Learning classification models of cognitive conditions from subtle behaviors in the digital Clock Drawing Test

Abstract

The Clock Drawing Test -- a simple pencil and paper test -- has been used for more than 50 years as a screening tool to differentiate normal elderly individuals from those with cognitive impairment, and has proven useful in helping to diagnose dementias, such as Alzheimer's disease, Parkinson's disease, and other conditions. A group of hospitals and clinics have been administering the test using a digitizing ballpoint pen that reports its position with considerable spatial and temporal precision, making available far more detailed data about the subject's performance. Using categorized stroke data from these drawings, we designed and computed a large collection of features, then explored the tradeoffs in performance and interpretability in classifiers built using a number of different subsets of these features and a variety of different machine learning techniques. We used traditional machine learning methods to build prediction models that achieve high accuracy. We operationalized widely used existing scoring algorithms so that we could use them as benchmarks for our models. We worked with clinicians to define guidelines for model interpretability, and constructed sparse linear models and decision lists designed to be as easy to use as scoring algorithms currently used by clinicians, but more accurate. We also extract insights from the data about the behavioral aspect of these conditions on patients. While our models will require additional testing with subjects for validation, they offer the possibility of substantial improvement in detecting cognitive impairment earlier than currently possible, a development with considerable potential impact in practice.