Autonomous Heart Rate Tracking Methodology Using Kalman Filter and the EM Algorithm
Bibliographic record
Abstract
Accurate heart rate monitoring during intense physical activity is a challenging problem due to high levels of motion artifacts (MA) in sensors that rely on stable physical proximity/contact for accurate measurement extraction. Photo-plethysmography (PPG) sensor is a non-invasive optical sensor that is widely used in wearable devices, such as smartwatches, to measure blood volume changes using the property of light reflection and absorption; these measurements can be used to extract the heart rate (HR) of an individual wearing that device. The PPG sensor is susceptible to the motion artifact which increases with physical activity. Since the frequency of the motion artifact is very close to the range of HR, estimation of HR information becomes very challenging. As a result, MA removal remains an active research topic over the last few years. Several approaches have been developed in the recent past for MA removal and accurate HR estimation. Among these recent works, a Kalman Filter (KF) based approach showed promising results for accurate estimation and tracking of HR based on PPG measurements. However, the previous KF based HR tracker was demonstrated for a particular dataset with manually tuned filter parameters. Such a custom tuned approach might not perform accurately in practical scenarios where the amount of motion artifact and the heart-rate variability depend on numerous, unpredictable factors. In this paper, we develop an approach to automatically tune the KF based HR tracker based on the expectation maximization (EM) algorithm. The applicability of the proposed approach is demonstrated using an open-source PPG database that was recorded during varying pre-determined physical activities.
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) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".