Cross-Platform Availability of Smartphone Sensors for Depression Indication Systems: Mixed-Methods Umbrella Review
Bibliographic record
Abstract
BACKGROUND: A popular trend in depression forecasting research is the development of machine learning models trained with various types of smartphone sensor data and periodic self-ratings to derive early indications of changes in depression severity. While most works focus on model performance, there is little concern about the universal usability and reliable operation of such systems across smartphone platforms. This review serves as foundational work for the MENTINA clinical trial, which investigates smartphone-based health self-management for depression. The usability and reliability of mobile apps for depression are commonly perceived through the lens of the approaches and interventions offered rather than the reliability of the built-in mobile phone functions to support effortless and exact delivery of intended interventions. OBJECTIVE: This work aimed to synthesize existing systematic reviews to identify smartphone sensor modalities used in mental health monitoring and, building on this foundation, assess the cross-platform availability of these data streams using PhoneDB to inform the design and implementation of digital depression indication systems. METHODS: To identify the already used hardware and software sensors and their purposes in mental health monitoring, an umbrella review was conducted. Three electronic databases, including PubMed, Web of Science Core Collection, and Scopus, were searched using smartphone, sensor data, and depression keyword combination to retrieve relevant literature reviews published within the last 5 years (2019-2024). Once the initial search was completed, the extracted hardware sensors were checked for availability on Android and iOS smartphones by analyzing device specifications in PhoneDB over the last 10 years. RESULTS: The extracted data streams observed across the 9 included studies covered 16 hardware and 3 software data streams. Hardware data streams included accelerometers, barometers, battery levels, Bluetooth, cameras, cellular networks, GPSs, gyroscopes, humidity, light sensors, magnetometers, proximity sensors, sound sensors, step counters, temperature sensors, and Wi-Fi. Software data streams included app usage, call and message logs, and screen status. Hardware component availability on Android and iOS systems showed the changes in component trends from 2014 to 2024 as of September 2024, with the accelerometers, batteries, cameras, and GPSs remaining consistent on Android and iOS, while components such as gyroscopes, step counters, and barometers gradually increased over the years, particularly on Android. CONCLUSIONS: Multiple data streams identified in the literature review showed a consistent increase in availability over time, enabling improved use of these outputs for depression forecasting and the training of machine learning models with diverse smartphone data, including sensor-derived information. For more precise and reliable data to be used in the mental health field, particularly in critical areas such as tracking and predicting changes in depression severity, further research is required to streamline smartphone data across varying mobile hardware and software configurations to provide reliable output for digital mental health purposes.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.021 | 0.092 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.009 | 0.011 |
| Bibliometrics | 0.021 | 0.014 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.006 | 0.004 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.008 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".