MétaCan
Menu
← Back to cohort
Record W4405166223 · doi:10.1101/2024.12.07.24318566

Early identification of Family Medicine residents at risk of failure using Natural Language Processing and Explainable Artificial Intelligence

2024· preprint· en· W4405166223 on OpenAlexaff
Abhisht Joshi, Pouria Mortezaagha, Diana Inkpen, Edward Seale, Douglas Archibald, Kendall Noel, Arya Rahgozar

Bibliographic record

VenuemedRxiv · 2024
Typepreprint
Languageen
FieldMedicine
TopicArtificial Intelligence in Healthcare and Education
Canadian institutionsUniversity of Ottawa
Fundersnot available
KeywordsArtificial intelligenceCertificationIdentification (biology)Computer scienceMachine learningChampionSupport vector machineArtificial neural networkNatural language processing

Abstract

fetched live from OpenAlex

Abstract Background During residency, each resident is observed and receives feedback based on their performance. Residency training is demanding, with some residents struggling with their academic performance. A competency-based residency training program’s success depends on its ability to identify residents with difficulty during their first year of post-graduate education and to provide them with timely intervention and support. Objective In large training programs such as Family Medicine, identifying residents at risk of failing their certification exams is difficult. We developed an AI system using state-of-the-art technologies in Machine Learning (ML), Deep Learning (DL), Natural Language Processing (NLP) and Explainable AI (XAI) to detect at-risk residents automatically. Materials and Methods The research was conducted in the 2023-24 academic year. We implemented ML, DL and NLP models for prediction and performance analysis. The target variable chosen for the prediction was the determination of whether the resident would fail or pass their certification exam. XAI was used to enhance the understanding of the model’s inner workings. Results In total, there were 1382 data points of residents. The final model, Support Vector Machine (SVM), achieved an accuracy of 89.05% and an F1 score of 74.54 for the multiclass classification when multimodal (text and tabular) data was used. This model outperformed the models that only used qualitative or quantitative data exclusively. Conclusion Combining qualitative and quantitative data represents a novel approach and provided better classification results. This research demonstrates the feasibility of an automated AI system for the early identification of residents at risk of academic struggle. Prior Abstract Presentation Abstract presented at AMEE (An International Association for Medical Education) Conference Basel, Switzerland, August 24-28, 2024.

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)0.001
metaresearch head score (Gemma)0.009
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: Empirical · Consensus signal: Empirical
Teacher disagreement score0.008
Threshold uncertainty score0.015

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.009
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0000.001
Bibliometrics0.0020.000
Science and technology studies0.0000.000
Scholarly communication0.0010.001
Open science0.0000.001
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0020.000

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.

Opus teacher head0.122
GPT teacher head0.423
Teacher spread0.301 · 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
GenreEmpirical

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

Citations1
Published2024
Admission routes1
Has abstractyes

Explore more

Same venuemedRxiv→Same topicArtificial Intelligence in Healthcare and Education→French-language works237,207→