Deep Learning Models to Screen Electronic Health Records for Breast and Colorectal Cancer Progression: Performance Evaluation Study
Bibliographic record
Abstract
BACKGROUND: Cancer progression is an important outcome in cancer research. However, it is frequently documented only in electronic health records (EHRs) as unstructured text, which requires lengthy and costly chart reviews to extract for retrospective studies. OBJECTIVE: This study aimed to evaluate the performance of 3 deep learning language models in determining breast and colorectal cancer progression in EHRs. METHODS: EHRs for individuals diagnosed with stage 4 breast or colorectal cancer between 2004 and 2020 in Manitoba, Canada, were extracted. A chart review was conducted to identify cancer progression in each EHR. Data were analyzed with pretrained deep learning language models (Bio+ClinicalBERT, Clinical-BigBird, and Clinical-Longformer). Sensitivity, positive predictive value, area under the curve, and scaled Brier scores were used to evaluate performance. Influential tokens were identified by removing and adding tokens to EHRs and examining changes in predicted probabilities. RESULTS: Clinical-BigBird and Clinical-Longformer models for breast and colorectal cancer cohorts demonstrated higher accuracy than the Bio+ClinicalBERT models (scaled Brier scores for breast cancer models: 0.70-0.79 vs 0.49-0.71; scaled Brier scores for colorectal cancer models: 0.61-0.65 vs 0.49-0.61). The same models also demonstrated higher sensitivity (breast cancer models: 86.6%-94.3% vs 76.6%-87.1%; colorectal cancer models: 73.1%-78.9% vs 62.8%-77.0%) and positive predictive value (breast cancer models: 77.9%-92.3% vs 80.6%-85.5%; colorectal cancer models: 81.6%-86.3% vs 72.9%-82.9%) compared to Bio+ClinicalBERT models. All models could remove more than 84% of charts from the chart review process. The most influential token was the word progression, which was influenced by the presence of other tokens and its position within an EHR. CONCLUSIONS: The deep learning language models could help identify breast and colorectal cancer progression in EHRs and remove most charts from the chart review process. A limited number of tokens may influence model predictions. Improvements in model performance could be obtained by increasing the training dataset size and analyzing EHRs at the sentence level rather than at the EHR level.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".