Multimodal Sequential Deep Learning for Agitation Detection in People Living with Dementia
Bibliographic record
Abstract
A common behavioral symptom in people living with dementia (PwD) is agitation. Agitation poses risks to the health and safety of both the patient and caregivers. Using sensor data from wearable devices is a promising means of detecting agitation events in a minimally invasive manner. Between 2017 - 2019, 600 days of sensor data was collected using an Empatica E4 wristband from 20 PwD. This paper investigates the application of sequential deep learning models on this unique sensor data to detect agitation in this population. Four deep learning architectures - Long Short-Term Models, Temporal Convolution Networks, Transformers and TS2Vec - are compared against each other and with previous results detecting agitation with classical machine learning models. We tested each model at various downsample factors and finds that Transformer-based models gave the highest AUC ROC and AUC PR scores. The findings also show the performance of the best performing deep learning models is comparable to the best performing machine learning models (random forest). This result underscores the potential of deep learning models in detecting agitation, as well as their potential to generalize to other similar clinical problems without the need for extensive feature engineering,
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 | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".