Probability-Invariant Random Walk Learning on Gyral Folding-Based Cortical Similarity Networks for Alzheimer's and Lewy Body Dementia Diagnosis
Researchers' probability-invariant random walk method analyzes gyral folding networks without requiring standardized brain maps.
Researchers led by Minheng Chen developed a probability-invariant random walk learning framework that diagnoses Alzheimer's and Lewy body dementia by analyzing individualized gyral folding patterns in cortical similarity networks. The method works without requiring node alignment or fixed network topology, overcoming limitations of atlas-based approaches. In experiments on clinical cohorts, it outperformed existing gyral folding and atlas-based models, demonstrating robust diagnostic potential for distinguishing between dementia types with overlapping symptoms.
Why It Matters
Enables more accurate, personalized dementia diagnosis by analyzing individual brain anatomy rather than forcing brains into standardized templates.