AI summaryⓘ
The authors created a computer system that uses data from a wrist device to spot and identify repetitive behaviors like hair pulling and skin picking, which are linked to anxiety and obsessive-compulsive disorders. This device collects different types of information, such as movement, heat, and distance from objects, to help the system understand the behaviors better. Their method combined several machine learning techniques and was very accurate in telling these behaviors apart from normal actions and also in distinguishing between different types of these behaviors. They found that movement and distance data were most helpful for recognizing the behaviors. This work could help develop wearable tools for real-time monitoring and support in mental health care.
body-focused repetitive behaviorsconvolutional neural networkgated recurrent unitautoencodertime-of-flight sensorinertial measurement unitShapley additive explanationsmultimodal sensor fusionbehavioral monitoringwearable device
Authors
Samaneh Rezaeimanesh, Mohsen Behradfar, Mohammad Fili, Guiping Hu
Abstract
Body-focused repetitive behaviors, such as hair pulling and skin picking, are compulsive motor actions commonly associated with obsessive-compulsive and anxiety disorders. Their early, objective detection remains difficult because the movements are subtle and overlap with ordinary, non-pathological gestures. We developed and evaluated a multimodal deep learning framework to detect and classify these behaviors from wrist-worn sensor data. The data, collected by the Child Mind Institute using the Helios wrist-worn device, combine inertial measurement units, thermopile sensors, and time-of-flight sensors, capturing kinematic, thermal, and proximity information. The framework combined a convolutional neural network with a gated recurrent unit, alongside modality-specific autoencoders and a late-fusion classifier, to exploit temporal and spatial dynamics. It achieved an F1 score of 0.985 and an area under the receiver operating characteristic curve of 0.997 for binary detection, distinguishing these behaviors from other activities, and a macro-averaged F1 score of 0.700 with an area under the curve of 0.963 across a nine-class scheme that distinguished each individual behavior from a single grouped Non-Target class, improving over single-modality baselines. Post-hoc interpretability based on Shapley additive explanations showed that the time-of-flight and inertial modalities dominated discriminative power by capturing spatial proximity and dynamic movement, while hierarchical clustering indicated that misclassifications were driven primarily by the anatomical region of the gesture. These findings demonstrate that multimodal sensor fusion enables accurate, objective, and continuous behavioral monitoring. This work establishes a foundation for real-time, wearable-assisted mental health diagnostics and personalized interventions in biomedical research and clinical care.