Connectome-based prediction of global cognitive performance in people with HIV

• Main Authors: Fan Nils Yang, Shiva Hassanzadeh-Behbahani, Margarita Bronshteyn, Matthew Dawson, Princy Kumar, David J. Moore, Ronald J. Ellis, Xiong Jiang
• Format: Article
• Language: English
• Published: Elsevier 2021-01-01
• Series: NeuroImage: Clinical
• Subjects: Magnetic resonance imaging; HIV; Machine Learning; Neurodegenerative Diseases; Cognitive function; Connectome-based predictive modeling
• Online Access: http://www.sciencedirect.com/science/article/pii/S2213158221001212

Global cognitive performance plays an important role in the diagnosis of HIV-associated neurocognitive disorders (HAND), yet to date, there is no simple way to measure global cognitive performance in people with HIV (PWH). Here, we performed connectome-based predictive modeling (CPM) to pursue a neural biomarker of global cognitive performance in PWH based on whole-brain resting-state functional connectivity. We built a CPM model that successfully predicted individual differences in global cognitive performance in the training set of 67 PWH by using leave-one-out cross-validation. This model generalized to both 33 novel PWH in the testing set and a subset of 39 PWH who completed a follow-up visit two years later. Furthermore, network strengths identified by the CPM model were significantly different between PWH with HAND and without HAND. Together, these results demonstrate that whole-brain functional network strengths could serve as a potential neural biomarker of global cognitive performance in PWH.