Artificial intelligence is moving beyond hospital labs into homes, blood tests, and drug discovery pipelines, offering doctors new ways to spot disease before symptoms appear. Three emerging applications show how AI is reshaping preventive medicine: detecting subtle changes in daily life that signal stroke risk, predicting joint disease years before damage occurs, and using blood biomarkers to understand what drives Alzheimer's disease. Yet researchers emphasize that AI's greatest promise lies not in replacing doctors, but in flagging high-risk patients early and helping clinicians intervene before irreversible harm occurs.

How Can AI Detect Disease Before Symptoms Show Up?

A team of researchers from South Korea's KAIST, Sungkyunkwan University, and Korea University Anam Hospital developed an AI system that analyzes daily activity patterns, sleep rhythms, and indoor environmental data to identify early warning signs of cerebrovascular disease. The system was trained on lifelog data collected from 1,224 older adults in their homes, analyzing 13,362 two-week samples of real-world behavior.

The AI identified a striking pattern: people in the early stages of cerebrovascular disease showed frequent continuous activity between 10 p.m. and 2 a.m., when the body normally prepares for sleep. As diagnosis approached, evening activity from 6 p.m. to 10 p.m. decreased noticeably, while inactive time increased. Low indoor humidity also emerged as a risk factor. When the AI classified data from within four weeks before diagnosis as "imminent risk" and data from 12 weeks before as "non-imminent," it distinguished between the two periods with 96.53% accuracy.

"The key point of this study is not that AI should replace a hospital diagnosis, but that it can first detect risk signals in small lifestyle changes at home and help connect patients to medical care at the right time," said Professor Lisa Lim, Department of Civil and Environmental Engineering at KAIST.

Professor Lisa Lim, Department of Civil and Environmental Engineering, KAIST

The researchers stressed that this technology is designed as a supportive tool for prevention and early consultation, not as a diagnostic replacement. Prospective validation in larger patient groups will be necessary before clinical deployment.

What Role Does AI Play in Predicting Joint Disease?

Ilya Burkov, PhD, now global head of healthcare and life sciences at Nebius, an AI cloud company, spent his early career studying osteoarthritis and osteoporosis. His breakthrough came when colleagues asked whether he had considered a machine learning approach to analyzing long-term medical imaging data. The concept was simple but powerful: if AI could identify patterns shared by patients who later developed joint disease, it might detect subtle biomarkers years before clinical symptoms appeared.

Burkov developed techniques capable of predicting early-onset osteoarthritis and osteoporosis with 80% to 90% accuracy. The AI models identified imaging features that consistently appeared years before patients required joint replacement surgery. In some cases, clinicians could tell patients that without lifestyle changes, they had a high probability of requiring a hip replacement within 10 to 15 years.

"That realization ultimately convinced me to transition from clinical medicine into industry, where I saw the opportunity to build technologies that could have a much broader impact," Burkov explained.

Ilya Burkov, PhD, Global Head of Healthcare and Life Sciences, Nebius

This shift from treating individual patients to creating scalable tools represents a fundamental change in how AI can support healthcare systems. Rather than applying clinical expertise one patient at a time, AI-powered technologies can help clinicians identify high-risk patients earlier and intervene before irreversible damage occurs.

How Are AI Models Uncovering the Drivers of Alzheimer's Disease?

Prima Mente, a London-based startup and previous AI Discovery Award winner, is tackling one of healthcare's most pressing challenges: understanding what actually drives Alzheimer's disease. The company combines blood-based biomarkers, multimodal biological data, and transformer-based AI (the same architecture underlying large language models like ChatGPT) to identify the molecular mechanisms behind neurodegenerative disorders.

The company's strategy draws inspiration from advances in cancer diagnostics, particularly liquid biopsy technologies that detect circulating tumor DNA in blood samples. Prima Mente wondered whether a similar approach could work for brain disease. Three years ago, many researchers thought the idea was unlikely; the prevailing view held that very little DNA from dying brain cells entered the bloodstream. The team has since demonstrated that cell-free DNA originating from neurons, microglia, and astrocytes can be detected in blood, and those fragments retain epigenetic information that reveals the biological state of brain cells before they died.

Rather than focusing solely on DNA sequences, Prima Mente analyzes methylation patterns carried on cell-free DNA. Methylation reflects how genes are regulated within specific cell types, providing insight into disease progression and cellular dysfunction. When cells die, they release fragmented DNA into the bloodstream, and those fragments preserve methylation signatures that reveal what state those brain cells were in.

"If ChatGPT can understand human language, our hypothesis is that similar models can understand biological languages," noted Hannah Madan, PhD, co-founder of Prima Mente.

Hannah Madan, PhD, Co-founder, Prima Mente

The company trains its AI models directly on raw biological sequences, including DNA, methylation signals, RNA transcripts, and proteomic data, rather than converting them into simplified numerical counts. By preserving more of the underlying biological information, Prima Mente believes its models could uncover relationships that conventional bioinformatics pipelines often overlook. The mission addresses a critical gap: dementia is the leading cause of death in the United Kingdom and the sixth highest in the United States, with Alzheimer's disease remaining the most common form of dementia worldwide.

What Challenges Remain for AI in Healthcare?

Despite these advances, experts emphasize that AI's success depends on more than sophisticated algorithms. Researchers highlighted the importance of several interconnected factors:

• Computing Infrastructure: Powerful computational resources are essential for training increasingly complex AI models that can process large biological datasets and identify subtle patterns in medical imaging and genomic data.
• High-Quality Biological Data: AI models are only as good as the data they learn from; researchers need access to diverse, well-annotated datasets that capture the full spectrum of disease presentations and patient populations.
• Laboratory Validation: Findings from AI models must be validated in controlled laboratory settings and clinical trials before they can be deployed in real-world healthcare settings to ensure safety and efficacy.
• Collaboration Across Sectors: Advancing preventive and personalized medicine requires partnerships among academia, healthcare institutions, industry, and governments to translate AI discoveries into clinical practice.

Participants at Nebius's AI Discovery Awards event in London emphasized that collaboration will be essential to move AI discoveries from the lab into clinical care. The event highlighted leading startups in biopharma, genomics, medical devices, and digital health that are using AI to deliver advances in healthcare and life sciences.

The emerging consensus is clear: AI is a powerful tool for identifying disease risk and understanding biological mechanisms, but it works best as part of a broader healthcare ecosystem that includes clinical judgment, patient examination, and diagnostic testing. The goal is not to replace doctors, but to give them better tools to catch disease earlier and intervene more effectively.