{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,19]],"date-time":"2026-10-19T09:40:41Z","timestamp":1792402841857,"version":"4.4.0"},"reference-count":23,"publisher":"Wiley","issue":"10","license":[{"start":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T00:00:00Z","timestamp":1789948800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T00:00:00Z","timestamp":1789948800000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["J Applied Clin Med Phys"],"published-print":{"date-parts":[[2026,10]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Breast Imaging Reporting and Data System (BI\u2010RADS) 4a nodules represent a diagnostic dilemma, with a malignancy rate ranging from 2% to 10%. The majority of these nodules prove benign after biopsy, leading to unnecessary invasive procedures, patient anxiety, and healthcare costs. Current clinical practice lacks a reliable, noninvasive tool to accurately distinguish benign from malignant BI\u2010RADS 4a lesions. Emerging evidence suggests that integrating multiparametric ultrasound data with clinical factors through artificial intelligence may improve diagnostic precision, but a validated multimodal model specifically for this ambiguous category is lacking.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>To develop and validate a multimodal artificial intelligence model that integrates ultrasound radiomics, hemodynamic features, and clinical data for the accurate differentiation of benign and malignant BI\u2010RADS 4a breast nodules, with the ultimate goal of supporting clinical decisionmaking and safely reducing unnecessary biopsies.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      This prospective, multicenter study enrolled 350 pathologically confirmed BI\u2010RADS 4a nodules from 331 patients in one tertiary hospital and two secondary hospitals. Clinical data, grayscale ultrasound images, and hemodynamic parameters were collected for each nodule. Radiomics features were extracted from the ultrasound images. A multimodal fusion model was constructed using the Light Gradient Boosting Machine (LightGBM) algorithm following feature selection via Least Absolute Shrinkage and Selection Operator (LASSO) regression. The model was developed and hyperparameter\u2010tuned on a training set (\n                      <jats:italic>n<\/jats:italic>\n                      \u00a0=\u00a0260) using nested cross\u2010validation and was evaluated on a temporally independent test set (\n                      <jats:italic>n<\/jats:italic>\n                      \u00a0=\u00a090). Model interpretability was enhanced using SHapley Additive exPlanations (SHAP).\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      On the independent test set, the multimodal LightGBM model demonstrated excellent diagnostic performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.95 (95% CI: 0.91\u20130.99), with a sensitivity of 100% (95% CI: 81.5%\u2013100%) and a specificity of 86.1% (95% CI: 75.8%\u201393.1%). Its performance was significantly superior to baseline classifiers including Logistic Regression, Support Vector Machine, and Random Forest (all\n                      <jats:italic>p<\/jats:italic>\n                      \u00a0&lt;\u00a00.05). SHAP analysis identified wavelet\u2010transformed texture features, the presence of a penetrating vessel, and BRCA mutation status as the top predictors of malignancy. Applying this model could have potentially avoided 32.4% (24\/74) of unnecessary biopsies on benign nodules while maintaining perfect sensitivity.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>A multimodal AI model based on LightGBM can accurately differentiate benign and malignant BI\u2010RADS 4a breast nodules. It shows significant potential as a clinical decision\u2010support tool to reduce unnecessary biopsies without compromising sensitivity, enabling more precise management for this challenging patient population.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/acm2.70800","type":"journal-article","created":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T04:48:31Z","timestamp":1790052511000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Enhancing diagnostic precision for BI\u2010RADS 4a breast nodules: A multimodal AI model integrating ultrasound radiomics, hemodynamic signatures, and clinical profiles"],"prefix":"10.1002","volume":"27","author":[{"given":"Jun","family":"Yang","sequence":"first","affiliation":[{"name":"Ultrasound Department Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Longman","family":"Long","sequence":"additional","affiliation":[{"name":"Ultrasound Department Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guang","family":"Li","sequence":"additional","affiliation":[{"name":"Department of Urology Surgery Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jian","family":"Zeng","sequence":"additional","affiliation":[{"name":"Ultrasound Department Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Caixiao","family":"Peng","sequence":"additional","affiliation":[{"name":"Ultrasound Department Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-1953-2277","authenticated-orcid":false,"given":"Jingting","family":"Yan","sequence":"additional","affiliation":[{"name":"Ultrasound Department Hengyang Central Hospital Affiliated to Hunan Normal University  Heyang Hunan China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,9,21]]},"reference":[{"key":"e_1_2_11_2_1","doi-asserted-by":"publisher","DOI":"10.1097\/MD.0000000000049036"},{"key":"e_1_2_11_3_1","doi-asserted-by":"publisher","DOI":"10.3390\/curroncol32010029"},{"key":"e_1_2_11_4_1","doi-asserted-by":"publisher","DOI":"10.1186\/s12880-025-01580-w"},{"issue":"6","key":"e_1_2_11_5_1","first-page":"1","article-title":"Application value of deep learning\u2010based AI model in the classification of breast nodules","volume":"86","author":"Shaogang Z","year":"2025","journal-title":"Br J Hosp Med (Lond)"},{"issue":"4","key":"e_1_2_11_6_1","first-page":"869","article-title":"Ultrasound radiomics\u2010based logistic regression model to differentiate between benign and malignant breast nodules","volume":"42","author":"Shanshan S","year":"2022","journal-title":"J Ultrasound Med"},{"key":"e_1_2_11_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.isci.2022.105692"},{"key":"e_1_2_11_8_1","doi-asserted-by":"publisher","DOI":"10.3389\/fonc.2022.1022441"},{"key":"e_1_2_11_9_1","doi-asserted-by":"publisher","DOI":"10.1097\/MD.0000000000047117"},{"key":"e_1_2_11_10_1","doi-asserted-by":"publisher","DOI":"10.3389\/fonc.2023.1192630"},{"key":"e_1_2_11_11_1","doi-asserted-by":"publisher","DOI":"10.26689\/par.v8i2.6405"},{"key":"e_1_2_11_12_1","doi-asserted-by":"publisher","DOI":"10.21037\/qims-22-1091"},{"key":"e_1_2_11_13_1","doi-asserted-by":"publisher","DOI":"10.46234\/ccdcw2025.056"},{"key":"e_1_2_11_14_1","doi-asserted-by":"publisher","DOI":"10.21037\/qims-2025-1092"},{"key":"e_1_2_11_15_1","doi-asserted-by":"publisher","DOI":"10.3390\/app13116764"},{"key":"e_1_2_11_16_1","doi-asserted-by":"publisher","DOI":"10.1186\/s12885-025-14760-2"},{"key":"e_1_2_11_17_1","doi-asserted-by":"publisher","DOI":"10.21037\/qims-23-283"},{"key":"e_1_2_11_18_1","doi-asserted-by":"publisher","DOI":"10.21037\/gs-24-457"},{"key":"e_1_2_11_19_1","doi-asserted-by":"publisher","DOI":"10.3389\/fpubh.2022.1098639"},{"key":"e_1_2_11_20_1","doi-asserted-by":"publisher","DOI":"10.11152\/mu-4510"},{"key":"e_1_2_11_21_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-024-71567-0"},{"issue":"12","key":"e_1_2_11_22_1","first-page":"1774","article-title":"Differential diagnosis model of benign and malignant breast BI\u2010RADS category 4 nodules based on serum SP70 and conventional laboratory indicators","volume":"56","author":"Ding HM","year":"2022","journal-title":"Zhonghua Yu Fang Yi Xue Za Zhi"},{"issue":"1","key":"e_1_2_11_23_1","article-title":"Maximizing diagnostic precision: evaluating the combined Yokohama and BI\u2010RADS scoring system for breast lesions","volume":"58","author":"Ahuja S","year":"2025","journal-title":"Rev Esp Patol"},{"key":"e_1_2_11_24_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2020.04.123"}],"container-title":["Journal of Applied Clinical Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/acm2.70800","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/acm2.70800","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/acm2.70800","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,19]],"date-time":"2026-10-19T08:53:11Z","timestamp":1792399991000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/acm2.70800"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,21]]},"references-count":23,"journal-issue":{"issue":"10","published-print":{"date-parts":[[2026,10]]}},"alternative-id":["10.1002\/acm2.70800"],"URL":"https:\/\/doi.org\/10.1002\/acm2.70800","archive":["Portico"],"relation":{},"ISSN":["1526-9914","1526-9914"],"issn-type":[{"value":"1526-9914","type":"print"},{"value":"1526-9914","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,21]]},"assertion":[{"value":"2026-04-21","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-08-21","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-21","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70800"}}