{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T16:38:59Z","timestamp":1790959139246,"version":"4.1.0"},"reference-count":37,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Form Res"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec sec-type=\"background\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Conventional nystagmus classification methods often rely on subjective observation by specialists, which is time-consuming and variable among clinicians. Recently, deep learning techniques have been used to automate nystagmus classification using convolutional and recurrent neural networks. These networks can accurately classify nystagmus patterns using video data. However, associated challenges including the need for large datasets when creating models, limited applicability to address specific image conditions, and the complexity associated with using these models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"objective\">\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>This study aimed to evaluate a novel approach for nystagmus classification that used the Generative Pre-trained Transformer 4 Vision (GPT-4V) model, which is a state-of-the-art large-scale language model with powerful image recognition capabilities.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"methods\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We developed a pupil-tracking process using a nystagmus-recording video and verified the optimization model\u2019s accuracy using GPT-4V classification and nystagmus recording. We tested whether the created optimization model could be evaluated in six categories of nystagmus: right horizontal, left horizontal, upward, downward, right torsional, and left torsional. The traced trajectory was input as two-dimensional coordinate data or an image, and multiple in-context learning methods were evaluated.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"results\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The developed model showed an overall classification accuracy of 37% when using pupil-traced images and a maximum accuracy of 24.6% when pupil coordinates were used as input. Regarding orientation, we achieved a maximum accuracy of 69% for the classification of horizontal nystagmus patterns but a lower accuracy for the vertical and torsional components.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"conclusions\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>We demonstrated the potential of versatile vertigo management in a generative artificial intelligence model that improves the accuracy and efficiency of nystagmus classification. We also highlighted areas for further improvement, such as expanding the dataset size and enhancing input modalities, to improve classification performance across all nystagmus types. The GPT-4V model validated only for recognizing still images can be linked to video classification and proposed as a novel method.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/70070","type":"journal-article","created":{"date-parts":[[2025,6,6]],"date-time":"2025-06-06T12:31:16Z","timestamp":1749213076000},"page":"e70070-e70070","source":"Crossref","is-referenced-by-count":3,"title":["Exploring Generative Pre-Trained Transformer-4-Vision for Nystagmus Classification: Development and Validation of a Pupil-Tracking Process"],"prefix":"10.2196","volume":"9","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6281-2782","authenticated-orcid":false,"given":"Masao","family":"Noda","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-0127-9137","authenticated-orcid":false,"given":"Ryota","family":"Koshu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8151-9071","authenticated-orcid":false,"given":"Reiko","family":"Tsunoda","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8335-2279","authenticated-orcid":false,"given":"Hirofumi","family":"Ogihara","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0490-8393","authenticated-orcid":false,"given":"Tomohiko","family":"Kamo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6863-4622","authenticated-orcid":false,"given":"Makoto","family":"Ito","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-4120-8641","authenticated-orcid":false,"given":"Hiroaki","family":"Fushiki","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2025,6,6]]},"reference":[{"issue":"4","key":"R1","doi-asserted-by":"publisher","first-page":"99","DOI":"10.1080\/2576117X.2018.1493311","article-title":"Dealing with nystagmus","volume":"68","author":"Ospina","journal-title":"J Binocul Vis Ocul Motil"},{"issue":"8","key":"R2","doi-asserted-by":"publisher","first-page":"798","DOI":"10.1001\/jama.2020.3911","article-title":"Nystagmus","volume":"325","author":"Kates","journal-title":"JAMA"},{"issue":"5","key":"R3","doi-asserted-by":"publisher","first-page":"378","DOI":"10.1097\/00055735-200110000-00010","article-title":"Nystagmus","volume":"12","author":"Gottlob","journal-title":"Curr Opin Ophthalmol"},{"issue":"10","key":"R4","doi-asserted-by":"publisher","first-page":"970","DOI":"10.1111\/j.1553-2712.2009.00523.x","article-title":"Disconnect between charted vestibular diagnoses and emergency department management decisions: a cross-sectional analysis from a nationally representative sample","volume":"16","author":"Newman-Toker","journal-title":"Acad Emerg Med"},{"issue":"S3","key":"R5","doi-asserted-by":"publisher","first-page":"S1","DOI":"10.1177\/0194599816689667","article-title":"Clinical practice guideline: Benign paroxysmal positional vertigo (update)","volume":"156","author":"Bhattacharyya","journal-title":"Otolaryngol--head neck surg"},{"issue":"6","key":"R6","doi-asserted-by":"publisher","first-page":"507","DOI":"10.1007\/s00405-004-0841-x","article-title":"Long-term outcome and health-related quality of life in benign paroxysmal positional vertigo","volume":"262","author":"Lopez-Escamez","journal-title":"Eur Arch Otorhinolaryngol"},{"key":"R7","doi-asserted-by":"crossref","unstructured":"Santini T Fuhl W Kasneci E . PuReST: robust pupil tracking for real-time pervasive eye tracking. Presented at: ETRA \u201918: 2018 Symposium on Eye Tracking Research and Applications. Jun 14-17, 2018:ACM. 1-5; Warsaw Poland. [doi: 10.1145\/3204493.3204578]","DOI":"10.1145\/3204493.3204578"},{"key":"#cr-split#-R8.1","doi-asserted-by":"crossref","unstructured":"Eivazi S Santini T Keshavarzi A K\u00fcbler TC Mazzei A . Improving real-time CNN-based pupil detection through domain-specific data augmentation. Presented at: ETRA '19. Jun 25-28, 2019:ACM. 1-6","DOI":"10.1145\/3314111.3319914"},{"key":"#cr-split#-R8.2","doi-asserted-by":"crossref","unstructured":"Denver Colorado. Jun 25, 2019.[doi: 10.1145\/3314111.3319914]","DOI":"10.1145\/3314111.3319914"},{"issue":"14","key":"R9","doi-asserted-by":"publisher","DOI":"10.1167\/15.14.11","article-title":"Knowing what the brain is seeing in three dimensions: A novel, noninvasive, sensitive, accurate, and low-noise technique for measuring ocular torsion","volume":"15","author":"Otero-Millan","journal-title":"J Vis"},{"key":"R10","doi-asserted-by":"publisher","DOI":"10.1016\/j.jneumeth.2019.05.016","article-title":"DeepVOG: Open-source pupil segmentation and gaze estimation in neuroscience using deep learning","volume":"324","author":"Yiu","journal-title":"J Neurosci Methods"},{"key":"R11","doi-asserted-by":"publisher","DOI":"10.1016\/j.bspc.2021.102616","article-title":"Deep learning based torsional nystagmus detection for dizziness and vertigo diagnosis","volume":"68","author":"Zhang","journal-title":"Biomed Signal Process Control"},{"issue":"5","key":"R12","doi-asserted-by":"publisher","first-page":"1122","DOI":"10.1016\/j.cell.2018.02.010","article-title":"Identifying medical diagnoses and treatable diseases by image-based deep learning","volume":"172","author":"Kermany","journal-title":"Cell"},{"key":"R13","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2024.108189","article-title":"A comprehensive evaluation of large language models on benchmark biomedical text processing tasks","volume":"171","author":"Jahan","journal-title":"Comput Biol Med"},{"key":"R14","unstructured":"GPT-4V(ision) system card. OpenAI. 2023. URL: https:\/\/openai.com\/research\/gpt-4v-system-card [Accessed 25-09-2023]"},{"key":"R15","doi-asserted-by":"publisher","DOI":"10.2196\/57054","article-title":"Performance of GPT-4V in answering the Japanese Otolaryngology Board Certification Examination questions: evaluation study","volume":"10","author":"Noda","journal-title":"JMIR Med Educ"},{"key":"R16","unstructured":"Harsha N Yin TL Sheng Z Dean C Richard E Nicolo F et al. Can generalist foundation models outcompete special-purpose tuning? Case study in medicine. arXiv. Preprint posted online on Nov 28, 2023. [doi: 10.48550\/arXiv.2311.16452]"},{"key":"R17","unstructured":"Nipun MS Sulaiman RB Kareem A . Efficiency comparison of AI classification algorithms for image detection and recognition in real-time. arXiv. Preprint posted online on Jun 12, 2022. [doi: 10.48550\/arXiv.2206.05842]"},{"key":"R18","unstructured":"Tian Y Kanade T Cohn JF . Dual-state parametric eye tracking. Presented at: Proceedings Fourth IEEE International Conference on Automatic Face and Gesture Recognition; Mar 28-30, 2000; Grenoble, France. URL: https:\/\/www.cs.cmu.edu\/~face\/Papers\/fg1camera.pdf [Accessed 26-04-2025]"},{"issue":"4","key":"R19","doi-asserted-by":"publisher","first-page":"466","DOI":"10.5281\/zenodo.1058133","article-title":"Evaluation of Haar cascade classifiers designed for face detection","volume":"6","author":"Padilla","journal-title":"World Acad Sci Eng Technol"},{"key":"R20","volume":"21","author":"Wilson","journal-title":"J Comput Sci Coll"},{"issue":"1","key":"R21","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pdig.0000433","article-title":"Performance of generative pretrained transformer on the National Medical Licensing Examination in Japan","volume":"3","author":"Tanaka","journal-title":"PLOS Digit Health"},{"issue":"11","key":"R22","doi-asserted-by":"publisher","first-page":"1217","DOI":"10.3950\/jibiinkotokeibu.126.11_1217","article-title":"A study of the performance of the generative pretrained transformer in the Japanese Otorhinolaryngology Specialty examination","volume":"126","author":"Noda","journal-title":"Nippon Jibiinkoka Tokeibugeka Gakkai Kaiho(Tokyo)"},{"key":"R23","unstructured":"Bsharat SM Myrzakhan A Shen Z . Principled instructions are all you need for questioning LLaMA-1\/2, GPT-3.5\/4. arXiv. Preprint posted online on Jan 18, 2024. [doi: 10.48550\/arXiv.2312.16171]"},{"key":"R24","unstructured":"Wei J Wang X Schuurmans D Bosma M Ichter B Xia F et al. Chain-of-thought prompting elicits reasoning in large language models. arXiv. Preprint posted online on Jan 10, 2023. [doi: 10.48550\/arXiv.2201.11903]"},{"key":"R25","unstructured":"Touvron H Lavril T Izacard G Martinet X Lachaux MA Lacroix T et al. LLaMA: open and efficient foundation language models. arXiv. Preprint posted online on Feb 27, 2023. [doi: 10.48550\/arXiv.2302.13971]"},{"key":"R26","unstructured":"Brown TB Mann B Ryder N Subbiah M Kaplan J Dhariwal P et al. Language models are few-shot learners. arXiv. Preprint posted online on Jul 22, 2020. [doi: 10.48550\/arXiv.2005.14165]"},{"key":"R27","doi-asserted-by":"publisher","DOI":"10.3389\/fneur.2022.963968","article-title":"aEYE: A deep learning system for video nystagmus detection","volume":"13","author":"Wagle","journal-title":"Front Neurol"},{"issue":"5","key":"R28","doi-asserted-by":"publisher","DOI":"10.3390\/jcm8050633","article-title":"Developing a diagnostic decision support system for benign paroxysmal positional vertigo using a deep-learning model","volume":"8","author":"Lim","journal-title":"J Clin Med"},{"issue":"1","key":"R29","doi-asserted-by":"publisher","first-page":"11975","DOI":"10.1038\/s41598-023-39104-7","article-title":"A nystagmus extraction system using artificial intelligence for video-nystagmography","volume":"13","author":"Lee","journal-title":"Sci Rep"},{"issue":"10","key":"R30","doi-asserted-by":"publisher","first-page":"990","DOI":"10.1016\/j.jacr.2023.05.003","article-title":"Evaluating GPT as an adjunct for radiologic decision making: GPT-4 versus GPT-3.5 in a breast imaging pilot","volume":"20","author":"Rao","journal-title":"J Am Coll Radiol"},{"issue":"8","key":"R31","doi-asserted-by":"publisher","first-page":"2589","DOI":"10.1007\/s11255-024-04009-5","article-title":"Adequacy of prostate cancer prevention and screening recommendations provided by an artificial intelligence-powered large language model","volume":"56","author":"Chiarelli","journal-title":"Int Urol Nephrol"},{"key":"R32","unstructured":"Jared K Sam M Tom H Tom BB Benjamin C Rewon C et al. Scaling laws for neural language models. arXiv. Preprint posted online on Jan 23, 2020. URL: https:\/\/arxiv.org\/pdf\/2001.08361 [Accessed 26-04-2025]"},{"issue":"16","key":"R33","doi-asserted-by":"publisher","DOI":"10.3390\/s19163630","article-title":"Development of an eye tracking-based human-computer interface for real-time applications","volume":"19","author":"Bozomitu","journal-title":"Sensors (Basel)"},{"key":"R34","doi-asserted-by":"publisher","DOI":"10.1016\/j.imavis.2018.04.002","article-title":"Unobtrusive and pervasive video-based eye-gaze tracking","volume":"74","author":"Cristina","journal-title":"Image Vis Comput"},{"key":"R35","doi-asserted-by":"crossref","unstructured":"Williams CYK Bains J Tang T et al. Evaluating large language models for drafting emergency department discharge summaries. medRxiv. Apr 4, 2024:2024.04.03.24305088. [doi: 10.1101\/2024.04.03.24305088] [Medline: 38633805]","DOI":"10.1101\/2024.04.03.24305088"},{"key":"R36","doi-asserted-by":"publisher","DOI":"10.2196\/43068","article-title":"A medical ethics framework for conversational artificial intelligence","volume":"25","author":"Fournier-Tombs","journal-title":"J Med Internet Res"}],"container-title":["JMIR Formative Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/formative.jmir.org\/2025\/1\/e70070","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:58:27Z","timestamp":1782842307000},"score":1,"resource":{"primary":{"URL":"https:\/\/formative.jmir.org\/2025\/1\/e70070"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,6]]},"references-count":37,"URL":"https:\/\/doi.org\/10.2196\/70070","relation":{},"ISSN":["2561-326X"],"issn-type":[{"value":"2561-326X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6,6]]},"article-number":"v9i4e70070"}}