Validation of an automated scoring program for a digital complex figure copy task within healthy aging and stroke

Area of Science:

• Neuropsychology
• Digital Health Technology
• Medical Diagnostics

Background:

• Complex Figure Copy Tasks are standard neuropsychological assessments.
• Manual scoring is time-consuming, requires expertise, and can yield inconsistent results.
• Automated scoring offers a potential solution to these limitations.

Purpose of the Study:

• To develop and evaluate a novel automated method for scoring tablet-based Complex Figure Copy Tasks.
• To assess the accuracy and reliability of automated scoring compared to manual scoring.
• To determine the program's ability to distinguish between different clinical impairment groups.

Main Methods:

• A cohort of 261 healthy adults and 203 stroke survivors completed a digital Figure Copy Task.
• Responses were scored by two human raters and a novel automated scoring program.
• Statistical analyses included intraclass correlation coefficients and Receiver Operating Curve analysis.

Main Results:

• The automated program achieved high sensitivity (92.10%) and specificity (90.20%) in identifying figure elements.
• Automated total scores showed strong agreement with manual scores (ICC = .83).
• The program accurately categorized impairment levels, distinguishing between healthy adults, subacute, and longer-term stroke survivors.

Conclusions:

• The automated scoring program reliably and accurately scores Figure Copy Task data, including in cases with fine-motor deficits.
• This technology provides immediate, unbiased, and reproducible scores.
• The program represents a significant advancement for clinical and research applications of the Figure Copy Task.