{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,6]],"date-time":"2026-07-06T17:14:34Z","timestamp":1783358074150,"version":"3.54.6"},"reference-count":46,"publisher":"SAGE Publications","license":[{"start":{"date-parts":[[2026,7,3]],"date-time":"2026-07-03T00:00:00Z","timestamp":1783036800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"funder":[{"DOI":"10.13039\/501100001691","name":"Japan Society for the Promotion of Science","doi-asserted-by":"crossref","award":["25K02910"],"award-info":[{"award-number":["25K02910"]}],"id":[{"id":"10.13039\/501100001691","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100002241","name":"Japan Science and Technology Agency","doi-asserted-by":"crossref","award":["JPMJBY24E"],"award-info":[{"award-number":["JPMJBY24E"]}],"id":[{"id":"10.13039\/501100002241","id-type":"DOI","asserted-by":"crossref"}]},{"name":"the Yoshiko Shinohara Memorial Foundation,","award":["2026"],"award-info":[{"award-number":["2026"]}]}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["Advances in Wound Care"],"abstract":"<jats:sec>\n                    <jats:title>Objective:<\/jats:title>\n                    <jats:p>To benchmark zero-shot generative pre-trained transformer (GPT)-based multimodal large language models (MLLMs) for pressure injury (PI) staging from photographs and quantify the effects of prompt strategy, structured outputs, and clinically meaningful label granularity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Approach:<\/jats:title>\n                    <jats:p>We performed a retrospective observational benchmark using 1,091 public, de-identified PI photographs labeled Stage I\u2013IV. In the standardized analysis, all 10 model\/prompt conditions were evaluated using a standardized, resume-safe application programming interface pipeline with per-image logging. We evaluated exact four-class staging, three-class staging (I\/II\/III\u2013IV), skin-break screening (I vs. II\u2013IV), and an advanced-intervention threshold (I\u2013II vs. III\u2013IV). Reporting followed Strengthening the Reporting of Observational Studies in Epidemiology and artificial intelligence\/machine learning guidance; metrics included accuracy, Wilson 95% confidence intervals, F1 scores, weighted kappa, and threshold sensitivity\/specificity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results:<\/jats:title>\n                    <jats:p>All conditions yielded parsed predictions for all 1,091 images. The best accuracy was 93.77% for skin-break screening (GPT-5.2 structured JavaScript Object Notation [JSON]), 90.28% for the advanced-intervention threshold (GPT-5.1 full prompt, low reasoning), 83.78% for three-class staging (GPT-5.1 full prompt, low reasoning), and 65.08% for exact four-class staging (GPT-5.2 structured JSON). The best advanced-intervention condition achieved sensitivity 96.72%, specificity 83.79%, and negative predictive value 96.19%. Stage IV undercalling remained safety-relevant; GPT-5.2-pro had Stage IV recall 8.79% and false-negative rate 91.21%.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Innovation:<\/jats:title>\n                    <jats:p>This prompt-transparent benchmark shows how outcome granularity, prompting, and structured outputs affect GPT-based PI staging and adds ordinal, threshold-based, and safety metrics.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion:<\/jats:title>\n                    <jats:p>GPT-based MLLMs may support clinician-supervised triage and prioritization, but image-only autonomous exact staging is not clinically ready.<\/jats:p>\n                    <jats:fig orientation=\"portrait\" position=\"float\">\n                      <jats:caption>\n                        <jats:p>Toshiaki Takahashi, RN, PhD<\/jats:p>\n                      <\/jats:caption>\n                      <jats:graphic xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" orientation=\"portrait\" position=\"float\" xlink:href=\"10.1177_21621918261465916-fig6.tif\"\/>\n                    <\/jats:fig>\n                  <\/jats:sec>","DOI":"10.1177\/21621918261465916","type":"journal-article","created":{"date-parts":[[2026,7,4]],"date-time":"2026-07-04T01:51:52Z","timestamp":1783129912000},"update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking Zero-Shot Generative Pre-Trained Transformer-Based Multimodal Large Language Models for Pressure Injury Staging"],"prefix":"10.1177","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5724-4784","authenticated-orcid":false,"given":"Toshiaki","family":"Takahashi","sequence":"first","affiliation":[{"name":"Department of Adult Nursing, Graduate School of Medicine, Yokohama City University, Yokohama, Kanagawa, Japan."},{"name":"Japan Institute for Health Security, Bureau of System Infrastructure Development, Shinjuku, Tokyo, Japan."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kengo","family":"Miyo","sequence":"additional","affiliation":[{"name":"Japan Institute for Health Security, Bureau of System Infrastructure Development, Shinjuku, Tokyo, Japan."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nao","family":"Tamai","sequence":"additional","affiliation":[{"name":"Department of Adult Nursing, Graduate School of Medicine, Yokohama City University, Yokohama, Kanagawa, Japan."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2026,7,3]]},"reference":[{"key":"e_1_3_3_2_1","doi-asserted-by":"publisher","DOI":"10.1097\/WON.0000000000000281"},{"key":"e_1_3_3_3_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijnurstu.2020.103546"},{"key":"e_1_3_3_4_1","doi-asserted-by":"publisher","DOI":"10.12968\/jowc.2019.28.11.702"},{"key":"e_1_3_3_5_1","doi-asserted-by":"publisher","DOI":"10.1111\/iwj.13071"},{"key":"e_1_3_3_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.amjsurg.2009.12.021"},{"key":"e_1_3_3_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijnurstu.2015.06.006"},{"key":"e_1_3_3_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijnurstu.2012.11.019"},{"key":"e_1_3_3_9_1","doi-asserted-by":"publisher","DOI":"10.3389\/fpubh.2022.964680"},{"key":"e_1_3_3_10_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.1365-2702.2004.00974.x"},{"key":"e_1_3_3_11_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.1365-2648.2007.04474.x"},{"key":"e_1_3_3_12_1","doi-asserted-by":"publisher","DOI":"10.1002\/sim.2877"},{"key":"e_1_3_3_13_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.1365-2702.2008.02569.x"},{"key":"e_1_3_3_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijnurstu.2009.03.014"},{"key":"e_1_3_3_15_1","doi-asserted-by":"publisher","DOI":"10.1097\/NCQ.0b013e3182169452"},{"key":"e_1_3_3_16_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.1365-2648.2006.03801.x"},{"key":"e_1_3_3_17_1","doi-asserted-by":"publisher","DOI":"10.1017\/S026646239915414X"},{"key":"e_1_3_3_18_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2019.101742"},{"key":"e_1_3_3_19_1","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2021.7234"},{"key":"e_1_3_3_20_1","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.2020200029"},{"key":"e_1_3_3_21_1","doi-asserted-by":"publisher","DOI":"10.7326\/0003-4819-147-8-200710160-00010"},{"key":"e_1_3_3_22_1","doi-asserted-by":"publisher","DOI":"10.2196\/36977"},{"key":"e_1_3_3_23_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-020-78799-w"},{"key":"e_1_3_3_24_1","doi-asserted-by":"publisher","DOI":"10.1089\/wound.2021.0091"},{"key":"e_1_3_3_25_1","doi-asserted-by":"publisher","DOI":"10.3389\/fmedt.2022.905074"},{"key":"e_1_3_3_26_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0264139"},{"key":"e_1_3_3_27_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jfma.2022.04.010"},{"key":"e_1_3_3_28_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-022-26812-9"},{"key":"e_1_3_3_29_1","doi-asserted-by":"publisher","DOI":"10.1111\/jan.15584"},{"key":"e_1_3_3_30_1","doi-asserted-by":"publisher","DOI":"10.1155\/2023\/3858997"},{"key":"e_1_3_3_31_1","doi-asserted-by":"publisher","DOI":"10.1089\/tmj.2023.0715"},{"key":"e_1_3_3_32_1","doi-asserted-by":"publisher","DOI":"10.12968\/jowc.2024.33.5.368"},{"key":"e_1_3_3_33_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jtv.2024.05.009"},{"key":"e_1_3_3_34_1","doi-asserted-by":"publisher","DOI":"10.1111\/wrr.13189"},{"key":"e_1_3_3_35_1","doi-asserted-by":"publisher","DOI":"10.1097\/ASW.0000000000000239"},{"key":"e_1_3_3_36_1","doi-asserted-by":"publisher","DOI":"10.1111\/jan.16905"},{"key":"e_1_3_3_37_1","doi-asserted-by":"publisher","DOI":"10.2196\/67356"},{"key":"e_1_3_3_38_1","doi-asserted-by":"publisher","DOI":"10.1097\/ASW.0000000000000352"},{"key":"e_1_3_3_39_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00521-022-07274-6"},{"key":"e_1_3_3_40_1","doi-asserted-by":"publisher","DOI":"10.1177\/21621918251380217"},{"key":"e_1_3_3_41_1","doi-asserted-by":"publisher","DOI":"10.2196\/62774"},{"key":"e_1_3_3_42_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jtv.2025.100894"},{"key":"e_1_3_3_43_1","doi-asserted-by":"publisher","DOI":"10.2196\/25704"},{"key":"e_1_3_3_44_1","unstructured":"43.Swift Medical. AI-Powered Advanced Wound Care Management. Available from: https:\/\/swiftmedical.com\/ [Last accessed: May26 2026]."},{"key":"e_1_3_3_45_1","unstructured":"44.Net Health. AI Wound Imaging & Wound Care Analytics. Tissue Analytics. Available from: https:\/\/www.nethealth.com\/tissue-analytics\/ [Last accessed: May26 2026]."},{"key":"e_1_3_3_46_1","unstructured":"45.eKare. Digital Wound Care. Wound Imaging Measurement Telehealth. Available from: https:\/\/ekare.ai\/ [Last accessed: May26 2026]."},{"key":"e_1_3_3_47_1","unstructured":"46.WoundSource. The WoundVision Scout. Available from: https:\/\/www.woundsource.com\/product\/woundvision-scout [Last accessed: May26 2026]."}],"container-title":["Advances in Wound Care"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/21621918261465916","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/21621918261465916","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/21621918261465916","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,6]],"date-time":"2026-07-06T16:17:24Z","timestamp":1783354644000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/21621918261465916"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,3]]},"references-count":46,"alternative-id":["10.1177\/21621918261465916"],"URL":"https:\/\/doi.org\/10.1177\/21621918261465916","relation":{},"ISSN":["2162-1918","2162-1934"],"issn-type":[{"value":"2162-1918","type":"print"},{"value":"2162-1934","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,3]]},"article-number":"21621918261465916"}}