Currently submitted to: Journal of Medical Internet Research
Date Submitted: Dec 9, 2024
Open Peer Review Period: Dec 10, 2024 - Feb 4, 2025
(closed for review but you can still tweet)
NOTE: This is an unreviewed Preprint
Warning: This is a unreviewed preprint (What is a preprint?). Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn (a note "no longer under consideration" will appear above).
Peer review me: Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period (in this case, a "Peer Review Me" button to sign up as reviewer is displayed above). All preprints currently open for review are listed here. Outside of the formal open peer-review period we encourage you to tweet about the preprint.
Citation: Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).
Final version: If our system detects a final peer-reviewed "version of record" (VoR) published in any journal, a link to that VoR will appear below. Readers are then encourage to cite the VoR instead of this preprint.
Settings: If you are the author, you can login and change the preprint display settings, but the preprint URL/DOI is supposed to be stable and citable, so it should not be removed once posted.
Submit: To post your own preprint, simply submit to any JMIR journal, and choose the appropriate settings to expose your submitted version as preprint.
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Deep CNN-Based Continuous Monitoring of Depression Using Smartphone Usage Patterns: Algorithm Development and Validation
ABSTRACT
Background:
Depression is a global mental health challenge, with traditional assessment methods like the Patient Health Questionnaire-9 (PHQ-9) limited by infrequent data collection and susceptibility to recall bias. Recent advances in digital phenotyping offer the potential to capture real-time behavioral data through smartphones. However, existing models primarily focus on binary classification, often overlooking the need for continuous and granular symptom-specific monitoring.
Objective:
This study aims to address limitations in traditional depression assessments by developing and validating a convolutional neural network (CNN)-based framework. The objectives are to predict continuous PHQ-9 scores, enable symptom-specific analysis, and provide confident classifications of depression severity using passive smartphone data.
Methods:
A novel digital phenotyping approach was employed, leveraging raster plots to encode smartphone usage patterns in 48-hour windows. Data from 491 participants were collected via a smartphone application, which tracked app usage, screen time, and interaction frequency. Participants also completed periodic PHQ-9 assessments. The CNN model was trained using five-fold cross-validation, optimized through grid search, and benchmarked against a random forest model using metrics such as precision, recall, mean absolute error (MAE), and fraction classified (ƒ).
Results:
The CNN model demonstrated superior performance over the random forest baseline, achieving an overall accuracy of 83.1%, a precision of 90.3% for positive cases, and a low MAE of 0.81 for motor activity predictions. The fraction classified (ƒ) metric indicated 95% of cases were confidently categorized as either negative or positive, with only 5% falling into the uncertain range (PHQ-9 scores 10–15). Continuous tracking of PHQ-9 scores illustrated the model’s ability to monitor stable and dynamic depressive trajectories. For instance, User ID 38867 showed strong alignment between predictions and self-reports, while User ID 56084 highlighted the model's sensitivity to symptom variability.
Conclusions:
This study introduces a robust framework for passive, continuous depression monitoring, advancing the application of digital phenotyping in mental health care. By leveraging CNNs and raster plots, the approach bridges gaps in traditional assessments, providing actionable, symptom-specific insights. The results emphasize the potential for personalized, scalable mental health monitoring and support future integration of multimodal data and real-time feedback mechanisms for enhanced clinical applicability.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.