Imagine you have a bag of mixed candies, but you’re not sure which ones are the best. Using a super-smart candy classifier and a special gadget that analyzes each candy’s flavor, texture, and color, we’ve discovered the top predictors that determine which candies will become your favorites and which ones will end up being just okay. Similarly, researchers have used machine learning algorithms and explainable artificial intelligence methods to identify the leading predictors of dementia in people with Parkinson’s disease (PD). By studying a group of PD patients over three years and analyzing multiple dementia risk factors like motor abilities, cognition, and imaging data using a Random Forest model, they were able to find the key indicators that differentiated those who later developed dementia from those who did not. The findings revealed that factors like poor gait, slower cognitive processing, specific molecular changes, older age, certain brain measurements, and daily lifestyle activities were all important markers for predicting dementia onset in PD patients. This exciting research demonstrates the potential for using advanced technologies to better understand and predict cognitive decline in Parkinson’s disease. If you want to dive deeper into this fascinating study, grab your research goggles and check out the full article below!

BackgroundPersons with Parkinson’s disease (PD) differentially progress to cognitive impairment and dementia. With a 3-year longitudinal sample of initially non-demented PD patients measured on multiple dementia risk factors, we demonstrate that machine learning classifier algorithms can be combined with explainable artificial intelligence methods to identify and interpret leading predictors that discriminate those who later converted to dementia from those who did not.MethodParticipants were 48 well-characterized PD patients (Mbaseline age = 71.6; SD = 4.8; 44% female). We tested 38 multi-modal predictors from 10 domains (e.g., motor, cognitive) in a computationally competitive context to identify those that best discriminated two unobserved baseline groups, PD No Dementia (PDND), and PD Incipient Dementia (PDID). We used Random Forest (RF) classifier models for the discrimination goal and Tree SHapley Additive exPlanation (Tree SHAP) values for deep interpretation.ResultsAn excellent RF model discriminated baseline PDID from PDND (AUC = 0.84; normalized Matthews Correlation Coefficient = 0.76). Tree SHAP showed that ten leading predictors of PDID accounted for 62.5% of the model, as well as their relative importance, direction, and magnitude (risk threshold). These predictors represented the motor (e.g., poorer gait), cognitive (e.g., slower Trail A), molecular (up-regulated metabolite panel), demographic (age), imaging (ventricular volume), and lifestyle (activities of daily living) domains.ConclusionOur data-driven protocol integrated RF classifier models and Tree SHAP applications to selectively identify and interpret early dementia risk factors in a well-characterized sample of initially non-demented persons with PD. Results indicate that leading dementia predictors derive from multiple complementary risk domains.

Read Full Article (External Site)