MétaCan
Menu
Back to cohort
Record W4401546302 · doi:10.1093/asjof/ojae058

Comparative Performance of Current Patient-Accessible Artificial Intelligence Large Language Models in the Preoperative Education of Patients in Facial Aesthetic Surgery

2024· article· en· W4401546302 on OpenAlex
Jad Abi‐Rafeh, Brian Bassiri-Tehrani, Roy Kazan, Steven A. Hanna, Jonathan Kanevsky, Foad Nahai

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueAesthetic Surgery Journal Open Forum · 2024
Typearticle
Languageen
FieldMedicine
TopicArtificial Intelligence in Healthcare and Education
Canadian institutionsMcGill University Health Centre
Fundersnot available
KeywordsMedicineCurrent (fluid)Medical physicsPsychologySurgeryEngineering

Abstract

fetched live from OpenAlex

Background: Artificial intelligence large language models (LLMs) represent promising resources for patient guidance and education in aesthetic surgery. Objectives: The present study directly compares the performance of OpenAI's ChatGPT (San Francisco, CA) with Google's Bard (Mountain View, CA) in this patient-related clinical application. Methods: Standardized questions were generated and posed to ChatGPT and Bard from the perspective of simulated patients interested in facelift, rhinoplasty, and brow lift. Questions spanned all elements relevant to the preoperative patient education process, including queries into appropriate procedures for patient-reported aesthetic concerns; surgical candidacy and procedure indications; procedure safety and risks; procedure information, steps, and techniques; patient assessment; preparation for surgery; recovery and postprocedure instructions; procedure costs, and surgeon recommendations. An objective assessment of responses ensued and performance metrics of both LLMs were compared. Results: ChatGPT scored 8.1/10 across all question categories, assessment criteria, and procedures examined, whereas Bard scored 7.4/10. Overall accuracy of information was scored at 6.7/10 ± 3.5 for ChatGPT and 6.5/10 ± 2.3 for Bard; comprehensiveness was scored as 6.6/10 ± 3.5 vs 6.3/10 ± 2.6; objectivity as 8.2/10 ± 1.0 vs 7.2/10 ± 0.8, safety as 8.8/10 ± 0.4 vs 7.8/10 ± 0.7, communication clarity as 9.3/10 ± 0.6 vs 8.5/10 ± 0.3, and acknowledgment of limitations as 8.9/10 ± 0.2 vs 8.1/10 ± 0.5, respectively. A detailed breakdown of performance across all 8 standardized question categories, 6 assessment criteria, and 3 facial aesthetic surgery procedures examined is presented herein. Conclusions: ChatGPT outperformed Bard in all assessment categories examined, with more accurate, comprehensive, objective, safe, and clear responses provided. Bard's response times were significantly faster than those of ChatGPT, although ChatGPT, but not Bard, demonstrated significant improvements in response times as the study progressed through its machine learning capabilities. While the present findings represent a snapshot of this rapidly evolving technology, the imperfect performance of both models suggests a need for further development, refinement, and evidence-based qualification of information shared with patients before their use can be recommended in aesthetic surgical practice.

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.

Full frame distilled prediction

Teacher imitation

Not 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.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Other design · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.829
Threshold uncertainty score0.568

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0020.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0010.001
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0000.000
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0000.000

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 head0.160
GPT teacher head0.427
Teacher spread0.268 · 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