Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study developed a multimodal deep learning framework to detect body-focused repetitive behaviors (BFRBs) using data from a wrist-worn sensor device that captures kinematic, thermal, and proximity information. By integrating a convolutional neural network with gated recurrent units and modality-specific autoencoders, the framework achieved an impressive F1 score of 0.985 for binary detection and 0.700 for a nine-class classification scheme. The results highlight the effectiveness of sensor fusion in providing accurate and continuous monitoring of subtle behaviors associated with mental health disorders, paving the way for real-time diagnostics and interventions.
Achieving a near-perfect F1 score for detecting subtle body-focused repetitive behaviors could revolutionize mental health monitoring and intervention strategies.
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.