Deep-Cog

Detecting abnormal cognitive decline before the appearance of initial symptoms is crucial for early intervention that will directly impact the quality of life and well-being of all patients, families, and society. Furthermore, it will also help to develop new therapies to mitigate the age-associated loss of function and provide information to better understand cognitive decline. In recent studies by our team, we showed through PET analysis that patients with dementia have alterations in the neuroinflammatory response [1] and that the cerebellum plays an important role in episodic memory (one of the best predictors for cognitive decline) through fMRI and psychometric assessments [2]. However, no clinical evaluation or isolated brain measurement (biochemical, structural, or functional) can detect abnormal cognitive decline in preclinical stages, making multimodal studies mandatory.
Therefore, we will implement a Multimodal Deep Neural Network (mDNN) and train it with a hybrid dataset composed of data from public repositories and data collected by our team.
Our secondary goal is to uncover structural and functional brain changes related to abnormal cognitive decline in aging. The detection of such changes will enhance our basic understanding of cognitive decline processes, but importantly, here they will also be used as biomarkers for detecting abnormal cognitive decline, which might be crucial for providing early intervention therapies.