大语言模型在本科物理治疗课程中的学业表现:子集分析
Academic performance of large language models across an undergraduate physiotherapy curriculum: a subset analysis
研究评估了GPT-5.0和Claude 4.5 Sonnet在34门本科物理治疗课程考试中的表现。两款模型均达到及格标准,Claude 4.5 Sonnet的平均总分为87.3分(GPA 3.39/4.00),略高于GPT-5.0的85.7分(GPA 3.28/4.00)。模型在基础科学课程上的表现优于临床导向课程,显示其在知识型评估中表现良好,但在临床整合领域存在相对局限性。
- Article
- Open access
- Published: 20 October 2026
Scientific Reports (2026) Cite this article
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
Abstract
Large language models (LLMs) are increasingly used in physiotherapy for information retrieval, educational support, and assistance in assessment and rehabilitation-related decision-making. Although their performance has often been evaluated using isolated examinations or domain-specific question sets, less is known about how these models perform across the breadth of an undergraduate physiotherapy curriculum. This study examined LLM performance using a curriculum-based academic assessment framework reflecting how theoretical knowledge is formally evaluated in physiotherapy education. Written exams from a subset of 34 of 54 eligible courses within an undergraduate Physiotherapy and Rehabilitation program were analyzed. Two LLMs, GPT-5.0 and Claude 4.5 Sonnet, were queried using a standardized zero-shot prompting approach, and responses were graded according to official answer keys and institutional criteria. Course-level examination scores, letter grades, and cumulative grade point averages were calculated for standardized comparison. Under the institutional grading system, both models achieved passing-level performance, with GPT-5.0 reaching a mean total score of 85.7 ± 10.4 (GPA-equivalent: 3.28/4.00) and Claude 4.5 Sonnet a mean total score of 87.3 ± 11.2 (GPA-equivalent: 3.39/4.00), both satisfying the institutional graduation threshold; Claude 4.5 Sonnet showed modestly but significantly higher overall performance. Both models performed better in basic science courses than in clinically oriented courses. These findings suggest that LLMs perform well on knowledge-based assessments in undergraduate physiotherapy education, while revealing relative limitations in clinically integrated domains.
Abbreviations
- LLM:
-
Large language model
- GPA:
-
Grade point average
- AI:
-
Artificial intelligence
- OSF:
-
Open Science Framework
Acknowledgements
The authors thank the course instructors who granted permission to access examination materials used in this study.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Department of Physiotherapy and Rehabilitation, Faculty of Health Sciences, Acibadem Mehmet Ali Aydinlar University, Istanbul, Turkey
Ali Ömer Acar, Sergen Öztürk, Dilek Çağrı Arslan & Nuray Alaca
Department of Cardiopulmonary Physiotherapy and Rehabilitation, Institute of Health Sciences, Bezmialem Vakif University, Istanbul, Türkiye
Dilek Çağrı Arslan
Department of Physiotherapy and Rehabilitation, Graduate School of Health Sciences, Acibadem Mehmet Ali Aydinlar University, Istanbul, Türkiye
Nuray Alaca
Authors
- Ali Ömer Acar
- Sergen Öztürk
- Dilek Çağrı Arslan
- Nuray Alaca
Corresponding author
Correspondence to Ali Ömer Acar.
Ethics declarations
Competing interests
The authors declare that they have no competing interests.
Ethics approval and consent to participate
This study did not involve human participants or identifiable personal data. Examination materials were fully anonymized prior to analysis, and ethical approval was not required in accordance with institutional and national guidelines. Permission to access and analyze examination materials was obtained from each responsible course instructor prior to data collection; materials were treated as proprietary academic content and used solely for the purpose of this study.
Declaration of generative AI use in manuscript preparation
The authors used ChatGPT (OpenAI) for language editing and formatting assistance during the preparation of this manuscript. All scientific content, data analysis, interpretation of findings, and conclusions are the sole responsibility of the authors. No generative AI tools were used to generate primary research content, statistical analyses, or substantive scientific text. The authors reviewed and verified all AI-assisted output and take full responsibility for the content of the manuscript.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Acar, A.Ö., Öztürk, S., Çağrı Arslan, D. et al. Academic performance of large language models across an undergraduate physiotherapy curriculum: a subset analysis. Sci Rep (2026). https://doi.org/10.1038/s41598-026-68822-x
Received: 29 January 2026
Accepted: 24 August 2026
Published: 20 October 2026
DOI: https://doi.org/10.1038/s41598-026-68822-x
Keywords
来源:Nature 健康护理主题 · nature.com