MétaCan
Menu
Back to cohort
Record W7125701415 · doi:10.1109/medai67139.2025.00015

Fine-Tuning Pre-trained Transformer-Based Models for Sentence-Level Medical Text Classification

2025· article· W7125701415 on OpenAlexaff
Tamanna Kaiser, Dan Wu

Bibliographic record

Venuenot available
Typearticle
Language
FieldComputer Science
TopicTopic Modeling
Canadian institutionsUniversity of Windsor
Fundersnot available
KeywordsDiceClass (philosophy)TransferabilityFunction (biology)Test (biology)Randomized controlled trial

Abstract

fetched live from OpenAlex

This paper presents the development of sentence-level medical text classifiers by fine-tuning eight pre-trained transformer-based models on the PubMed 20k RCT dataset. The models span both general-purpose and biomedical-specific architectures. To enhance performance and address class imbalance, a composite loss function combining cross-entropy, focal loss, and dice loss was applied during training. The fine-tuned models were trained on PubMed 20k RCT and then applied, without further adaptation, to the MTSamples dataset using balanced and imbalanced test subsets. ClinicalBERT achieved the highest results, reaching 97.15% accuracy and 96.93% F1-score on PubMed 20k RCT, 95.20% accuracy and 95.10% F1-score on the balanced MTSamples subset, and 91.80% accuracy and 90.60% F1-score on the imbalanced subset, indicating strong transferability across structured and unstructured medical texts. These outcomes highlight the effectiveness of domain-specific fine-tuning combined with optimized training strategies in building accurate and adaptable medical sentence classifiers.

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 imitation

Not 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.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.008
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.004
Threshold uncertainty score0.013

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0020.008
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0010.001
Science and technology studies0.0000.000
Scholarly communication0.0010.002
Open science0.0010.001
Research integrity0.0010.002
Insufficient payload (model declined to judge)0.0020.003

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.

Opus teacher head0.099
GPT teacher head0.320
Teacher spread0.221 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreMethods

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".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same topicTopic ModelingFrench-language works237,207