Machine learning-based prediction of cognitive outcomes in de novo Parkinson’s disease

Abstract

Cognitive impairment is a debilitating symptom in Parkinson's disease (PD). We aimed to establish an accurate multivariate machine learning (ML) model to predict cognitive outcome in newly diagnosed PD cases from the Parkinson's Progression Markers Initiative (PPMI). Annual cognitive assessments over an eight-year time span were used to define two cognitive outcomes of i) cognitive impairment, and ii) dementia conversion. Selected baseline variables were organized into three subsets of clinical, biofluid and genetic/epigenetic measures and tested using four different ML algorithms. Irrespective of the ML algorithm used, the models consisting of the clinical variables performed best and showed better prediction of cognitive impairment outcome over dementia conversion. We observed a marginal improvement in the prediction performance when clinical, biofluid, and epigenetic/genetic variables were all included in one model. Several cerebrospinal fluid measures and an epigenetic marker showed high predictive weighting in multiple models when included alongside clinical variables.

Predicting the occurrence of mild cognitive impairment in Parkinson’s disease using structural MRI data

“…To date, only a few studies have explored the potential of machine learning techniques to predict cognitive outcomes in PD before clinical symptoms arise, and these studies primarily utilize clinical variables for this purpose (Smith et al, 2021;Harvey et al, 2022). For instance, a predictive model trained on clinical and biological parameters exhibited robust accuracy in predicting cognitive impairment and maintaining normal cognition over an 8-year follow-up period, with an AUC of 0.86 (Harvey et al, 2022).…”

“…To date, only a few studies have explored the potential of machine learning techniques to predict cognitive outcomes in PD before clinical symptoms arise, and these studies primarily utilize clinical variables for this purpose (Smith et al, 2021;Harvey et al, 2022). For instance, a predictive model trained on clinical and biological parameters exhibited robust accuracy in predicting cognitive impairment and maintaining normal cognition over an 8-year follow-up period, with an AUC of 0.86 (Harvey et al, 2022). The relevance of clinical metrics, such as anxiety and olfactory impairment, as well as biological markers like DNA methylation, is also highlighted in this study, indicating their possibility of being used as indicators for cognitive outcomes in PD (Harvey et al, 2022).…”

“…For instance, a predictive model trained on clinical and biological parameters exhibited robust accuracy in predicting cognitive impairment and maintaining normal cognition over an 8-year follow-up period, with an AUC of 0.86 (Harvey et al, 2022). The relevance of clinical metrics, such as anxiety and olfactory impairment, as well as biological markers like DNA methylation, is also highlighted in this study, indicating their possibility of being used as indicators for cognitive outcomes in PD (Harvey et al, 2022). The efficacy of using cortical structure was also assessed in predicting cognitive performance in PD patients, at least 3 years before the onset of MCI symptoms, yielding an AUC of 0.72 (Smith et al, 2021).…”

“…4 The predictor object, comprising the finalized model and the test dataset, was used to calculate the Shapley values of each sample with 10,000 Monte Carlo simulations. The absolute Shapley values were then averaged across all the samples, giving an overall assessment of the global Shapley contribution of each variable, as described in Harvey et al (2022).…”

“…Machine learning technologies have been utilized to develop algorithms that classify PD-MCI vs. PD-HC based on structural T1-weighted (T1w) MRI images, and identified the right anterior entorhinal cortex (BA 34) (Cho, 2019) and right caudate nucleus (Shibata et al, 2022) as most contributive regions for this classification. There have been only a limited number of studies that have investigated the use of machine-learning methods to predict cognitive outcomes in PD before the onset of clinical symptoms, and these studies have primarily relied on clinical data for their analysis (Smith et al, 2021;Harvey et al, 2022).…”