A study on the exploration of mild cognitive impairment in Parkinson's disease based on decision-making cognitive computing.

Mild cognitive impairment in Parkinson's disease (PD-MCI) as an independent risk factor for dementia in Parkinson's disease has prognostic value in predicting dementia in PD patients. It was found that the calculation of cognitive function decision-making could better evaluate the cognitive function of PD-MCI. Therefore, this study explored deficits in decision-making cognitive function in PD-MCI population, and mined novel digital biomarkers for recognizing early cognitive decline in PD-MCI through an independently designed maze decision-making digital assessment paradigm. This study included 30 healthy controls 37 PD with normal cognition (PD-NC) and 40 PD-MCI patients. Through difference comparison and stepwise regression analysis, two digital decision-making biomarkers, total decision time and performance average acceleration, were screened, and their joint area under curve for the ability to discriminate between PD-MCI and PD-NC was 0.909, and for the ability to discriminate between PD-MCI and NC was 0.942. In addition, it was found that maze digital decision-making biomarkers had greater early warning efficacy in men than in women. Unlike traditional methods, this study used digital dynamic assessment to reveal possible decision-making cognitive deficits in the PD-MCI populations, which provides new ideas for effective screening for PD-MCI.