The Role of AI in Physician Decision Support

Written by Darwish Thajudeen  

The Global Diagnostic Gap

About 5% of US adult patients experience diagnostic errors, or approximately 12 million US adults every year (1). For comparison, the country of Belgium has a population of 11.8 Million (2). So, how would we feel if the entire population of Belgium experienced diagnostic errors every year? Would you be willing to investigate the possibilities of AI helping in this regard?

Every single medical data point could share an insight, but in many parts of the world, there aren’t enough doctors or diagnostic tools to analyze it carefully. Across Africa, there is a shortage of healthcare workers (3) and many healthcare professionals are making life-saving decisions with limited diagnostic support (4). In settings where specialist access is scarce and infrastructure is stretched thin, physicians often rely on clinical instinct alone to diagnose serious conditions, which at times can be unknown to them. At MI4People, we believe AI can be a game-changer, by empowering them with the right set of AI assistance. And that is how we could achieve our goals of developing Artificial Intelligence for good and Machine Intelligence for people (MI4People).  

From Static Rules to Adaptive Intelligence  

Traditional decision support systems were built on static rules—but AI brings adaptive intelligence into the clinic. Through machine learning, AI systems trained on thousands of health records and data can now detect subtle abnormalities that might be missed by even the most experienced physicians. These models could become capable of identifying complex patterns and correlations, providing predictive insights and early warnings for complicated conditions and rare diseases.

Physicians today—especially in low-resource settings—are often overwhelmed and under-supported. Time pressure, patient load, and diagnostic uncertainty are part of their daily reality. But society, authorities and physicians must research and help in research to understand that AI could be a great aid in this direction, because AI could act as a trusted second opinion accessible anytime, if trained with the most accurate data, tested with the help of field experts and approved by policymakers. This would be especially powerful for early-career doctors or rural practitioners who may not have a specialist nearby.  

Building AI That Serves Doctors and Saves Lives   

Let us share about two such great tools aiding the decision support system, that our kindhearted volunteers at MI4People help build.

Project 1- Sustain AI ECG

At MI4People, we’re working to bring life-saving diagnostic tools to doctors in underserved regions. One of our flagship initiatives focuses on deploying AI-powered ECG analysis in Ethiopia, where many hospitals lack access to cardiologists and advanced diagnostic systems.

Thanks to the Technische Hochschule Mittelhessen (THM), who generously made their AI-powered ECG tool available to MI4People for free use in Africa, and with financial support from the Bavarian State Chancellery, we are now able to roll out a life-saving, cloud-based solution built on AWS infrastructure in 10 clinics. This allows physicians to upload ECG data using low-cost devices — and receive AI-supported diagnostic assistance in real time.

The pilot phase is already underway in Ethiopian hospitals, aiming to help general practitioners detect cardiac risks earlier and with greater confidence. Our long-term goal is to scale this model across the continent, enabling healthcare providers everywhere to access high-quality, affordable diagnostic support. This project reflects our core mission: to amplify the skills of physicians through AI and ensure that the benefits of technology reach those who need it most.

Project 2- AI for pediatric rare diseases 

Children with rare diseases face the most difficult journey. Because many in their early stages are widely misdiagnosed or receive delayed diagnoses after being passed from doctor to doctor by when the disease has grown to different stages. With over 95% of rare diseases lacking approved treatments, timely and accurate diagnosis becomes even more critical. Meanwhile, modern medicine is generating vast amounts of complex medical data that could hold the key to better care—but it’s simply too much for any one physician to analyze alone.

This is where AI and machine learning can make a transformative difference. These technologies can sift through massive datasets, identify hidden patterns, and support doctors in making more accurate and personalized medical decisions—especially for complex and rare conditions in children. For adults, AI can also detect early warning signs of neurological or cardiovascular issues, allowing for better prevention and targeted treatment.

In partnership with the Care-for-Rare Foundation and Dr. von Haunerschen Children’s Hospital in Munich, MI4People is developing a secure, decentralized AI platform to support the diagnosis and treatment of children with rare diseases. This platform will empower medical professionals with AI tools while ensuring that sensitive patient data remains fully protected within each clinic.

The impact could be life-changing: faster diagnosis, fewer unnecessary tests and treatments, reduced exposure to harmful medications, and ultimately, saving lives. And together, we can ensure that no child with a rare disease is left behind.

AI empowers, doctors decide

To summarize our thoughts, we always believe AI would never replace a doctor, but it can make sure they have more confidence in their decisions, backed by millions of data and support of evidence-based information when diagnosing a life-threatening condition. And to make that happen, each and every one of us has a larger role to play, such as advocating for AI for good movements, spreading awareness of AI. And if you can take a step further, help in democratizing data, supporting research, encouraging policymakers, and aiding AI initiatives. Thankfully we are honored to have you and a huge community that supports our initiatives. We are grateful for partners, donors, advisors, volunteers and global followers - thank you. Your belief in technology for good is helping us make a real difference in real lives. We believe, someday one of our tools that we could build with your unwavering support and dedication, may be able to save the life of a child to grant him a bright future filled with joy rather than pain and suffering.

The journey from data to diagnosis is not just about machines; it’s about enhancing care, and saving lives in places where it matters most. And we’re just getting started. If you're passionate about AI for good, let’s connect, and help save lives.

With heartfelt gratitude, 

MI4People Team