Automated Assessment of Manual Dexterity Using a Sensorized Nine-Hole Peg Test Board: Reproducibility and Innovative Quantitative Metrics.

The Nine-Hole Peg Test (9-HPT) is widely used in clinical settings but typically relies on the assessor's expertise to record execution time. Here, we propose a novel sensorized 9-HPT capable of automatically measuring total execution time and, importantly, extracting a set of newly defined temporal parameters that enable a more detailed and objective characterization of task performance. We first demonstrated concurrent validity between the sensorized 9-HPT and stopwatch-based measurements recorded by an assessor (ρ always > 0.98; p < 0.001) in healthy participants. Agreement between methods was further supported by the Bland-Altman analysis, showing negligible bias and narrow limits of agreement. A linear mixed-effects model confirmed no systematic differences between methods but showed significant differences between dominant and non-dominant hands. Test-retest reliability of total completion time, assessed across two sessions, was good for both the dominant (ICC = 0.81) and non-dominant (ICC = 0.74) hands. The newly introduced temporal parameters also showed significant reliability (ICC = 0.73-0.78), particularly for the dominant hand. Overall, these findings support the reliability of the sensorized 9-HPT for standard outcome measures and highlight its added value in providing novel temporal metrics that more precisely capture the different phases of task execution.