AI models shifting healthcare from detection to prediction
New reports show AI is moving beyond detecting diseases to predicting future health risks using blood proteins and organ analysis.
The landscape of modern medicine is undergoing a fundamental transformation. Rather than simply identifying illnesses after they have already taken hold, Artificial Intelligence (AI) is now being leveraged to forecast future health risks. According to a recent report by Longevitty.ai, the medical field is moving from a reactive stance of disease detection to a proactive era of disease prediction.
Scientists are no longer just looking at individual health tests to see if a patient is currently sick. Instead, they are utilizing advanced AI to understand the trajectory of a person's health. The core question is shifting from "Do you have this condition?" to "Which direction is your body heading, and how fast is it moving?"
Predicting the future through blood proteins
One of the most significant breakthroughs involves the study of biological ageing and blood proteins. In 2024, researchers utilized AI to analyze protein data from 45,441 individuals within the UK Biobank. The study revealed that a specific combination of 204 proteins could effectively estimate a person's biological age.
The implications for preventative medicine are massive. The research found that individuals whose protein profiles suggested they were "older" than their chronological age faced a significantly higher risk of 18 major conditions, including:
- Heart disease
- Kidney disease
- Lung disease
- Diabetes
- Cancer
Organ-level analysis and the Delphi-2M model
Beyond bloodwork, AI is being used to study the human body organ by organ. Scientists at Stanford conducted research involving 44,498 people, discovering that the brain and the immune system serve as some of the most powerful predictors of longevity and overall health quality. Interestingly, the "age" of these organs fluctuates based on lifestyle choices and medication use.
Another leap forward is the European-developed AI model known as Delphi-2M. Functioning much like the predictive text on a smartphone keyboard, Delphi-2M does not predict words, but rather the next disease a person might develop and the timing of its onset. Key details of this model include:
- Training: Developed using data from 400,000 people in the UK.
- Testing: Validated against 1.9 million people in Denmark.
- Capability: Capable of predicting more than 1,000 different diseases.
The rise of Super Intelligence in clinics
The integration of these technologies is happening faster than many realize. Max Chopra, Founder and CEO of Longevitty.ai, noted that "Super Intelligence" is rapidly becoming a staple in clinical settings. A survey conducted this year revealed that 81 per cent of doctors are now using AI in their medical work, a massive jump from just 38 per cent in 2023.
However, this rapid adoption comes with caution. While doctors are embracing the technology, they have expressed concerns regarding patients using AI tools to self-diagnose without professional medical supervision. Chopra emphasized the importance of integrated health records, suggesting that a future "Super Intelligence" could synthesize a patient's entire medical history with their blood data to provide a complete health narrative.
As researchers continue to combine blood protein analysis, organ-level data, and comprehensive health records, the goal of preventing disease before it starts is moving closer to reality.