A Novel AI technology allowed researchers to capture the FUTURE 4D brain degeneration for the first time. This combo algorithms portray not only the patterns of regional brain cortical thinning in 4D but also the phenotypical behavior in standard clinical dementia rating (CDR) scores, potentially facilitating early detection of dementia in sub-healthy group, delaying the aging process,contributing to the development of early interventionpreventive medicine.

Personalized 4D brain degeneration maps AI prediction module for healthy agingdementia combined the multiple data types of omicsimaging biomarkers by innovative deep learning algorithm, conditional variational autoencoder - generative adversarial network (CVAE-GAN), to establish a precisereliable disease risk assessment platform, which can predict future clinical dementia rating (CDR) score (accuracy=89.7)determine the phenotype of symptoms (accuracy=88.0) as well as the dementia subtypes (accuracy=90.3). This automated system can provide guidelines for early treatment strategies for dementia.

The technique of “Personalized 4D Prediction of Brain Degeneration Maps in Healthy AgingDementia” comprised four core technologies jointing efforts from multi-disciplinary teams (clinical experts specializing in neurologyneuroradiology, big data analystsAI experts),internationallocal industry manufacturers. For clinicalindustrial applications, the groundbreaking core technologies are extremely applicable for earlyaccurate prediction of future dementia for sub-healthy group, assisting in risk screening, patient stratification,personalized treatment.