MAISON-LLF: Multimodal Sensor Dataset for Monitoring Older Adults Post Lower-Limb Fractures in Community Settings
Bibliographic record
Abstract
The MAISON-LLF dataset was collected from 10 older adult participants living alone in the community following lower limb fractures. Each participant contributed data for over 8 weeks, beginning from their first-week post-discharge. This resulted in a total of 560 days of continuous multimodal sensor data, complemented by biweekly clinical questionnaire data. The MAISON-LLF dataset is organized into a directory tree, as shown below. maison-llf/├── sensor-data/│ ├── p01/...│ ├── p10/│ │ ├── acceleration-data.csv│ │ ├── heartrate-data.csv│ │ ├── motion-data.csv│ │ ├── position-data.csv│ │ ├── sleep-data.csv│ │ ├── step-data.csv├── features/│ ├── p01/...│ ├── p10/│ │ ├── acceleration-features.csv│ │ ├── heartrate-features.csv│ │ ├── motion-features.csv│ │ ├── position-features.csv│ │ ├── sleep-features.csv│ │ ├── step-features.csv│ │ ├── clinical.csv├── dataset/│ ├── all-features.csv│ ├── all-features-imputed.csv│ ├── dataset-daily.pt│ ├── dataset-weekly.pt│ ├── dataset-biweekly.pt│ ├── In ‘sensor-data’ folder, the dataset includes 60 CSV files containing data from six sensor types for 10 participants. Each file includes a ‘timestamp’ column indicating the date and time of the recorded sensor data, accurate to milliseconds (‘yyyy-MM-dd HH:mm:ss.SSS’), along with the corresponding sensor measurements. For instance, the ‘acceleration-data.csv’ files include four columns: timestamp, and x, y, and z coordinates, while the ‘heartrate-data.csv’ files contain two columns: timestamp and heart rate value. The dataset also includes 70 CSV files containing daily features extracted from the sensor data, along with clinical questionnaire data and physical test results. Each feature CSV file includes a timestamp column representing the date (‘yyyy-MM-dd’) of the sensor data from which the daily features were extracted, alongside the corresponding sensor features. For example, the ‘acceleration-features.csv’ files contain eight columns: timestamp and the seven acceleration features and the ‘heartrate-features.csv’ files include five columns: timestamp and the four heart rate features. Additionally, the ‘clinical.csv’ files provide values for individual items of the SIS (‘sis-01’ to ‘sis-06’), OHS (‘ohs-01’ to ‘ohs-12’), and OKS (‘oks-01’ to ‘oks-12’) questionnaires, along with their final scores (‘sis’, ‘ohs’, and ‘oks’). These files also include results for the TUG and 30-second chair stand tests. Each participant has four sets of clinical data, with each set sharing the same ‘timestamp’ corresponding to the date (‘yyyy-MM-dd’) on which the clinical data were collected. To provide a comprehensive overview of the dataset, the ‘all-features.csv’ and ‘all-features-imputed.csv’ files in ‘dataset’ folder combine all daily features, clinical data, and demographic information into single CSV files, representing the data before and after missing value imputation (as explained in subsection 2.2.4). Additionally, the Python PyTorch files are structured datasets designed to facilitate supervised and unsupervised machine learning model development for estimating clinical outcomes. ‘dataset-daily.pt’ in ‘dataset’ folder contains a NumPy array with dimensions num_days × num_features, representing the daily features for all 10 participants. Alongside this array, it includes a num_days IDs array that maps each day to a participant (IDs 1 to 10). Additionally, the file contains three separate num_days arrays for SIS, OHS, and OKS scores, each assigned to the corresponding days in the daily features array. ‘dataset-weekly.pt’ in ‘dataset’ folder provides an array with dimensions num_weeks × 7 × num_features, which includes the weekly sequential features for all participants. This file also includes a num_weeks IDs array to identify the participant (1 to 10) associated with each week in the samples array. Similar to the daily dataset, it contains three separate num_weeks arrays for the SIS, OHS, and OKS scores, each assigned to the respective weeks in the weekly features array. ‘dataset-biweekly.pt’ in ‘dataset’ folder provides an array with dimensions num_biweeks × 14 × num_features, which includes the biweekly sequential features for all participants. This file also includes a num_biweeks IDs array to identify the participant (1 to 10) associated with each biweekly period in the samples array. Similar to the daily dataset, it contains three separate num_biweeks arrays for the SIS, OHS, and OKS scores, each assigned to the respective biweekly periods in the biweekly features array. CitationCite the related pre-print: A. Abedi, C. H. Chu, and S. S. Khan, "Multimodal Sensor Dataset for Remote Monitoring of Older Adults Post Lower-Limb Fractures in the Community,"
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 | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.009 | 0.007 |
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".