<?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[Gotta Know Healthtech - Medium]]></title>
        <description><![CDATA[Technology advancements in the health tech made simple - Medium]]></description>
        <link>https://medium.com/gotta-know-healthtech?source=rss----c5d76f3cd9e2---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Gotta Know Healthtech - Medium</title>
            <link>https://medium.com/gotta-know-healthtech?source=rss----c5d76f3cd9e2---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 17:32:43 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/gotta-know-healthtech" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[AI in Healthcare: Beyond the Illusion of a Magic Bullet!]]></title>
            <link>https://medium.com/gotta-know-healthtech/ai-in-healthcare-beyond-the-illusion-of-a-magic-bullet-4f0618100c66?source=rss----c5d76f3cd9e2---4</link>
            <guid isPermaLink="false">https://medium.com/p/4f0618100c66</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[healthcare]]></category>
            <dc:creator><![CDATA[Sanju Venugopal]]></dc:creator>
            <pubDate>Thu, 08 Jun 2023 07:14:42 GMT</pubDate>
            <atom:updated>2023-06-08T07:14:41.897Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/660/1*3bgDRRZGKoER5ijq8wIsyw.jpeg" /></figure><p><em>There is a perception that AI has the potential to solve all problems and revolutionize the healthcare industry. The integration of AI in healthcare has generated great enthusiasm for its potential to transform various avenues of healthcare like drug discovery, diagnostics, treatment, and healthcare delivery. AI’s advancements have indeed showcased impressive abilities in analyzing medical data, predicting outcomes, and aiding in intricate medical procedures.</em></p><p>However, it is important to critically examine its limitations and consider the limitations and challenges that persist.</p><h3>Some of the key limitations and challenges of AI in healthcare</h3><blockquote>1. Data Concerns</blockquote><blockquote>2. Need for More Human Involvement</blockquote><blockquote>3. Understandability Issue</blockquote><blockquote>4. Hurdles in AI Adoption and Implementation</blockquote><blockquote>5. Ethical Considerations</blockquote><h3><strong>1. Data Concerns</strong></h3><blockquote>Quality of Data</blockquote><blockquote>Difficulty in Data Collection, Access, and Aggregation</blockquote><blockquote>Fragmented Data</blockquote><blockquote>Limited Data</blockquote><blockquote>Regulations on Data Privacy and Protection</blockquote><p><strong><em>AI needs lots of good-quality data to learn from.</em></strong></p><ol><li>Healthcare data is often seen as incomplete, inconsistent, or contain errors.</li><li>This data is often difficult to collect, access, and aggregate.</li><li>The data is fragmented across different applications, institutions, and formats. <em>It’s challenging to pool such data across hospitals or across healthcare providers to aggregate and analyze data from multiple sources.</em></li><li>There isn’t enough information about people with rare diseases, specific health conditions, and health data of people of specific ethnicity and demographics.</li><li>The data is highly sensitive and subject to stringent privacy regulations.</li></ol><p><strong>AI Impact</strong></p><ol><li>Low-quality data (missing, inaccurate, inconsistent, incomplete) and limited availability of data undermine the performance and reliability of AI models and <strong>become labor-intensive to pre-process data for AI</strong> to get trained on.</li><li>Fragmentation <strong>hampers the ability to build comprehensive datasets</strong> necessary for training AI models.</li><li>It becomes challenging to train AI models specifically for these rare serious health conditions. This can <strong>result in biased predictions</strong> and worsen healthcare disparities.</li><li>Stringent regulations <strong>limit access to the necessary quality data</strong> for training AI models.</li></ol><h3>2. Need for More Human Involvement</h3><blockquote>Personalized care</blockquote><blockquote>Considering patient preferences</blockquote><blockquote>Detailed evaluation of health history</blockquote><blockquote>Diagnostic ambiguity</blockquote><blockquote>Legally and ethically sound decisions</blockquote><p><strong><em>A patient should be involved in decision-making about his or her own health</em></strong><em>.</em></p><ol><li>A Physician establishes a level of empathy, intuition, and personal communication. <em>They tailor the treatments to individual patients and consider factors like patient preferences, values, and individual circumstances. These are extremely valuable for patients.</em></li><li>The treatment decisions may involve assessing risks and benefits, considering potential interactions with other medications, and incorporating patient-specific factors.</li><li>There can be diagnostic ambiguity and uncertainty. <em>In these situations, employing clinical judgment, expertise, and critical thinking skills plays a very crucial role.</em></li><li>Healthcare decisions are subject to legal and ethical framework and medical professionals are made the final responsibility of these decisions.</li></ol><p><strong>AI Impact</strong></p><ol><li>Healthcare decisions often involve considerations that <strong>extend beyond purely data-driven approaches</strong> as seen in the AI applications.</li><li>AI may <strong>not fully capture the unique circumstances</strong> <strong>and preferences</strong> of every patient.</li><li>Data Analysis may require comprehensive evaluation and interpretation of various diagnostic tests and imaging studies, which <strong>require complex algorithms to be developed and trained on</strong>.</li><li>AI algorithms <strong>may not fully replicate the complex decision-making</strong> <strong>process</strong> where factors are prone to change over time, which is relative to individual patients.</li><li>While AI can provide decision support, the f<strong>inal responsibility lies with healthcare professionals</strong> to make legally and ethically sound decisions.</li></ol><h3>3. Understandability Issue</h3><blockquote>Interpretability</blockquote><blockquote>Transparency</blockquote><blockquote>Explainability</blockquote><blockquote>Trust and Confidence</blockquote><blockquote>Ethical and Legal Concerns</blockquote><p><strong><em>AI is assumed to play a key role in enhancing clinical decision-making</em></strong></p><ol><li>If the AI outputs are not clearly understandable, the workflow currently followed in healthcare takes time to integrate and adopt AI into their decision-making processes.</li><li>Being unable to establish AI transparency would make it difficult to interpret the AI’s system output and apply it appropriately.</li><li>Explainability involves gaining insights into how AI algorithms arrive at their predictions, recommendations, or decisions. <em>Some of the most well-known AI algorithms cannot explain themselves well.</em></li><li>To enable trust and confidence, AI should allow healthcare people to validate and inspect the AI results. <em>This ensures they align with their knowledge and expertise.</em></li><li>Healthcare professionals and patients have the right to understand the basis for AI-driven decisions that impact patient care. <em>If not understood, they can compromise ethical standards and make it difficult to meet legal requirements.</em></li></ol><p><strong>AI Impact</strong></p><p>1. AI integration into existing healthcare workflows <strong>will take more time</strong> than other sectors.</p><p>2. Without transparency, trust in using the AI systems appropriately cannot be built and may <strong>lead to skepticism and resistance among the healthcare community.</strong></p><p>3. Without explainability, the <strong>reliability of AI tends to decrease</strong>.</p><p>4. It may also become <strong>difficult to explain to the regulatory agencies</strong> if there are findings on its capabilities and results.</p><h3>4. Hurdles in Adoption and Implementation</h3><blockquote>Regulatory framework</blockquote><blockquote>Cost</blockquote><blockquote>Resistance</blockquote><blockquote>Interoperability</blockquote><blockquote>Training</blockquote><p><strong><em>Widespread AI adoption has the possibility to transform healthcare for good.</em></strong></p><ol><li>The regulatory agencies (for example FDA) are yet to make the regulatory approach effective for AI/ML-enabled devices <em>(except for “Guidance for Clinical Decision Support Systems” which is effective). The regulatory considerations pose a big challenge to the widespread adoption of AI.</em></li><li>The cost of developing, maintaining, and upgrading AI systems can be substantial (AI cloud solution providers are around though!). <em>Economic impact (should turn out positive) from AI implementation is seen as a key here.</em></li><li>Any introduction of new technology would face resistance to change, especially in healthcare where a considerable number of non-techies are involved. <em>Addressing concerns and demonstrating the value and benefits of AI should be carried out.</em></li><li>Aligning legacy systems, EHRs systems, and processes followed in healthcare to the AI can be complex where technical and compatibility issues may arise. <em>Only interoperability among systems and seamless data exchange can harness the full potential of AI while meeting HIPAA requirements.</em></li><li>Upskilling healthcare professionals and helping them understand the limitations <em>and have appropriate use of AI-driven insights go a long way in the adoption of AI.</em></li></ol><p><strong>AI Impact</strong></p><p>1.<strong> </strong>The non-availability of a regulatory framework, where <strong>developing AI solutions/capabilities first and then aligning to the regulations (once effective), involves time, cost, and effort</strong>.</p><p>2. Cost may be a factor where <strong>increased ROI and positive economic impact</strong> are a major focus for stakeholders.</p><p>3. A <strong>pushback may be expected</strong> on AI technology being relatively new.</p><p>4. Relatively <strong>seamless interoperability of AI</strong> with existing healthcare systems and their data <strong>may take time</strong> to achieve in the immediate future.</p><p>5. <strong>Education and training play key roles </strong>in the adoption of any new technology in healthcare.</p><h3>5. Ethical Considerations</h3><blockquote>Bias</blockquote><blockquote>Patients’ right to know</blockquote><blockquote>Unintended risks and consequences</blockquote><ol><li>As AI requires a lot of data to be trained on, if the data predominantly represents only a particular demographic group, it may lead to inaccurate results. <em>These results may affect decisions made, for example, treatment suggestions.</em></li><li>The patient has to consent to use their data and must be empowered on the key decisions made from data and AI-driven treatments.</li><li>Usage of new technologies like AI may be prone to have unintended risks and consequences, which were never documented before.<em> So understanding the technology and having mitigation strategies in place would help to prevent harmful outcomes especially related to patients’ health.</em></li></ol><p><strong>AI Impact</strong></p><p>1. Availability of <strong>diverse data is a mandate to train AI</strong> for healthcare. This may be difficult to get or have access to.</p><p>2. Making patients aware of their <strong>data being used for training AI may have a pushback</strong> since there is always a risk to data privacy and data protection.</p><p>3. It’s <strong>difficult to foresee the risks and unintended consequences of AI</strong> usage before its widespread usage, which may raise some eyebrows.</p><p><em>Healthcare is a deeply personal and emotionally complex field that requires the empathy, compassion, and expertise of healthcare professionals. Nevertheless, it is crucial to approach AI in healthcare with a critical mindset, acknowledging its limitations and understanding the complexities and challenges that still exist. By balancing enthusiasm with careful consideration, we can harness the true potential of AI in healthcare.</em></p><h3>References</h3><ol><li><a href="https://healthtechmagazine.net/article/2020/03/artificial-intelligence-healthcare-why-transparency-matters-perfcon"><em>https://healthtechmagazine.net/article/2020/03/artificial-intelligence-healthcare-why-transparency-matters-perfcon</em></a></li><li><a href="https://www.frontiersin.org/articles/10.3389/frai.2022.879603/full"><em>https://www.frontiersin.org/articles/10.3389/frai.2022.879603/full</em></a></li><li><a href="https://www.holisticai.com/blog/ai-transparency"><em>https://www.holisticai.com/blog/ai-transparency</em></a></li><li><a href="https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-021-04368-1"><em>https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-021-04368-1</em></a></li><li><a href="https://www.fda.gov/media/109618/download"><em>https://www.fda.gov/media/109618/download</em></a></li><li><a href="https://healthitanalytics.com/news/fda-releases-ai-ml-enabled-medical-device-modification-guidance"><em>https://healthitanalytics.com/news/fda-releases-ai-ml-enabled-medical-device-modification-guidance</em></a></li><li><a href="https://www.fda.gov/regulatory-information/search-fda-guidance-documents/marketing-submission-recommendations-predetermined-change-control-plan-artificial"><em>https://www.fda.gov/regulatory-information/search-fda-guidance-documents/marketing-submission-recommendations-predetermined-change-control-plan-artificial</em></a></li><li><a href="https://www.brookings.edu/research/why-is-ai-adoption-in-health-care-lagging/"><em>https://www.brookings.edu/research/why-is-ai-adoption-in-health-care-lagging/</em></a></li><li><a href="https://www.quantib.com/blog/how-healthcare-ai-can-improve-patient-centeredness"><em>https://www.quantib.com/blog/how-healthcare-ai-can-improve-patient-centeredness</em></a></li><li><a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8521931/"><em>https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8521931/</em></a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4f0618100c66" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gotta-know-healthtech/ai-in-healthcare-beyond-the-illusion-of-a-magic-bullet-4f0618100c66">AI in Healthcare: Beyond the Illusion of a Magic Bullet!</a> was originally published in <a href="https://medium.com/gotta-know-healthtech">Gotta Know Healthtech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is AI? What has healthcare to do with it?]]></title>
            <link>https://medium.com/gotta-know-healthtech/what-is-ai-what-has-healthcare-to-do-with-it-2f88b642a7e?source=rss----c5d76f3cd9e2---4</link>
            <guid isPermaLink="false">https://medium.com/p/2f88b642a7e</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[healthcare-technology]]></category>
            <category><![CDATA[healthtech]]></category>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Sanju Venugopal]]></dc:creator>
            <pubDate>Wed, 31 May 2023 07:25:28 GMT</pubDate>
            <atom:updated>2023-05-31T07:25:28.217Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/893/1*b7x3jN591A_6HCjcYUC8RQ.jpeg" /></figure><h3>Contents</h3><blockquote>1. AI</blockquote><blockquote>2. What AI has to do with healthcare?</blockquote><blockquote>3. Things to take care of to effectively use AI in healthcare</blockquote><h3>1. AI</h3><p>Artificial Intelligence (AI) is like having a smart robot friend who learns and understands by looking at lots of information and finding patterns in them.</p><p>Let’s assume you have a smart robot friend. For instance, if you train your friend about pictures of dogs, the friend will learn what dogs look like. Then, if you show some random pictures (which includes pictures of dogs!) it can point you to the pictures of a dog with fair accuracy based on the training it was given.</p><p><em>This tells us that AI should always learn with any new information given. So we need to train them correctly and with the right information and use them carefully.</em></p><p><strong>US FDA states that</strong></p><blockquote>Artificial Intelligence has been broadly defined as the science and engineering of making intelligent machines, especially intelligent computer programs (McCarthy, 2007).</blockquote><blockquote>Artificial intelligence can use different techniques, including models based on statistical analysis of data, expert systems that primarily rely on if-then statements, and machine learning.</blockquote><p><strong><em>Looks cool! But does AI have applications in healthcare?</em></strong></p><p><strong><em>Let’s see.</em></strong></p><h3>2. What AI has to do with healthcare?</h3><p><strong>Here are some great examples of what AI can do</strong></p><blockquote>Medical Imaging</blockquote><blockquote>Personalized Care</blockquote><blockquote>Discovering Medicines</blockquote><blockquote>Remote Check-ups</blockquote><blockquote>Keeping Things Organized</blockquote><ol><li><strong>Medical Imaging </strong><br>Sometimes, doctors need to scan our bodies using X-rays, MRIs, and CT scans to diagnose. AI can look at these scans and help the doctor find any problems or things that need attention. It’s like having a second pair of eyes that can spot things the doctor might have missed!</li><li><strong>Personalized Care </strong><br>Everyone is different, and our bodies work in unique ways. AI can learn about each person and their specific needs by looking at their medical records, genetic information, and other important details. Then, it can help the doctor give personalized care and treatment to each person, making sure they get exactly what they need to feel better.</li><li><strong>Discovering Medicines</strong><br>Scientists are always trying to find new medicines to help people get better. AI can help them discover new medicines faster by doing a lot of calculations and tests. It’s like having a smart helper who can do lots of experiments in no time!</li><li><strong>Remote Check-ups</strong><br>Sometimes, people cannot visit the doctor in person, especially when they live far away. AI can help doctors check on these people from a distance using special devices. It’s like having a virtual doctor who can talk to us and make sure we’re doing well, even from far away!</li><li><strong>Keeping Things Organized</strong><br>Doctors have lots of important things to remember, like appointments, medical records, and insurance information. AI can help them remember and organize all these things. It’s like having a smart assistant who keeps everything in order so the doctor can focus on helping people.</li></ol><p><em>AI would be a smart assistant for healthcare professionals and scientists, making things easier and helping them help us feel better. There are many more areas that AI can help healthcare professionals.</em></p><p><strong><em>Yeah Absolutely.</em></strong></p><p><strong><em>But wait! What is required to use AI in healthcare which is a highly regulated industry?</em></strong></p><p><strong><em>Let’s see.</em></strong></p><h3>3. Things to take care of to effectively use AI in healthcare</h3><p><strong>Here are things</strong></p><blockquote>Lots of Good Information</blockquote><blockquote>Keeping Information Safe</blockquote><blockquote>Important Checks to be done</blockquote><blockquote>Being Fair and Honest</blockquote><blockquote>Working Together</blockquote><blockquote>Learning and Getting Better</blockquote><ol><li><strong>Lots of Good Information (Access to high-quality and diverse data)</strong><br>AI needs lots of good, diverse, and helpful information to learn from. This information includes things like people’s medical records, imaging data, genetic information, and other important details about their health. <strong><em>This is crucial for training AI models and ensuring their reliability.</em></strong></li><li><strong>Keeping Information Safe (Data Privacy and Security)</strong><br>Healthcare data is sensitive, and maintaining data privacy and data security is of utmost importance. The information used by AI is very private and needs to be kept safe. We need to make sure that nobody else can access or use it without permission. <strong><em>This can be achieved by implementing secure data storage, encryption techniques, and strict access controls to prevent unauthorized use or disclosure of sensitive data</em></strong><em>.</em></li><li><strong>Important Checks to be done (Validation and following Regulations)</strong><br>AI systems should go through tests (undergo rigorous validation and testing processes) to make sure they are safe, effective, and reliable for helping people. <strong><em>Regulatory bodies like Food and Drug Administration (FDA), and European Medical Agency (EMA) play a crucial role in evaluating and approving AI technologies for use in healthcare to protect patients and ensure adherence to established standards.</em></strong></li><li><strong>Being Fair and Honest (No Bias)</strong><br>AI needs to be fair and honest, just like we are. It should treat everyone the same way and not show any preferences. It should also be clear and explain how it makes decisions so that everyone understands. <strong><em>This ensures fairness and avoids bias in algorithm design, and considers the ethical implications of AI such as consent, accountability, and potential social impact.</em></strong></li><li><strong>Working Together (Collaboration and Integration</strong>)<br>AI in health tech needs different people, like doctors, scientists, and technology experts, to work together. They share their knowledge and skills to make AI better and help more people. <strong><em>Collaboration helps ensure that AI solutions are designed to address real healthcare needs and can be seamlessly integrated into existing healthcare workflows.</em></strong></li><li><strong>Learning and Getting Better (Continuous Learning and Improvement)</strong><br>AI needs to keep learning and improving, just like we do. It can learn from new information and experiences to become smarter and help people in better ways. <strong><em>This requires mechanisms for ongoing monitoring, evaluation, and improvement of AI algorithms and models which refine and enhance the performance of AI systems over time.</em></strong></li></ol><p><em>When all these things come together, AI can be really helpful in health tech. It can help healthcare professionals and scientists find problems in our bodies, give personalized care, discover new medicines, and keep important information organized. AI is like having a smart friend who helps healthcare professionals to take care of us and keep us healthy.</em></p><h3>References</h3><ol><li><em>FDA — </em><a href="https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device#whatis"><em>https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device#whatis</em></a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2f88b642a7e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gotta-know-healthtech/what-is-ai-what-has-healthcare-to-do-with-it-2f88b642a7e">What is AI? What has healthcare to do with it?</a> was originally published in <a href="https://medium.com/gotta-know-healthtech">Gotta Know Healthtech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>