Fig. 1: Analysis pipeline.

a Participants read and immediately retold an AT and an nAT. b P-RSF scores were extracted from each subject’s AT and nAT retelling. c Statistical between-group comparisons were made via ANCOVAs, covarying for MoCA and IFS scores. d Classification analyses were based on support vector machines, with results represented via receiver operating characteristic (ROC) curves, confusion matrices, and distribution plots of P-RSF scores. These analyses were applied to discriminate between (i) all PD patients and all HCs, (ii) PD-nMCI patients and HCs, (iii) PD-MCI patients and HCs, and (iv) PD-nMCI and PD-MCI patients. AT action text, nAT non-action text, ANCOVA analysis of covariance, MoCA Montreal Cognitive Assessment, IFS INECO Frontal Screening, P-RSF Proximity-to-Reference-Semantic-Field, PD Parkinson’s disease, PD-MCI Parkinson’s disease with mild cognitive impairment, PD-nMCI Parkinson’s disease without mild cognitive impairment.