providers</em> are simply functions which take in an <strong>Image ID</strong> and specified metadata type and return the metadata itself. This metadata information will be important for our lesion annotations. Therefore, it will be part of the information consumed by the Deep Learning Algorithms (DLA).</p><h3>Image Annotations</h3><p>A key distinguishing feature of projects is its support for standardized formats for image annotations. Already referenced before, the image metadata also includes information about the image, such as the name of the imaging procedure and how or when the image was acquired. Our system supports controlled terminologies, enabling semantic interoperability. In the use case of cancer lesion annotation, the value of our projects is <strong>recording lesion identifiers</strong>, <strong>anatomic locations of lesions</strong>, <strong>lesion</strong> and <strong>study types</strong>. This semantic information is critical for automating the generation of tabular summaries of lesions. Also, it enables automating comparing the response assessment in patients according to different imaging biomarkers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9wYvJPv3PB64xG0BHAACCQ.png" /><figcaption>The viewer and annotation window. Images are displayed in the web viewer and the Radiologist records<br>image annotations in using drawing tools and an annotation window. [<a href="https://www.researchgate.net/publication/333984786_Viewer_and_Annotation_Window">DOI: 10.13140/RG.2.2.13981.15846</a>]</figcaption></figure><p>As the Radiologist makes annotations on images in the viewer, it creates JSON files. All annotations are stored in several JSON <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a>. The annotation <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> are accessible via functions from our several <a href="https://github.com/BreastScreening/meta/wiki/Tools">tools</a>. These are the <em>queried</em> to track lesions and summarize changes in the cancer treatment response.</p><h3>Discussion</h3><p>Many research studies that require viewing and annotating radiology images for making measurements of lesions or extracting <a href="https://www.radiomics.io/">radiomics</a> features [11] from them could benefit from using our Deep Learning Algorithms (DLA) and techniques. The use of the <a href="https://cornerstonejs.org/">CornerstoneJS</a> library and <a href="https://www.orthanc-server.com/">Orthanc</a> servers across our <a href="https://github.com/MIMBCD-UI/prototype-cornerstone">Cornerstone Prototype</a> and projects facilitates the collection of annotations and measurements targeting lesions in compliance with imaging standards.</p><p>From the acquired annotations, the calculated feature values on the JSON files can be used in applications, such as training a <a href="https://towardsdatascience.com/review-densenet-image-classification-b6631a8ef803">DenseNet</a> model [12] or <a href="https://l.messenger.com/l.php?u=https%3A%2F%2Fmedium.com%2Foppr%2Fan-initial-approach-to-feed-forward-neural-networks-3cb766feacb3&amp;h=AT3DHtxlo2p6ZEP3OCfzWUSIu8jsYjBP1G436u8QZ3CO_3YIEIr-6fl84LCmyjICfFH8vV28nVWDJkqMxseVrCL2yy3E6DuTIbk0EDkXThBscT5PMuUrXasR6j_-8V4maCg3sN9-">another kind of Machine Learning models</a>. At this stage, we just need: (i) <strong>medical images</strong> to train the models and for <strong>image segmentation</strong>; and (ii) <strong>lesion annotations</strong> provided by Radiologists (i.e., <a href="https://medium.com/oppr/applying-new-paradigms-in-human-computer-interaction-to-health-informatics-86d6a7046130">Human-In-The-Loop</a>) so that the model can learn. By computing a variety of image biomarkers on cohorts of patients, we can accumulate a substantial amount of data. This data can permit studies comparing the effectiveness of different imaging biomarkers as indicators of cancer treatment response. In the end, we will enhance clinicians’ workflow, as well as improved diagnosis. Thus, providing patients with better healthcare.</p><h3>Conclusions</h3><p>For our projects (<em>i.e.</em>, <a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a>, <a href="https://mida-project.github.io/">MIDA</a>, and <a href="https://breastscreening.github.io/">BreastScreening</a>), we are trying to implement and improve a novel Clinical Decision Support System (CDSS) serving from several <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> and <a href="https://github.com/BreastScreening/meta/wiki/Tools">tools</a>. Those <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> and <a href="https://github.com/BreastScreening/meta/wiki/Tools">tools</a> are implemented in order to reflect our requirements and clinicians’ needs on the main goal. We also used external tools, such as <a href="https://cornerstonejs.org/">CornerstoneJS</a> library and <a href="https://www.orthanc-server.com/">Orthanc</a> servers to promote the beginning of these projects. So that we can generate our <a href="https://github.com/BreastScreening/meta/wiki/Datasets"><em>datasets</em></a> while providing them to the <a href="http://libccv.org/doc/doc-convnet/">Deep ConvNet</a> [14] models on medical image analysis applications.</p><p>On the next post, titled as “<a href="https://medium.com/oppr/medical-imaging-downloader-for-cornerstonejs-and-orthanc-d08c3a508d9b">Medical Imaging Downloader for CornerstoneJS and Orthanc</a>”, we will describe our implementation of a medical imaging downloader for both <a href="https://cornerstonejs.org/">CornerstoneJS</a> and <a href="https://www.orthanc-server.com/">Orthanc</a> platforms. This tool implementation will retrieve and provide the stored DICOM images on an <a href="https://www.orthanc-server.com/">Orthanc</a> server addressed on a <a href="https://cornerstonejs.org/">CornerstoneJS</a> based platform. Such a tool was important during the entire project. Therefore, we want to share it with all the community. For instance, we use these set of tools for the work done in proceedings of conference publications such as the “<a href="https://dl.acm.org/citation.cfm?id=3134111">Towards Touch-Based Medical Image Diagnosis Annotation</a>” paper at <a href="https://iss2017.acm.org/">ISS’17</a>.</p><h3>Acknowledgments</h3><p>This post is supported by the <a href="https://github.com/MIMBCD-UI/meta/wiki">case studies</a> of <a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a>, <a href="https://mida-project.github.io/">MIDA</a>, and <a href="https://breastscreening.github.io/">BreastScreening</a> projects at <a href="http://tecnico.ulisboa.pt/">IST</a> from <a href="https://www.ulisboa.pt/">ULisboa</a>. The three projects are strongly sponsored by <a href="https://www.fct.pt/">FCT</a>, a <a href="https://www.portugal.gov.pt/">Portuguese</a> public agency that promotes science, technology, and innovation, in all scientific domains. The <a href="https://breastscreening.github.io/">BreastScreening</a> project is an <a href="https://cs.adelaide.edu.au/~dlmmbia/">ARC Discovery Project (DP140102794)</a> in collaboration with <a href="http://tecnico.ulisboa.pt/">IST</a>, <a href="http://www.adelaide.edu.au/">UAdelaide</a>, and <a href="https://www.uq.edu.au/">UQueensland</a>. The genesis of this post was a research work between <a href="http://welcome.isr.tecnico.ulisboa.pt/">ISR-Lisboa</a> and <a href="https://iti.larsys.pt/">ITI</a>, both associated laboratories of <a href="http://larsys.pt/">LARSyS</a>. From these institutions, I would like to convey a special thanks to Professor <a href="http://users.isr.ist.utl.pt/~jan/">Jacinto C. Nascimento</a> and Professor <a href="http://nunonunes.info/">Nuno Nunes</a> for advisor me during my research work. Last but not least, I would like to thank several important people of this noble organization called <a href="https://oppr.io/">oppr</a>. A special thanks to <a href="https://www.linkedin.com/in/gustavo-passos-de-gouveia-a8300764/">Gustavo Passos de Gouveia</a>, <a href="https://github.com/bruno-oliveira">Bruno Oliveira</a>, <a href="https://medium.com/@joaobernardomadeiracampos">João Campos</a>, and <a href="https://medium.com/@bruno.filipe.silva.dias">Bruno Dias</a> for reviewing this article giving me great inputs. Last but not least, a special thanks to <a href="https://www.linkedin.com/in/chafey/">Chris Hafey</a>, the propelling person of <a href="https://cornerstonejs.org/">CornerstoneJS</a>, who also developed the <a href="https://github.com/chafey/cornerstoneDemo">cornerstoneDemo</a>. Not forgetting the three supporters of the <a href="https://cornerstonejs.org/">CornerstoneJS</a> library, <a href="https://www.linkedin.com/in/alois-dreyfus">Aloïs Dreyfus</a>, <a href="http://dannyrb.com/">Danny Brown</a>, and <a href="https://www.npmjs.com/~swederik">Erik Ziegler</a>. We also would like to give a special thanks to <a href="https://www.npmjs.com/~swederik">Erik Ziegler</a> who supports several <a href="https://groups.google.com/forum/#!forum/cornerstone-platform">issues</a> during this path. In the end, a great thank to all Orthanc project team, but especially to <a href="https://www.linkedin.com/in/jodogne/">Sébastien Jodogne</a>.</p><h3>Supporters</h3><p>Our organization is a non-profit organization. However, we have many expenses across our activity. From infrastructure to service expenses, we need some money, as well as help, to support our team and projects. For the expenses, we created several channels that will mitigate this problem. First of all, you can support us by being one of our <a href="https://patreon.com/oppr">Patreons</a>. Second, you can support us on <a href="https://opencollective.com/oppr">Open Collective</a> page. Thirdly, you can <a href="https://ko-fi.com/opprGroup">buy one coffee</a> (or more) for us. Fourth, you can also support us on our <a href="https://liberapay.com/oppr">Liberapay</a> page. Last but not least, you can directly support us on <a href="https://paypal.me/oppr">PayPal</a>. On the other hand, we also need help in the development of our projects. Therefore, if you have the knowledge we welcome you to support our projects. Just follow our channels and repositories.</p><h3>References</h3><p>[1] Jodogne, S., Bernard, C., Devillers, M., Lenaerts, E. and Coucke, P., 2013, April. Orthanc-A lightweight, restful DICOM server for healthcare and medical research. In 2013 IEEE 10th International Symposium on Biomedical Imaging (pp. 190–193). IEEE.</p><p>[2] Jodogne, S., 2018. The Orthanc Ecosystem for Medical Imaging. Journal of digital imaging, 31(3), pp.341–352.</p><p>[3] Hostetter, J., Khanna, N. and Mandell, J.C., 2018. Integration of a zero-footprint cloud-based picture archiving and communication system with customizable forms for radiology research and education. Academic Radiology, 25(6), pp.811–818.</p><p>[4] Sedghi, A., Hamidi, S., Mehrtash, A., Ziegler, E., Tempany, C., Pieper, S., Kapur, T. and Mousavi, P., 2019, March. Tesseract-medical imaging: an open-source browser-based platform for artificial intelligence deployment in medical imaging. In Medical Imaging 2019: Image-Guided Procedures, Robotic Interventions, and Modeling (Vol. 10951, p. 109511R). International Society for Optics and Photonics.</p><p>[5] Calisto, F.M., Lencastre, H., Nunes, N.J. and Nascimento, J.C., Medical Imaging Diagnosis Assistant: AI-Assisted Radiomics Framework User Validation.</p><p>[6] Calisto, F.M., Lencastre, H., Nunes, N.J. and Nascimento, J.C., Medical Imaging Diagnosis Assistant: AI-Assisted Radiomics Framework User Validation.</p><p>[7] Calisto, F.M., Miraldo, P., Nunes, N. and Nascimento, J.C., BreastScreening: A Multimodality Diagnostic Assistant.</p><p>[8] Santiago, C., Nascimento, J.C. and Marques, J.S., 2013, July. Performance evaluation of point matching algorithms for left ventricle motion analysis in MRI. In 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 4398–4401). IEEE.</p><p>[9] Calisto, F.M., Lencastre, H., Nunes, N.J. and Nascimento, J.C., BreastScreening: Towards Breast Cancer Clinical Decision Support Systems.</p><p>[10] Cai, C.J., Reif, E., Hegde, N., Hipp, J., Kim, B., Smilkov, D., Wattenberg, M., Viegas, F., Corrado, G.S., Stumpe, M.C. and Terry, M., 2019, April. Human-centered tools for coping with imperfect algorithms during medical decision-making. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (p. 4). ACM.</p><p>[11] Lambin, P., Rios-Velazquez, E., Leijenaar, R., Carvalho, S., Van Stiphout, R.G., Granton, P., Zegers, C.M., Gillies, R., Boellard, R., Dekker, A. and Aerts, H.J., 2012. Radiomics: extracting more information from medical images using advanced feature analysis. European journal of cancer, 48(4), pp.441–446.</p><p>[12] Ting, D.S., Liu, Y., Burlina, P., Xu, X., Bressler, N.M. and Wong, T.Y., 2018. AI for medical imaging goes deep. Nature medicine, 24(5), p.539.</p><p>[13] Calisto, F.M., Ferreira, A., Nascimento, J.C. and Gonçalves, D., 2017, October. Towards Touch-Based Medical Image Diagnosis Annotation. In Proceedings of the 2017 ACM International Conference on Interactive Surfaces and Spaces (pp. 390–395). ACM.</p><p>[14] Liao, Z. and Carneiro, G., 2017. A deep convolutional neural network module that promotes competition of multiple-size filters. Pattern Recognition, 71, pp.94–105.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c9675819c33a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/oppr/using-cornerstonejs-and-orthanc-to-support-deep-learning-projects-c9675819c33a">Using CornerstoneJS and Orthanc to Support Deep Learning Projects</a> was originally published in <a href="https://medium.com/oppr">oppr</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[.NET Core — Using Entity Framework Core in a separate Project]]></title>
            <link>https://medium.com/oppr/net-core-using-entity-framework-core-in-a-separate-project-e8636f9dc9e5?source=rss----9de3d09b6cbe---4</link>
            <guid isPermaLink="false">https://medium.com/p/e8636f9dc9e5</guid>
            <category><![CDATA[entity-framework-core]]></category>
            <category><![CDATA[net-core-2]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[net-core]]></category>
            <dc:creator><![CDATA[Rodrigo Santos]]></dc:creator>
            <pubDate>Sat, 13 Apr 2019 13:12:23 GMT</pubDate>
            <atom:updated>2021-01-15T19:19:08.406Z</atom:updated>
            <content:encoded><![CDATA[<h3>.NET Core — Using Entity Framework Core in a separate Project</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/591/1*4HIU0YdDkj0dmGySVC5D_g.png" /></figure><blockquote>From the .NET documentation, <em>“Entity Framework Core is a cross-platform version of Entity Framework data access technology”</em>. It is used as an Object Relational Mapper (ORM), which makes the work of dealing with the database easier, since developers can interact with it using .NET objects (models), avoiding the usage of database specific code.</blockquote><p>Link to the same article on my blog <a href="https://www.rodrigosantosdev.com/Blog/net-core--using-entity-framework-core-in-a-separate-project">here</a>.</p><p>Most of the times, developers try to ensure that the Data Access Layer is entirely separated from the Business Logic of an application. These and others are considered best practices regarding software development, providing modularity and the principle of separation of concerns to an application.</p><p>In this post, we will build an example application that separates a .NET Core solution into three separates projects: a business logic project, a data access layer project and a web API project.</p><h3>Setting up the solution in Visual Studio</h3><p>Open up Visual Studio and create a solution that contains three projects with the following names:</p><ul><li><strong>MyCookingMaster.BL (Class Library)</strong> → corresponds to the business logic where we will place our Models.</li><li><strong>MyCookingMaster.DAL (Class Library)</strong> → corresponds to the data access layer. Here we are going to place everything regarding Entity Framework Core (DbContext and Migrations).</li><li><strong>MyCookingMaster.API (ASP.NET Core Web Application — API)</strong> → our startup project.</li></ul><p>Next, we’ll set the dependencies in each one of the projects. The DAL project has a dependency to the BL project in order to use the models that we will build. To do this, right click in the dependencies of the DAL project and select <em>Add Reference…</em> In the window that pops up, select the BL project and click <em>Ok</em>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h0EpwG-K9mLH3bcHnbHJaw.png" /></figure><p>In the API project, follow the same steps and add references to the BL and DAL project.</p><h3>Business Layer Project</h3><p>In this project we will build our models. It will be just a simple Recipes application model. Users can create recipes and see other recipes that were created by other users. Of course, for simplification reasons, it is not an exaustive model because it is not the main focus of the post.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/44a02f2c15a6c7a1b1149bd381dbbccd/href">https://medium.com/media/44a02f2c15a6c7a1b1149bd381dbbccd/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/ffd2baa4448c66de7196c23476bbac95/href">https://medium.com/media/ffd2baa4448c66de7196c23476bbac95/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c8ed6cd52f4ff55fee6537135f968920/href">https://medium.com/media/c8ed6cd52f4ff55fee6537135f968920/href</a></iframe><h3>Adding a database connection</h3><p>Access the appsettings.json file in the API project and provide a database connection.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/f8d1043aefca330e8122fdfe19445924/href">https://medium.com/media/f8d1043aefca330e8122fdfe19445924/href</a></iframe><h3>Data Access Layer Project</h3><p>Now that we have our Models in the BL Project and our database connection string in the API project, let’s go ahead and configure the DAL project.</p><p>We will need to have a reference to the following packages:</p><p>→ <strong>Microsoft.EntityFrameworkCore</strong></p><p>→ <strong>Microsoft.EntityFrameworkCore.Design</strong> (This package is required for the Entity Framework Core Tools to work)</p><p>→ <strong>Microsoft.EntityFrameworkCore.SqlServer</strong> (We are going to use a SQL Server database)</p><p>It is time to define our DbContext.</p><p>Inside the DAL project, create a class named <em>ApplicationDbContext</em> containing the following code.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/88e51e62d6e748a08c432a7f1f8333f5/href">https://medium.com/media/88e51e62d6e748a08c432a7f1f8333f5/href</a></iframe><p>Now, try to run one of the following commands inside the DAL project.</p><p><strong><em>.NET CLI</em></strong> → dotnet ef migrations add Initial</p><p><strong><em>NuGet Package Manager Console</em></strong> → Add-Migration Initial</p><blockquote>“Unable to create an object of type ‘ApplicationDbContext’. For the different patterns supported at design time, see https://go.microsoft.com/fwlink/?linkid=851728&quot;</blockquote><p><strong>Error</strong>? Erm… Hmm… What happened?</p><p>When we try to run the migrations, Entity Framework Tools tries to obtain the DbContext object from the application’s service provider. The problem is that the DAL project is a Class Library Project and <strong>NOT </strong>an ASP.NET Core Application/API. Next, it tries to create an instance using a constructor with no parameters from the derived DbContext (ApplicationDbContext in our case), which also does not apply.</p><p>There is other way we can tell Entity Framework Tools how to create our DbContext by implementing the <strong>IDesignTimeDbContextFactory&lt;TContext&gt;</strong> interface. From the .NET Core documentation:</p><blockquote>“If a class implementing this interface is found in either the same project as the derived DbContext or in the application’s startup project, the tools bypass the other ways of creating the DbContext and use the design-time factory instead.”</blockquote><p>Keeping all this in mind, let’s add the following code to <em>ApplicationDbContext.cs</em> file:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c744f0ba9e634bc7bb9406b6d64da169/href">https://medium.com/media/c744f0ba9e634bc7bb9406b6d64da169/href</a></iframe><blockquote>In order to resolve SetBasePath add the <strong>Microsoft.Extensions.Configuration.FileExtensions</strong> package. To resolve AddJsonFile, add a reference to the <strong>Microsoft.Extensions.Configuration.Json</strong> package.</blockquote><p>We are now ready to run an initial migration.</p><p><strong><em>.NET CLI</em></strong> → dotnet ef migrations add Initial</p><p><strong><em>NuGet Package Manager Console</em></strong> → Add-Migration Initial</p><p>And there you have it! An initial migration was created in the DAL project in the Migrations folder. Everytime we need to make a new migration, we just need to run the command inside the DAL project.</p><p>We can now create the database.</p><p><strong><em>.NET CLI</em></strong> → dotnet ef database update</p><p><strong><em>NuGet Package Manager Console</em></strong> → Update-Database</p><h3>Seeding the database (Optional)</h3><p>In order to test all our implementation, let’s also seed the database with some dummy data to be returned by the API. Create a file named <em>DataSeeder.cs</em> and add the following code.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9fd370470af7a9ef91760077e24dfb40/href">https://medium.com/media/9fd370470af7a9ef91760077e24dfb40/href</a></iframe><p>We need also to add some code in our API project, namely in <em>Program.cs </em>.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/ad13389f85819ba25ee5265649bccccc/href">https://medium.com/media/ad13389f85819ba25ee5265649bccccc/href</a></iframe><h3>API Project</h3><p>In order to our API project use the context we defined in the DAL project, it must be added through Dependency Injection in <em>Startup.cs</em> file. We also add some JSON Options just to avoid reference loops when returning JSON to the client.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/6a878ffb96f5abcd206167a78effbfc3/href">https://medium.com/media/6a878ffb96f5abcd206167a78effbfc3/href</a></iframe><p>Finally, we create a controller that can handle CRUD operations on the Clients.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c0d90f5c3022c9e30963b81933b9c53e/href">https://medium.com/media/c0d90f5c3022c9e30963b81933b9c53e/href</a></iframe><p>Run the API and access the ‘api/clients’ endpoint. If the database was seeded previously, it will show the data in JSON format.</p><h3>Conclusion</h3><p>And there you have it, our business logic and data access layer/entity framework are completely separated from our API. With this approach, all database logic is encapsulated into one project and can be shared across many projects in case we need.</p><p>Link to the same article on my blog <a href="https://www.rodrigosantosdev.com/Blog/net-core--using-entity-framework-core-in-a-separate-project">here</a>.</p><p>Source code: <a href="https://github.com/Rudyzio/EFCoreSeparateProject">GitHub</a></p><h3>Supporters</h3><p>Our organization is a non-profit organization. However, we have many expenses across our activity. From infrastructure to service expenses, we need some money, as well as help, to support our team and projects. For the expenses, we created several channels that will mitigate this problem. First of all, you can support us by being one of our <a href="https://patreon.com/oppr">Patreons</a>. Second, you can support us on <a href="https://opencollective.com/oppr">Open Collective</a> page. Thirdly, you can <a href="https://ko-fi.com/opprGroup">buy one coffee</a> (or more) for us. Fourth, you can also support us on our <a href="https://liberapay.com/oppr">Liberapay</a> page. Last but not least, you can directly support us on <a href="https://paypal.me/oppr">PayPal</a>. On the other hand, we also need help in the development of our projects. Therefore, if you have the knowledge we welcome you to support our projects. Just follow our channels and repositories.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e8636f9dc9e5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/oppr/net-core-using-entity-framework-core-in-a-separate-project-e8636f9dc9e5">.NET Core — Using Entity Framework Core in a separate Project</a> was originally published in <a href="https://medium.com/oppr">oppr</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Applying New Paradigms in Human-Computer Interaction to Health Informatics]]></title>
            <link>https://medium.com/oppr/applying-new-paradigms-in-human-computer-interaction-to-health-informatics-86d6a7046130?source=rss----9de3d09b6cbe---4</link>
            <guid isPermaLink="false">https://medium.com/p/86d6a7046130</guid>
            <category><![CDATA[user-centered-design]]></category>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[humancomputer-interaction]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[health-informatics]]></category>
            <dc:creator><![CDATA[Francisco Maria Calisto]]></dc:creator>
            <pubDate>Sun, 22 Apr 2018 18:36:00 GMT</pubDate>
            <atom:updated>2019-10-21T09:40:23.488Z</atom:updated>
            <content:encoded><![CDATA[<p>Nowadays, data acquisition and infrastructure for data processing and predictive capabilities are becoming cheaper and faster for <em>small</em> &amp; <em>low</em> resource organizations. As a result, researchers turn, at an ever-increasing rate, to <strong>Machine Learning (ML)</strong> and <strong>Data Science (DS)</strong> to implement better <strong>Human-Computer Interaction (HCI)</strong> in <strong>Health Informatics (HI)</strong>. In this blog post, we will give you an overview of how researchers are leveraging data, technologies and (<em>cheap</em> &amp; <em>fast</em>) infrastructure to improve the <strong>User eXperience (UX)</strong> and the decision making quality in <strong>HI</strong> and what new paradigms in <strong>HCI</strong> are setting the narrative for future research.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f1ViysfXNrVdvnAqqCEucw.jpeg" /><figcaption>Radiology Software</figcaption></figure><p>This article explores the new paradigms of <strong>HCI</strong> as a discipline on the development of <strong>User Interfaces (UIs)</strong> regarding <strong>HI</strong>, the current and future challenges for researchers, and the various trends, stating its innovations and developments. The implementation of those <strong>UIs</strong> establishes a direct interaction between clinical information availability and medical process accessibility. Thus, a need for workflow optimization changes the way in which we obtain the information, helping the clinical decision-making process. With the emergence of new information datasets and communication technologies, new opportunities have been created in the design and development of <strong>UIs</strong> by the application of <strong>HCI</strong> discipline. As an interdisciplinary field, the <strong>HCI</strong> theoretical roots enclose many several subjects that are also healthcare supportive. These include computing, ergonomics, cognitive sciences, and psychology. Applying <strong>HCI</strong> to the clinical domain allows less costly solutions but still presents theoretical and practical challenges. This opportunity has led to an impact in the society due to a change in the clinician’s user profile and the use of new interaction techniques. As suggested in the previous paragraphs, <strong>HI</strong> will also have an impact on <strong>HCI</strong> by opening new opportunities for applying <strong>HCI</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9lF-mF1SUYBq1R_GTAKrqA.jpeg" /><figcaption>Human-Computer Interaction of Health Informatics</figcaption></figure><p>Healthcare (<strong>HI</strong>) solutions are giving <strong>HCI </strong>enough space to extend its field of study. There are many ways of applying either standard and non-standard techniques to the existing <em>new paradigms</em>. For instance, one of my research papers [1] related to <a href="http://www.who.int/diagnostic_imaging/en/"><strong>Medical Imaging</strong></a> via touch-based surfaces, is highly significant and novel, but still operates within fewer novel techniques, the touch-based, and mobile-based device surfaces. The healthcare infrastructure is now ready to support systems by feeding them with their information. That said, we are starting to see <a href="https://mimbcd-ui.github.io/">orders</a> growing together and allowing <strong>Machine Learning (ML)</strong> techniques to be applied. These techniques will determine a trend in which the output of the system is a recommendation for the clinician. Through <strong>ML</strong> the machine can predict the most severe cases to be prioritized for or even give pathology predictions to clinicians. This behavior, therefore, raises specific <strong>UI</strong> issues. For example, in that research work [1], we have considered how to visualize <a href="https://www.dicomstandard.org/">DICOM</a> images of the patients. Also, this work, enables clinicians to act on such data by using touch devices. It requires a more profound knowledge regarding the <a href="https://www.interaction-design.org/literature/article/5-stages-in-the-design-thinking-process">Design Thinking</a> of the whole system. All of the medical information is being often imperceptibly gathered and integrated by different operations on the overall healthcare (<strong>HI</strong>) infrastructure.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DfPaZHQbbuD4vMW-5xzZWw.jpeg" /><figcaption><a href="https://dl.acm.org/citation.cfm?id=3134111">Touch-Based</a> User Interface (UI) for Medical Imaging [1]</figcaption></figure><p>Health systems <strong>UIs</strong> are a vital part of ubiquitous computing that takes an <strong>HCI</strong> advantage. Creating such UIs is difficult in the clinical domain because of the medical image processing [2] and dynamic healthcare knowledge required for creating classifiers. <strong>Interactive Machine Learning (iML)</strong> [3] can allow clinicians to view, classify and train machines by correcting the classifications.</p><blockquote>Algorithms that can interact with agents and can optimize their learning behaviour through these interactions, where the agents can also be humans.</blockquote><p>The quote above by Andreas Holzinger [4] defines <strong>iML</strong>. But first, we need to address the goal of <strong>Machine Learning (ML)</strong>: the development of algorithms improving the prediction ability over time through learning. Most of the researchers are putting their efforts by focusing on <strong>Automatic Machine Learning (aML)</strong> which brings us a significant advantage as an automatic approach benefits from big data with many training sets. This approach is also a disadvantage, since in the health domain; we are in the presence of a little number of events or available people to generate those datasets, where approaches like <strong>aML</strong> suffer from insufficient training samples. Here, the author brings a solution to this. The <strong>iML</strong> is, therefore, a viable approach for <strong>HI</strong>, having its roots in <strong>Active Learning (AL)</strong> [7] and <strong>Reinforcement Learning (RL)</strong> [8].</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pHqlxg4pWHv3UUaEcUzi8w.png" /><figcaption><a href="https://hci-kdd.org/project/iml/">iML with the Human-in-the-Loop</a> [9]</figcaption></figure><p>A more powerful learning process, one that can achieve higher accuracy with less information, can be obtained by combining <strong>AL</strong> with <strong>RL</strong>. A system with this approach may pose queries to be annotated by a radiologist. <strong>AL</strong> lends itself to our purpose as non-annotated medical images may not be abundant or difficult to obtain since radiologists annotations are expensive and time-consuming. The authors [4, 9] strongly recommend applying <strong>ML</strong> within <strong>HI</strong>. This procedure requires a concerted effort of researchers from different areas. Thereby, combining human cognition with <strong>iML</strong> approaches can be of particular interest to solve problems in <strong>HI</strong>, although there is a substantial lack of datasets.</p><p>By using a <strong>Human-In-The-Loop (HITL)</strong> approach, several difficulties could be mitigated while combining both human expertise with computer efficiency. The <a href="https://sloanreview.mit.edu/article/justifying-human-involvement-in-the-ai-decision-making-loop/">tests indicate that humans make significantly better decisions</a> [10, 11] when they lean on explanations that are extracted by these techniques. Therefore, the work done by these authors is essential to us. It is also an introduction to the <strong>HITL</strong> topic. For instance, clinicians are great at looking at complex medical images and selecting lesions. Therefore, a machine needs to understand what a lesion is. Different lesions and patients can give machine knowledge through the learning process. By teaching the machine to see those various lesions and patients (i.e. <em>machine intelligence</em>), it will improve machine algorithms in an incredibly quick and accurate way, thanks to a robust dataset (i.e. <em>human intelligence</em>) of the labeled (annotated) images.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/694/1*Cpkt6V6W6S3iDMtdQMI_TQ.png" /><figcaption><a href="https://www.crowdflower.com/machine-learning-with-crowdflower/">Human-in-the-loop for Machine Learning</a></figcaption></figure><p>The research domain is extraordinarily complex, while the focus is on <strong>HI</strong>. For that reason, it is essential to do qualitative work, not only quantitative, taking into account the clinical environment by studying the <em>clinical teams</em> and <em>patients</em>, observing them and, eventually, becoming one of them (or at least try). It is a <strong>User-Centered</strong> participatory process [5]. This approach, demands interviews and <strong>User Research (UR)</strong> by the use of artifacts [6], like video, designing workshops and focus groups. Therefore, it will complement the quantitative analysis, with answers regarding how is the <em>clinicians</em> work and <em>patients</em> care. Mixing both approaches, the qualitative and quantitative will feed our research with stronger data from our users (<em>clinicians</em> or <em>patients</em>) on a sensible domain field.</p><p>As far as we assess the clinical field of the domain, a multidisciplinary, hierarchical role-based coordination is acknowledged. We have surgeons, emergency physicians, nurses, technicians and so on. A lot of different specialists depend on the information. The hierarchy is fundamental for task delegation as other people are executing the tasks. There is even a distributed role for specific positions to allow division labor over the health infrastructure, bridging coordination work regarding the information. For instance, several clinical departments have a <em>doctor leader</em>, that coordinates care and assigns tasks using <strong>Information Systems (IS)</strong> to support communication. Typically, <em>nurses</em> are also setting up equipment and reporting states to an <strong>IS</strong>. Then a <em>scribe</em> will document the clinical events, recording the workflow. The health infrastructure transforms into a huge lab for <strong>HCI</strong> researchers to understand their users. That said, by observation, we have opportunities to bring here technological innovation and enough space for improvements. For improving their interactions. For enhancing <strong>HI</strong> supported by <strong>HCI</strong>.</p><h3>Acknowledgments</h3><p>This article is supported by the <a href="https://github.com/MIMBCD-UI/meta/wiki">case studies</a> of the <a href="https://mimbcd-ui.github.io/">MIMBCD-UI</a> and <a href="https://mida-project.github.io/">MIDA</a> projects. The two projects are strongly sponsored by <a href="https://www.fct.pt/">FCT</a>, a <a href="https://www.portugal.gov.pt/">Portuguese</a> public agency that promotes science, technology, and innovation, in all scientific domains. The genesis of this article was a research work between <a href="http://welcome.isr.tecnico.ulisboa.pt/">ISR-Lisboa</a> and <a href="http://www.inesc-id.pt/">INESC-ID</a>, both associated laboratories of <a href="http://tecnico.ulisboa.pt/">IST</a> at <a href="https://www.ulisboa.pt/">ULisboa</a>. Not forgetting the collaboration of <a href="https://www.m-iti.org/">M-ITI</a> that pairwise with <a href="http://welcome.isr.tecnico.ulisboa.pt/">ISR-Lisboa</a> are two research laboratories of <a href="http://larsys.pt/">LARSyS</a>. From these institutions, I would like to convey a special thanks to Professor Jacinto Nascimento and Professor Daniel Gonçalves for advisor me during my research work. Other special thanks to Professor Alfredo Ferreira, for co-authoring with several research papers. I also would like to thank Professor Joaquim Jorge who was one of my research career precursors. Last but not least, I would like to thank several important people of this noble organization called <a href="https://oppr.io/">oppr</a>. A special thanks to João Campos, Bruno Dias, and Rodrigo Lourenço for review this article giving me great inputs.</p><h3><strong>Supporters</strong></h3><p>Our organization is a non-profit organization. However, we have many expenses across our activity. From infrastructure to service expenses, we need some money, as well as help, to support our team and projects. For the expenses, we created several channels that will mitigate this problem. First of all, you can support us by being one of our <a href="https://patreon.com/oppr">Patreons</a>. Second, you can support us on <a href="https://opencollective.com/oppr">Open Collective</a> page. Thirdly, you can <a href="https://ko-fi.com/opprGroup">buy one coffee</a> (or more) for us. Fourth, you can also support us on our <a href="https://liberapay.com/oppr">Liberapay</a> page. Last but not least, you can directly support us on <a href="https://paypal.me/oppr">PayPal</a>. On the other hand, we also need help in the development of our projects. Therefore, if you have the knowledge we welcome you to support our projects. Just follow our channels and repositories.</p><h3>References</h3><p>[1] Calisto, Francisco M., et al. “Towards Touch-Based Medical Image Diagnosis Annotation.” <em>Proceedings of the 2017 ACM International Conference on Interactive Surfaces and Spaces</em>. ACM, 2017.</p><p>[2] McInerney, Tim, and Demetri Terzopoulos. “Deformable models in medical image analysis: a survey.” <em>Medical image analysis</em> 1.2 (1996): 91–108.</p><p>[3] Fails, Jerry Alan, and Dan R. Olsen Jr. “Interactive machine learning.” <em>Proceedings of the 8th international conference on Intelligent user interfaces</em>. ACM, 2003.</p><p>[4] Andreas Holzinger. Interactive machine learning for health informatics: when do we need the human-in-the-loop? Brain Informatics, 3(2):119–131, 2016.</p><p>[5] Brunner, Julian, et al. “User-centered design to improve clinical decision support in primary care.” <em>International journal of medical informatics</em> 104 (2017): 56–64.</p><p>[6] Gray, Kathleen, and Cecily Gilbert. “Digital Health Research Methods and Tools: Suggestions and Selected Resources for Researchers.” <em>Advances in Biomedical Informatics</em>. Springer, Cham, 2018. 5–34.</p><p>[7] Settles, Burr. “From theories to queries: Active learning in practice.” <em>Active Learning and Experimental Design workshop In conjunction with AISTATS 2010</em>. 2011.</p><p>[8] Busoniu, Lucian, Robert Babuska, and Bart De Schutter. “A comprehensive survey of multiagent reinforcement learning.” <em>IEEE Trans. Systems, Man, and Cybernetics, Part C</em> 38.2 (2008): 156–172.</p><p>[9] Holzinger, Andreas, et al. “Towards interactive Machine Learning (iML): applying ant colony algorithms to solve the traveling salesman problem with the human-in-the-loop approach.” <em>International Conference on Availability, Reliability, and Security</em>. Springer, Cham, 2016.</p><p>[10] Schirner, Gunar, et al. “The future of human-in-the-loop cyber-physical systems.” <em>Computer</em> 46.1 (2013): 36–45.</p><p>[11] Oliveira, Eugénio. “Beneficial AI: the next battlefield.” <em>Journal of Innovation Management</em> 5.4 (2018): 6–17.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=86d6a7046130" width="1" height="1" alt=""><hr><p><a href="https://medium.com/oppr/applying-new-paradigms-in-human-computer-interaction-to-health-informatics-86d6a7046130">Applying New Paradigms in Human-Computer Interaction to Health Informatics</a> was originally published in <a href="https://medium.com/oppr">oppr</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>