{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,18]],"date-time":"2026-11-18T02:09:46Z","timestamp":1794967786690,"version":"4.9.1"},"reference-count":26,"publisher":"Ovid Technologies (Wolters Kluwer Health)","issue":"12","license":[{"start":{"date-parts":[[2024,7,29]],"date-time":"2024-07-29T00:00:00Z","timestamp":1722211200000},"content-version":"unspecified","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["lww.com","ovid.com"],"crossmark-restriction":true},"short-container-title":["Invest Radiol"],"published-print":{"date-parts":[[2024,12]]},"abstract":"<jats:sec>\n            <jats:title>Objectives<\/jats:title>\n            <jats:p>Deep learning (DL) studies for the detection of clinically significant prostate cancer (csPCa) on magnetic resonance imaging (MRI) often overlook potentially relevant clinical parameters such as prostate-specific antigen, prostate volume, and age. This study explored the integration of clinical parameters and MRI-based DL to enhance diagnostic accuracy for csPCa on MRI.<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Materials and Methods<\/jats:title>\n            <jats:p>We retrospectively analyzed 932 biparametric prostate MRI examinations performed for suspected csPCa (ISUP \u22652) at 2 institutions. Each MRI scan was automatically analyzed by a previously developed DL model to detect and segment csPCa lesions. Three sets of features were extracted: DL lesion suspicion levels, clinical parameters (prostate-specific antigen, prostate volume, age), and MRI-based lesion volumes for all DL-detected lesions. Six multimodal artificial intelligence (AI) classifiers were trained for each combination of feature sets, employing both early (feature-level) and late (decision-level) information fusion methods. The diagnostic performance of each model was tested internally on 20% of center 1 data and externally on center 2 data (n = 529). Receiver operating characteristic comparisons determined the optimal feature combination and information fusion method and assessed the benefit of multimodal versus unimodal analysis. The optimal model performance was compared with a radiologist using PI-RADS.<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Results<\/jats:title>\n            <jats:p>Internally, the multimodal AI integrating DL suspicion levels with clinical features via early fusion achieved the highest performance. Externally, it surpassed baselines using clinical parameters (0.77 vs 0.67 area under the curve [AUC], <jats:italic toggle=\"yes\">P<\/jats:italic> &lt; 0.001) and DL suspicion levels alone (AUC: 0.77 vs 0.70, <jats:italic toggle=\"yes\">P<\/jats:italic> = 0.006). Early fusion outperformed late fusion in external data (0.77 vs 0.73 AUC, <jats:italic toggle=\"yes\">P<\/jats:italic> = 0.005). No significant performance gaps were observed between multimodal AI and radiologist assessments (internal: 0.87 vs 0.88 AUC; external: 0.77 vs 0.75 AUC, both <jats:italic toggle=\"yes\">P<\/jats:italic> &gt; 0.05).<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Conclusions<\/jats:title>\n            <jats:p>Multimodal AI (combining DL suspicion levels and clinical parameters) outperforms clinical and MRI-only AI for csPCa detection. Early information fusion enhanced AI robustness in our multicenter setting. Incorporating lesion volumes did not enhance diagnostic efficacy.<\/jats:p>\n          <\/jats:sec>","DOI":"10.1097\/rli.0000000000001102","type":"journal-article","created":{"date-parts":[[2024,7,29]],"date-time":"2024-07-29T18:00:05Z","timestamp":1722276005000},"page":"854-860","update-policy":"https:\/\/doi.org\/10.1097\/lww.0000000000001000","source":"Crossref","is-referenced-by-count":49,"title":["Multimodal AI Combining Clinical and Imaging Inputs Improves Prostate Cancer Detection"],"prefix":"10.1097","volume":"59","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3575-4465","authenticated-orcid":false,"given":"Christian","family":"Roest","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Derya","family":"Yakar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dorjan Ivan","family":"Rener Sitar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joeran S.","family":"Bosma","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dennis B.","family":"Rouw","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stefan Johannes","family":"Fransen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Henkjan","family":"Huisman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas C.","family":"Kwee","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"276","published-online":{"date-parts":[[2024,7,29]]},"reference":[{"key":"bib1-20250802","first-page":"229","article-title":"Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries","volume":"74","year":"2024","journal-title":"CA Cancer J Clin"},{"key":"bib2-20250802","doi-asserted-by":"crossref","first-page":"583","DOI":"10.1016\/j.juro.2017.03.131","article-title":"Prospective evaluation of PI-RADS\u2122 version 2 using the International Society of Urological Pathology Prostate Cancer Grade Group System","volume":"198","year":"2017","journal-title":"J Urol"},{"key":"bib3-20250802","doi-asserted-by":"crossref","first-page":"e87","DOI":"10.1097\/PAS.0000000000001497","article-title":"The 2019 International Society of Urological Pathology (ISUP) consensus conference on grading of prostatic carcinoma","volume":"44","year":"2020","journal-title":"Am J Surg Pathol"},{"key":"bib4-20250802","doi-asserted-by":"crossref","first-page":"474","DOI":"10.1002\/jmri.27595","article-title":"Performance of deep learning and genitourinary radiologists in detection of prostate cancer using 3-T multiparametric magnetic resonance imaging","volume":"54","year":"2021","journal-title":"J Magn Reson Imaging"},{"key":"bib5-20250802","doi-asserted-by":"crossref","first-page":"799","DOI":"10.1097\/RLI.0000000000000791","article-title":"Fully automatic deep learning in bi-institutional prostate magnetic resonance imaging: effects of cohort size and heterogeneity","volume":"56","year":"2021","journal-title":"Invest Radiol"},{"key":"bib6-20250802","doi-asserted-by":"crossref","first-page":"6867","DOI":"10.1007\/s00330-020-07030-1","article-title":"Using decision curve analysis to benchmark performance of a magnetic resonance imaging\u2013based deep learning model for prostate cancer risk assessment","volume":"30","year":"2020","journal-title":"Eur Radiol"},{"key":"bib7-20250802","doi-asserted-by":"crossref","first-page":"3318","DOI":"10.3390\/cancers13133318","article-title":"Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging: a systematic review and recommendations for future studies","volume":"13","year":"2021","journal-title":"Cancers (Basel)"},{"key":"bib8-20250802","doi-asserted-by":"crossref","first-page":"243","DOI":"10.1016\/j.eururo.2020.09.042","article-title":"EAU-EANM-ESTRO-ESUR-SIOG guidelines on prostate cancer-2020 update. Part 1: screening, diagnosis, and local treatment with curative intent","volume":"79","year":"2021","journal-title":"Eur Urol"},{"key":"bib9-20250802","doi-asserted-by":"crossref","first-page":"1773","DOI":"10.1038\/s41591-022-01981-2","article-title":"Multimodal biomedical AI","volume":"28","year":"2022","journal-title":"Nat Med"},{"key":"bib10-20250802","doi-asserted-by":"crossref","first-page":"1095","DOI":"10.1016\/j.ccell.2022.09.012","article-title":"Artificial intelligence for multimodal data integration in oncology","volume":"40","year":"2022","journal-title":"Cancer Cell"},{"key":"bib11-20250802","doi-asserted-by":"crossref","first-page":"1261","DOI":"10.1111\/exd.13777","article-title":"Multimodal skin lesion classification using deep learning","volume":"27","year":"2018","journal-title":"Exp Dermatol"},{"key":"bib12-20250802","first-page":"429","article-title":"Multimodal priors guided segmentation of liver lesions in MRI using mutual information based graph co-attention networks","year":"2020","journal-title":"Medical Image Computing and Computer Assisted Intervention\u2013MICCAI"},{"key":"bib13-20250802","doi-asserted-by":"crossref","first-page":"16","DOI":"10.1016\/j.eururo.2015.08.052","article-title":"PI-RADS prostate imaging\u2014reporting and data system: 2015, version 2","volume":"69","year":"2016","journal-title":"Eur Urol"},{"key":"bib14-20250802","doi-asserted-by":"crossref","first-page":"103","DOI":"10.1002\/jmri.27546","article-title":"Performance of prostate imaging reporting and data system version 2.1 for diagnosis of prostate cancer: a systematic review and meta-analysis","volume":"54","year":"2021","journal-title":"J Magn Reson Imaging"},{"key":"bib15-20250802","doi-asserted-by":"crossref","first-page":"e230031","DOI":"10.1148\/ryai.230031","article-title":"Semisupervised learning with report-guided pseudo labels for deep learning\u2013based prostate cancer detection using biparametric MRI","volume":"5","year":"2023","journal-title":"Radiol Artif Intell"},{"key":"bib16-20250802","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1038\/s41592-020-01008-z","article-title":"nnU-net: a self-configuring method for deep learning\u2013based biomedical image segmentation","volume":"18","year":"2021","journal-title":"Nat Methods"},{"key":"bib17-20250802","article-title":"Algorithms for hyper-parameter optimization","volume":"24","year":"2011","journal-title":"Adv Neural Inform Process Syst"},{"key":"bib19-20250802","doi-asserted-by":"crossref","first-page":"837","DOI":"10.2307\/2531595","article-title":"Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach","volume":"44","year":"1988","journal-title":"Biometrics"},{"key":"bib20-20250802","first-page":"65","article-title":"A simple sequentially rejective multiple test procedure","volume":"6","year":"1979","journal-title":"Scand J Stat"},{"issue":"3\u20134","key":"bib21-20250802","first-page":"353","article-title":"Notes on bias in estimation","volume":"43","year":"1956","journal-title":"Biometrika"},{"key":"bib22-20250802","first-page":"1490","article-title":"Comparative performance of deep learning and radiologists for the diagnosis and localization of clinically significant prostate cancer at MRI: a systematic review","volume":"12","year":"2022","journal-title":"Life (Basel)"},{"key":"bib23-20250802","doi-asserted-by":"crossref","first-page":"2224","DOI":"10.1007\/s00330-021-08320-y","article-title":"Deep learning\u2013assisted prostate cancer detection on bi-parametric MRI: minimum training data size requirements and effect of prior knowledge","volume":"32","year":"2022","journal-title":"Eur Radiol"},{"key":"bib24-20250802","doi-asserted-by":"crossref","first-page":"e230275","DOI":"10.1148\/radiol.230275","article-title":"Prediction variability to identify reduced AI performance in cancer diagnosis at MRI and CT","volume":"308","year":"2023","journal-title":"Radiology"},{"key":"bib25-20250802","doi-asserted-by":"crossref","first-page":"99","DOI":"10.1109\/TMI.2019.2920608","article-title":"Integrative analysis of pathological images and multi-dimensional genomic data for early-stage cancer prognosis","volume":"39","year":"2020","journal-title":"IEEE Trans Med Imaging"},{"key":"bib26-20250802","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1111\/bju.13649","article-title":"Lesion volume predicts prostate cancer risk and aggressiveness: validation of its value alone and matched with prostate imaging reporting and data system score","volume":"120","year":"2017","journal-title":"BJU Int"},{"key":"bib27-20250802","doi-asserted-by":"crossref","first-page":"12434","DOI":"10.1038\/s41598-021-91467-x","article-title":"Inter-vendor performance of deep learning in segmenting acute ischemic lesions on diffusion-weighted imaging: a multicenter study","volume":"11","year":"2021","journal-title":"Sci Rep"}],"container-title":["Investigative Radiology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.lww.com\/10.1097\/RLI.0000000000001102","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,2]],"date-time":"2025-08-02T19:06:08Z","timestamp":1754161568000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.lww.com\/10.1097\/RLI.0000000000001102"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,7,29]]},"references-count":26,"journal-issue":{"issue":"12","published-print":{"date-parts":[[2024]]}},"URL":"https:\/\/doi.org\/10.1097\/rli.0000000000001102","relation":{},"ISSN":["1536-0210","0020-9996"],"issn-type":[{"value":"1536-0210","type":"electronic"},{"value":"0020-9996","type":"print"}],"subject":[],"published":{"date-parts":[[2024,7,29]]}}}