Luckett P, Paul RH, Navid J, et al. - Given the utility of deep learning algorithms of cerebral blood flow for classification of cognitive impairment and frailty in people living with HIV (PLWH) and that brain region identified in feature extraction techniques are the strongest predictors, researchers sought to classify virologically suppressed (< 50 copies/mL) PLWH (n = 125) on combination antiretroviral therapy with respect to their cognitive impairment and frailty. PLWH were classified using trained deep neural network (DNN) models as cognitively unimpaired or impaired based on neuropsychological tests (Hopkins Verbal Learning Test-Revised and Brief Visuospatial Memory Test-Revised, Trail making, Letter-Number Sequencing, Verbal Fluency, and Color Word Interference), as well as frail, prefrail, or nonfrail based on the Fried phenotype criteria (at least 3 of the following 5: weight loss, physical inactivity, exhaustion, grip strength, walking time). Outcomes indicate DNN models as capable of allowing classification of cognitive impairment and frailty status in PLWH with high accuracy. For cognitive impairment, the strongest predictors were cortical (parietal, occipital, and temporal) and subcortical (amygdala, caudate, and hippocampus) regions, whereas for frailty, the strongest predictors were subcortical (amygdala, caudate, hippocampus, thalamus, pallidum, and cerebellum).