Learning Models for Diagnosis and Prognosis from Electrocardiogram Data
Bibliographic record
Abstract
The electrocardiogram (ECG) records the electrical activity of a patient’s heart movement. It is one of the standard routine healthcare tests as it is non-invasive and easy to apply. In this thesis, we analyze 2 million ECGs and over 260,000 patients’ health records from the Alberta Health Service, and propose frameworks for learning diagnostic and prognostic models based on supervised learning methods, including ones for survival prediction. First, we learned many models that each use a patient’s ECG to determine if s/he has a specific disease, corresponding to an ICD-10 diagnosis code. Our results show that these diagnosis models can accurately predict numerous health conditions, beyond cardiovascular conditions. Second, we develop ECG diagnosis models for COVID-19 and then use transfer learning to produce models with superior performance. Finally, motivated by the evidence from earlier tasks, we develop binary classification ECG models for predicting all-cause (fixed time) mortality for hospitalized (resp., emergency) patients, and also survival models that produce meaningful survival predictions for each patient. We demonstrate state-of-the-art performance for predicting the time-until-death by using machine learning techniques that first re-express each ECG in latent representations.
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 | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 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".