Mobile and wearable data fusion improves mental health severity predictions
MC-TRCM: Observation-Aware Recursive Fusion for Incomplete Mobile and Wearable Mental-Health Feature Views
Machine Learning
Summary
Mental health data from phones and wearables often come in summaries that miss some details or sources. The authors developed a new method that treats each data source separately and knows when data is missing. Their approach combines this varied information over time to better predict depression and anxiety severity. Tests showed it improved prediction accuracy compared to other methods on real mental health datasets.
What this means in practice
- •For mental health platform developers: Integrate modality-aware fusion to improve prediction accuracy from incomplete phone and wearable mental health data streams.
- •For mobile app developers: Develop apps that combine diverse sensor summaries while handling missing data to provide better mental health monitoring tools.
Authors
Wentao Wang, Lifeng Han, Zining Ren, Hengyu Zhong, Guangyu Zou
Abstract
Public mobile and wearable mental-health datasets often provide summarized feature tables rather than synchronized raw sensor streams. In these releases, each anchor corresponds to a survey or label time and may combine phone or wearable summaries, prior symptom scores, demographics, clinical variables, and source-availability indicators. We propose the Modality-Conditioned Temporal Recursive Context Model (MC-TRCM), which preserves each feature source as a separate token and incorporates missingness as part of the input context. Observed sources are encoded with values and missingness summaries, absent sources use learned absence tokens, dataset and task embeddings condition fusion, and a recursive prediction head refines each output over validation-selected steps. We evaluated MC-TRCM on six predefined endpoints from DepreST-CAT and Prediction of Severity Change-Depression (PSYCHE-D) using participant-level splits and validation-only model selection. MC-TRCM achieved the lowest mean absolute error on DepreST-CAT Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7) severity, improving over the best tabular reference by 0.181 and 0.217 scale points. Classification endpoints showed task-dependent behavior: MC-TRCM matched the best rounded GAD-7 category balanced accuracy, was numerically highest by 0.002 balanced-accuracy points on PSYCHE-D multiclass prediction, and remained close to the strongest references on PHQ-9 category and PSYCHE-D binary prediction. Ablations support Feature-wise Linear Modulation, absence tokens, missingness projections, and recursive refinement, while calibration and feature-source controls characterize endpoint behavior. Our code is available at https://github.com/Botwwt/MC-TRCM.