Latest news with #MAI-DxO
Mint
7 days ago
- Health
- Mint
Mint Explainer: Can AI diagnose you better than a doctor?
Few sectors are witnessing AI's disruptive power as dramatically as healthcare. Trained on millions of historical patient records, advanced AI models are now matching—and in some cases surpassing—human expertise. With real-time analysis across medical histories, imaging, and genetic profiles, AI transforms clinical complexity into diagnostic clarity. AI spots patterns even seasoned physicians might overlook. Earlier this month, Microsoft unveiled a platform–Microsoft AI Diagnostic Orchestrator (MAI-DxO)—that can diagnose medical cases with an accuracy rate of up to 85.5%, far exceeding the 20% accuracy of experienced physicians under the same conditions. Mint explains the role and impact of AI in healthcare and diagnostics. Does Microsoft's MAI-DxO improve diagnostic accuracy? MAI-DxO leverages vast datasets, advanced probabilistic reasoning, and continuous learning from clinical cases to surpass human diagnostic accuracy. Unlike doctors who rely on experience and pattern recognition, MAI-DxO evaluates multiple variables simultaneously—lab results, symptoms, imaging—and predicts outcomes. MAI‑DxO was tested on 304 real-world clinical scenarios sourced from the New England Journal of Medicine. In these cases, MAI‑DxO achieved an accuracy rate of 85.5%, surpassing the average 20% success rate of 21 experienced physicians from the UK and the US. Instead of relying on a single model, MAI-DxO coordinates multiple large language models (LLMs) that interact like a team of doctors, reviewing, challenging, and refining each other's suggestions before settling on a final diagnosis. Microsoft has emphasised that the tool is not designed to replace doctors, but to work alongside them. What medical datasets does MAI-DxO use? MAI-DxO is trained on anonymised patient records, peer-reviewed literature, medical guidelines and global clinical datasets from hospitals and research institutions. Its database spans symptoms, biomarkers, comorbidities, and disease trajectories across demographics, allowing it to contextualize findings in real-world scenarios. So if a patient suddenly feels numbness on her left side, it could indicate stroke, nerve damage, heart problem or some other issue. Doctors will wait for test results (blood and scans) to diagnose the problem, while the AI platform will interpret it faster, helping the doctor make quicker decisions. MAI-DXO's strength lies in its ability to absorb medical nuance across specialties—mirroring the cumulative expertise of several doctors and researchers. Will AI replace doctors? AI will not replace but complement doctors and other health professionals. They need to navigate ambiguity and build trust with patients. Clinical roles will evolve with AI, giving medical staff the ability to automate routine tasks, identify diseases earlier, personalize treatment plans, and potentially prevent some diseases altogether. For consumers, they will provide better tools for self-management and shared decision-making. These will be particularly helpful in remote areas where there's a shortage of doctors. MAI-DxO operates on statistical inference rather than intuition. It learns to recognise patterns and contextual subtleties from data, simulating judgment. While human intuition draws on tacit knowledge—emotion, empathy, gut feel—AI taps structured data and outcome probabilities. Platforms like MAI-DxO can augment decision-making. These are especially useful when doctors second-guess themselves—making MAI-DxO an ally, not a rival. What risks do AI-driven diagnostics pose for patient care? Key risks include overreliance, bias from skewed training data, and lack of transparency in AI reasoning. If doctors defer judgment too easily, subtle clinical clues could be overlooked. Additionally, AI platforms recommendations may vary based on demographics or comorbidities not well represented in its training data, potentially exacerbating health disparities. For instance, if they are trained only on American or European patient data, they might overlook or misinterpret common conditions among patients in South Asia, say tuberculosis (TB) or type 2 diabetes or complications caused by chewing tobacco. There's also the risk of false positives and patient anxiety. Most critically, AI tools must be accountable — patients need clarity on how diagnoses are made. The challenge lies in integrating AI without ignoring the human touch in care. What are the regulatory frameworks for AI deployment in hospitals? In India, every AI tool must be approved by the Delhi-based Central Drugs Standard Control Organisation (CDSCO)— equivalent to the US Food and Drug Administration (FDA). Globally, regulatory bodies like the FDA (US), European Medicines Agency (EMA-Europe) are developing guidelines for AI in medicine. Back home, AI platforms need clearance as a software as a medical device (SaMD), requiring evidence of clinical safety, efficacy, and data privacy compliance. What is the role of AI in the future of medicine? Current developments position them as early warning systems, helping doctors rather than making final decisions. According to the World Economic Forum 4.5 billion people are currently without access to essential healthcare services and a health worker shortage of 11 million is expected by 2030. AI has the potential to help bridge that gap. AI can also assist paramedics in situations where a patient is being transferred via an ambulance. AI models trained on factors such as a patient's mobility, pulse and blood oxygen levels, chest pain, etc., can relay information to doctors, helping them make decisions faster. AI can detect early signs in an individual that are likely to result in diseases like Alzheimer's, heart disease, kidney disease, and so on. This could help doctors suggest preventive action. Can MAI-DxO evolve into a real-time medical assistant across specialties? At present, MAI-DxO is a demonstration of AI capability and research. With real-time access to patient records, labs, and imaging, MAI-DxO could become an assistant, offering diagnostic suggestions, tracking progress, and alerting doctors to anomalies as they emerge. Integrated into electronic health systems, it could support rounds, triage, and even remote consultations. Much like AI embedded medical devices (ultrasound devices, bedside X-ray machines) are now being used in some hospitals, including Max healthcare, to interpret scans and assist doctors. The key lies in continual updates, feedback loops and building trust.
International Business Times
23-07-2025
- Business
- International Business Times
Microsoft Recruits 24 AI Experts from Google DeepMind, Including Gemini Chatbot Veteran
Microsoft has hired 24 top engineers, researchers, and product specialists from Google DeepMind, according to a report by the Financial Times . The most high-profile recruit is Amar Subramanya, who previously led engineering for Google's Gemini chatbot. He has now joined Microsoft as Corporate Vice President of AI. X Subramanya spent 16 years at Google. In a LinkedIn post, he praised Microsoft's work culture, calling it "low ego" and "ambitious." He also said the environment reminded him of a startup—fast-moving, collaborative, and focused on building cutting-edge AI products. Subramanya is now working on Microsoft's consumer-facing tools, including Bing and Copilot. This hiring spree is part of Microsoft's broader AI strategy led by Mustafa Suleyman, co-founder of DeepMind and now Microsoft's head of consumer AI. Suleyman joined Microsoft in 2024 following the company's $650 million "acqui-hire" of his startup, Inflection AI. Since then, he has built a strong team by leveraging his DeepMind connections. Other DeepMind alumni who have joined Microsoft include engineering lead Sonal Gupta, senior engineer Adam Sadovsky, and product manager Tim Frank. Sadovsky, who spent nearly 18 years at Google, now serves as a corporate VP at Microsoft AI. The tech industry is currently in a fierce race to secure AI talent. OpenAI CEO Sam Altman recently claimed Meta offered some of his staff $100 million signing bonuses. Meta also reportedly recruited Scale AI founder Alexandr Wang, though Meta denies some of the high figures being reported. Microsoft and Google continue to compete closely in the AI space. Suleyman now competes directly with DeepMind CEO Demis Hassabis, his former colleague. However, sources close to Google say DeepMind's attrition remains below the industry average. Meanwhile, under Suleyman's leadership, Microsoft recently unveiled an AI diagnostic system claimed to be four times more accurate than human doctors. Called MAI-DxO, the tool combines multiple AI models to assess symptoms and deliver diagnoses. Despite these advancements, Microsoft recently laid off around 9,000 employees globally, raising questions about balancing large-scale AI investment with broader cost-cutting.
7NEWS
17-07-2025
- Health
- 7NEWS
Microsoft claims its AI medical tool can outperform human doctors in diagnosing complex cases
In June, Microsoft unveiled a groundbreaking AI-powered diagnostic tool that could revolutionise how doctors detect and treat complex diseases. Called the Microsoft AI Diagnostic Orchestrator (MAI-DxO), the system mimics a panel of five virtual doctors, each with its own specialty, working together to debate and diagnose. In trials, it solved nearly 86 per cent of difficult medical cases published in The New England Journal of Medicine, which is known for complexity. That is over four times more accurate than experienced human doctors under the same conditions, according to the tech's co-founder Dr Mustafa Suleyman. 'We're talking about models that are not just better, but dramatically better — faster, cheaper, and more accurate than human doctors,' Suleyman said. Experts urge caution — it's early days. The system made the diagnoses using patient data already provided, and clinical use is still a way off, with peer-review and safety testing ahead. University of Sydney biomedical informatics professor Adam Dunn said it is not a real representation of what doctors do. 'These tools can be useful as an assistant, but they're never going to replace doctors,' Dunn told 7NEWS. 'The doctor has to quickly look at you and understand what kinds of questions to ask you to gather the right kinds of information. 'The second important thing to think about for ethics is related to who the AI is intended to support. 'In a lot of cases if the data are from one hospital in Boston, it's not a representative population. 'The decisions that an AI can make may be unfair. They may not be protecting our most marginalised and vulnerable populations.' Artificial intelligence is currently being used across healthcare in many areas, with one standing out in terms of popularity. Known as AI Scribes, the system listens to the conversation between a patient and a doctor or healthcare provider, and then quickly summarises the chat for medical records. Australian Medical Association chair of public health Dr Michael Bonning said it allows the clinician to spend more time with the patient and less on paperwork. 'We know, and many practitioners will say this, that because of the time pressures on their day, they will often write a short series of notes and then rely on their memory of the consultation as well to provide further insight down the track,' Bonning told 7NEWS. 'In a six or eight-hour day, you might spend an hour and a half of that typing. And if in that circumstance you can let the AI do that, that's a huge time saving for a clinician.' Doctors still review the transcription. 'Cannot replace humans' Members of the public told 7NEWS they welcome the use of artificial intelligence as a tool to assist doctors, but would not trust the computer in place of a human doctor. 'I think it can be used as a support b, butot completely replacing humans,' one person said. 'I think we still have to have human reviewing in there.' Another said a 'human doctor is still better, but I guess everything is artificial intelligence'. 'It's scary, but at the same time you don't know what the future holds with AI. It's both good and bad.'
Time of India
02-07-2025
- Business
- Time of India
Microsoft's new AI tool a medical genius? Tech giant claims it is 4x more accurate than real doctors
Tech giant Microsoft, recently hit with a fresh round of layoffs, has developed a new medical AI tool that performs better than human doctors at complex health diagnoses, creating a 'path to medical superintelligence'. The Microsoft AI team shared research that demonstrated how AI can sequentially investigate and solve medicine's most complex diagnostic challenges—cases that expert physicians struggle to answer. Tech company's AI unit, led by the British tech pioneer Mustafa Suleyman , has developed a system that imitates a panel of expert physicians tackling 'diagnostically complex and intellectually demanding' cases. Microsoft AI Diagnostic Orchestrator (MAI-DxO) correctly diagnosed up to 85% of NEJM case proceedings, a rate more than four times higher than a group of experienced physicians. MAI-DxO also gets to the correct diagnosis more cost-effectively than physicians, the company said in a blog post. ALSO READ: Microsoft layoffs: Tech giant's sales head Judson Althoff asked to go on two-month leave. Here's why Microsoft says AI system better than doctors The Microsoft AI Diagnostic Orchestrator', or MAI-DxO for short, the AI-powered tool is developed by the company's AI health unit, which was founded last year by Mustafa Suleyman. The tech giant said when paired with OpenAI's advanced o3 AI model, its approach 'solved' more than eight of 10 case studies specially chosen for the diagnostic challenge. When those case studies were tried on practising physicians – who had no access to colleagues, textbooks or chatbots – the accuracy rate was two out of 10. Microsoft said it was also a cheaper option than using human doctors because it was more efficient at ordering tests. When benchmarked against real-world case records, the new medical AI tool 'correctly diagnoses up to 85% of NEJM case proceedings, a rate more than four times higher than a group of experienced physicians' while being more impressive is that these cases are from the New England Journal of Medicine and are very complex and require multiple specialists and tests before doctors can reach any conclusion. Live Events According to The Wired, the Microsoft team used 304 case studies sourced from the New England Journal of Medicine to devise a test called the Sequential Diagnosis Benchmark. A language model broke down each case into a step-by-step process that a doctor would perform in order to reach a diagnosis. ALSO READ: Melania should be on first boat: Deportation calls for US' First Lady gains traction amid Trump's immigration crackdown Microsoft new AI tool diagnosed 85% cases For this, the company used different large language models from OpenAI, Meta, Anthropic, Google, xAI and DeepSeek. Microsoft said that the new AI medical tool correctly diagnosed 85.5 per cent of cases, which is way better compared to experienced human doctors, who were able to correctly diagnose only 20 per cent of the cases. "This orchestration mechanism—multiple agents that work together in this chain-of-debate style—that's what's going to drive us closer to medical superintelligence,' Suleyman told The Wired. Microsoft announced it is building a system designed to mimic the step-by-step approach of real-world clinicians—asking targeted questions, ordering diagnostic tests, and narrowing down possibilities to reach an accurate diagnosis. For example, a patient presenting with a cough and fever might be guided through blood tests and a chest X-ray before the system determines a diagnosis like pneumonia. ALSO READ: Sean Diddy Combs' secret plan against his ex Jennifer Lopez emerges amid sex-trafficking trial Microsoft said its approach was able to wield a 'breadth and depth of expertise' that went beyond individual physicians because it could span multiple medical disciplines. It added: 'Scaling this level of reasoning – and beyond – has the potential to reshape healthcare. AI could empower patients to self-manage routine aspects of care and equip clinicians with advanced decision support for complex cases.' Microsoft acknowledged its work is not ready for clinical use. Further testing is needed on its 'orchestrator' to assess its performance on more common symptoms, for instance. Economic Times WhatsApp channel )
Miami Herald
02-07-2025
- Health
- Miami Herald
Microsoft is working on a surprising way to help you live longer
Several years ago, I developed a strange disease. Too much sitting and too much stress from my IT job took its toll. Describing the symptoms is very difficult. It feels like something is moving in my calves. It is not painful, but I'd rather be in pain than feel that strange sensation. The doctors were clueless. They did all the tests. Ultrasound. Electromyoneurography. MRI of the lumbar part of the spine. The radiologist was having so much fun with me that he suggested I should also do an MRI of my brain. Related: OpenAI makes shocking move amid fierce competition, Microsoft problems I was looking for different opinions, and I never got a diagnosis. Not that specialists didn't have "great" ideas for experiments on me. That is what happens when you don't have a run-of-the-mill disease. Surprisingly, Microsoft, which isn't exactly known for being a medical company, may have a solution to finding the proper diagnosis, especially for difficult cases. Dominic King and Harsha Nori, members of the Microsoft (MSFT) Artificial Intelligence team, blogged on June 30th about their team's activities. According to them, generative AI has advanced to the point of scoring near-perfect scores on the United States Medical Licensing Examination and similar exams. But this test favors memorization over deep understanding, which isn't difficult for AI to do. The team is aware of this test's inadequacy and is working on improving the clinical reasoning of AI models, focusing on sequential diagnosis capabilities. This is the usual process you go through with the doctor: questions, tests, more questions, or tests until the diagnosis is found. Related: Analyst sends Alphabet warning amid search market shakeup They developed a Sequential Diagnosis Benchmark based on 304 recent case records published in the New England Journal of Medicine. These cases are extremely difficult to diagnose and often require multiple specialists and diagnostic tests to reach a diagnosis. What they created reminds me of the very old text-based adventure games. You can think about each of the cases they used as a level you need to complete by giving a diagnosis. You are presented with a case, and you can type in your questions or request diagnostic tests. You get responses, and you can continue with questions or tests until you figure out the diagnosis. Obviously, to know what questions to type in, you have to be a doctor. And like a proper game, it shows how much money you have spent on tests. The goal of the game is to spend the least amount of money to give the correct diagnosis. Because the game (pardon me, benchmark) is in the form of chat, it can be played by chatbots. They tested ChatGPT, Llama, Claude, Gemini, Grok, and DeepSeek. To better harness the power of the AI models, the team developed Microsoft AI Diagnostic Orchestrator (MAI-DxO). It emulates a virtual panel of physicians. MAI-DxO paired with OpenAI's o3 was the most efficient, correctly solving 85.5% of the NEJM benchmark cases. They also evaluated 21 practicing physicians, each with at least 5 years of clinical experience. These experts achieved a mean accuracy of 20%; however, they were denied access to colleagues and textbooks (and AI), as the team deemed such comparison to be more fair. More Tech Stocks: Amazon tries to make AI great again (or maybe for the first time)Veteran portfolio manager raises eyebrows with latest Meta Platforms moveGoogle plans major AI shift after Meta's surprising $14 billion move I strongly disagree with the idea that the comparison is fair. If a doctor is facing a difficult to diagnose issue and does not consult a colleague or refer you to a specialist, or look through his books to jog his memory, what kind of doctor is that? The team noted that further testing of MAI-DxO is needed to assess its performance on more common, everyday presentations. However, there is an asterisk. I write a lot about AI, and I think it is just pattern matching. The data on which models have been trained is typically not disclosed. If o3 has been trained on NEJM cases, it's no wonder it can solve them. The same is true if it was trained on very similar cases. Back to my issue. My friend, who is a retired pulmonologist, had a solution. Who'd ask a lung doctor for a disease affecting the legs? Well, she is also an Ayurvedic doctor and a Yoga teacher. She thinks outside the box. I was given a simple exercise that solved my problem. Years have passed, and if I stop doing it regularly, my symptoms return. What I know for sure is that no AI could ever come up with it. Another problem is that even if this tool works, and doctors start using it, they'll soon have less than a 20% success rate on the "benchmark." You lose what you don't use. Related: How Apple may solve its Google Search problem The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc.
