Cambridge Researchers Develop AI Tool To Predict Early Dementia Progression

Scientists at the University of Cambridge have created an innovative Artificial Intelligence (AI) tool designed to identify early-stage dementia patients and predict their likelihood of developing Alzheimer’s disease.

This groundbreaking advancement comes in response to the growing global healthcare challenge posed by dementia, which currently affects over 55 million individuals and incurs an estimated annual cost of $820 billion.

With cases projected to nearly triple over the next 50 years, the need for effective diagnostic tools is critical.

The Cambridge researchers developed their AI model using routinely collected, non-invasive, and cost-effective patient data, including cognitive assessments and structural MRI scans that reveal grey matter shrinkage.

The study involved over 400 participants from a research cohort in the United States.

The model’s effectiveness was further validated with real-world patient data from an additional 600 participants in the US, along with longitudinal data from 900 individuals in memory clinics across the United Kingdom and Singapore.

Remarkably, the algorithm demonstrated a capability to distinguish between individuals with stable mild cognitive impairment and those who would go on to develop Alzheimer’s within a three-year timeframe.

According to findings published in the journal ‘EClinicalMedicine’, the model successfully identified individuals who developed Alzheimer’s disease with an accuracy of 82%, while correctly recognizing those who did not progress with an accuracy of 81%.

These promising results however indicate the potential for the model’s reliability in clinical settings.

Professor Zoe Kourtzi from the Department of Psychology at Cambridge emphasized the tool’s predictive capabilities, noting its real-world applicability allows for broader generalizations.

Ben Underwood, also from the University of Cambridge, highlighted the tool’s potential to alleviate concerns for patients and their families, offering a clearer understanding of their health trajectories.

Also Read: Researchers Develop Tool To Predict Breast Cancer Treatment Nerve Damage