{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T10:32:00Z","timestamp":1785407520537,"version":"3.56.0"},"reference-count":31,"publisher":"MDPI AG","issue":"11","license":[{"start":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T00:00:00Z","timestamp":1780358400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"China Medical University Hsinchu Hospital","award":["CMUHCH-DMR-113-001"],"award-info":[{"award-number":["CMUHCH-DMR-113-001"]}]},{"name":"China Medical University Hsinchu Hospital","award":["CMUHCH-DMR-112-024"],"award-info":[{"award-number":["CMUHCH-DMR-112-024"]}]},{"name":"China Medical University Hsinchu Hospital","award":["CMUHCH-DMR-112-027"],"award-info":[{"award-number":["CMUHCH-DMR-112-027"]}]},{"name":"Taiwan National Science and Technology Council","award":["112-2314-B-039-059MY3"],"award-info":[{"award-number":["112-2314-B-039-059MY3"]}]},{"award":["112-2314-B-039-059MY3"],"award-info":[{"award-number":["112-2314-B-039-059MY3"]}],"id":[{"id":"https:\/\/ror.org\/02kv4zf79","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Diagnostics"],"abstract":"<jats:p>Background\/Objectives: To evaluate the effect of detector architecture and dataset characteristics on intracranial hemorrhage (ICH) subtype localization on noncontrast head CT, with emphasis on bidirectional cross-dataset generalization. Methods: This retrospective study analyzed two publicly available datasets: the Brain Hemorrhage Extended (BHX) dataset and the RSNA 2019+ dataset. Models were trained and internally validated on one dataset and externally tested on the other dataset in both directions: BHX-to-RSNA+ and RSNA+-to-BHX. Six representative deep learning detectors, including CNN-based one-stage and two-stage detectors and a Swin Transformer-based RT-DETR (Swin-RT-DETR) variant, were evaluated. Localization performance was assessed using mean average precision at a bounding-box intersection-over-union threshold of 0.5 (mAP@50), bounding-box Dice similarity coefficient (BB-DSC), and bounding-box intersection-over-union (BB-IoU). Image-level and patient-level analyses were performed, with Bonferroni correction applied for statistical comparisons. Dataset characterization analyses were performed to compare subtype prevalence, bounding-box geometry, lesion burden, annotation density, and spatial distribution. Results: Under internal validation, Swin-RT-DETR achieved competitive or superior performance across several ICH subtypes, but its advantage was subtype-dependent rather than uniform. Faster R-CNN with a ResNeXt101 backbone achieved comparable IVH performance and higher IPH BB-DSC and BB-IoU, whereas Swin-RT-DETR performed better for SAH, SDH, and EDH. External validation showed substantial performance degradation across architectures, subtypes, and validation directions. Absolute BB-DSC reductions for Swin-RT-DETR ranged from approximately 0.54\u20130.79 in the BHX-to-RSNA+ direction and 0.17\u20130.74 in the RSNA+-to-BHX direction. Similar degradation patterns were observed at the patient level. Statistical comparisons showed fewer significant model-level differences under external validation, suggesting attenuation of architecture-specific advantages under domain shift. Dataset characterization analysis demonstrated differences in subtype distribution, bounding-box geometry, lesion burden, annotation density, and spatial localization patterns between BHX and RSNA+. Conclusions: ICH subtype localization performance is strongly influenced by dataset characteristics, annotation heterogeneity, and domain shift. Although Transformer-based hierarchical feature extraction showed subtype-dependent advantages under internal validation, these advantages diminished under bidirectional external validation. These findings highlight the need for dataset characterization, external validation, patient-level evaluation, and task-specific clinical benchmarks before automated ICH localization models can be considered for real-world clinical integration.<\/jats:p>","DOI":"10.3390\/diagnostics16111705","type":"journal-article","created":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T10:14:57Z","timestamp":1780395297000},"page":"1705","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Cross-Dataset Generalization of Deep Learning-Based Detectors for Intracranial Hemorrhage Subtype Localization on Noncontrast Head CT: A Comparative Study"],"prefix":"10.3390","volume":"16","author":[{"given":"Chiao-Hua","family":"Lee","sequence":"first","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hikam","family":"Muzakky","sequence":"additional","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cheng-En","family":"Juan","sequence":"additional","affiliation":[{"name":"China Medical University Hospital, Taichung 404, Taiwan"},{"name":"Master\u2019s Program of Biomedical Informatics and Biomedical Engineering, Feng Chia University, Taichung 407, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chia-Ching","family":"Chang","sequence":"additional","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"},{"name":"Department of Management Science, National Yang Ming Chiao Tung University, Hsinchu 300, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ya-Hui","family":"Li","sequence":"additional","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"},{"name":"Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei 106, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tung-Yang","family":"Lee","sequence":"additional","affiliation":[{"name":"Show Chwan Memorial Hospital, Changhua 500, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cheng-Hsuan","family":"Juan","sequence":"additional","affiliation":[{"name":"Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei 106, Taiwan"},{"name":"Department of Medical Imaging, China Medical University Hospital, Taichung 404, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1495-5353","authenticated-orcid":false,"given":"Ming-Ting","family":"Tsai","sequence":"additional","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5219-0406","authenticated-orcid":false,"given":"Chun-Jung","family":"Juan","sequence":"additional","affiliation":[{"name":"Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan"},{"name":"Department of Medical Imaging, China Medical University Hospital, Taichung 404, Taiwan"},{"name":"Department of Radiology, School of Medicine, College of Medicine, China Medical University, Taichung 406, Taiwan"},{"name":"Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu 300, Taiwan"},{"name":"Department of Computer Science and Information Engineering, National Taiwan University, Taipei 106, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,6,2]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"e282","DOI":"10.1161\/STR.0000000000000407","article-title":"2022 Guideline for the Management of Patients With Spontaneous Intracerebral Hemorrhage: A Guideline From the American Heart Association\/American Stroke Association","volume":"53","author":"Greenberg","year":"2022","journal-title":"Stroke"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"16","DOI":"10.1002\/ana.26519","article-title":"A Causal Classification System for Intracerebral Hemorrhage Subtypes","volume":"93","author":"Raposo","year":"2023","journal-title":"Ann. Neurol."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"e230159","DOI":"10.1148\/ryai.230159","article-title":"Examination-Level Supervision for Deep Learning-based Intracranial Hemorrhage Detection on Head CT Scans","volume":"6","author":"Teneggi","year":"2024","journal-title":"Radiol. Artif. Intell."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"4611","DOI":"10.1038\/s41598-025-34596-x","article-title":"A multi-task deep learning pipeline for classification, detection, and weakly supervised 3D segmentation of intraparenchymal hematoma on brain CT","volume":"16","author":"Juan","year":"2026","journal-title":"Sci. Rep."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"e240032","DOI":"10.1148\/ryai.240032","article-title":"Applying Conformal Prediction to a Deep Learning Model for Intracranial Hemorrhage Detection to Improve Trustworthiness","volume":"7","author":"Gamble","year":"2025","journal-title":"Radiol. Artif. Intell."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"9","DOI":"10.1038\/s41746-017-0015-z","article-title":"Advanced machine learning in action: Identification of intracranial hemorrhage on computed tomography scans of the head with clinical workflow integration","volume":"1","author":"Arbabshirani","year":"2018","journal-title":"npj Digit. Med."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"e230296","DOI":"10.1148\/ryai.230296","article-title":"Precise Image-level Localization of Intracranial Hemorrhage on Head CT Scans with Deep Learning Models Trained on Study-level Labels","volume":"6","author":"Wu","year":"2024","journal-title":"Radiol. Artif. Intell."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Nijiati, M., Tuersun, A., Zhang, Y., Yuan, Q., Gong, P., Abulizi, A., Tuoheti, A., Abulaiti, A., and Zou, X. (2022). A symmetric prior knowledge based deep learning model for intracerebral hemorrhage lesion segmentation. Front. Physiol., 13.","DOI":"10.3389\/fphys.2022.977427"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"103085","DOI":"10.1016\/j.bspc.2021.103085","article-title":"Detecting hemorrhage types and bounding box of hemorrhage by deep learning","volume":"71","year":"2022","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Ferdi, A., Benierbah, S., and Ferdi, Y. (2023, January 23\u201325). YOLOv3-based Intracranial Hemorrhage Localization from CT Images. Proceedings of the 13th International Symposium on Advanced Topics in Electrical Engineering (ATEE), Bucharest, Romania.","DOI":"10.1109\/ATEE58038.2023.10108273"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"106611","DOI":"10.1016\/j.bspc.2024.106611","article-title":"Quadratic Convolution-based YOLOv8 (Q-YOLOv8) for localization of intracranial hemorrhage from head CT images","volume":"96","author":"Ferdi","year":"2024","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"141309","DOI":"10.1109\/ACCESS.2023.3339560","article-title":"YOLOv5s-CAM: A Deep Learning Model for Automated Detection and Classification for types of Intracranial Hematoma in CT Images","volume":"11","author":"Vidhya","year":"2023","journal-title":"IEEE Access"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"104378","DOI":"10.1016\/j.bspc.2022.104378","article-title":"Localization of mixed intracranial hemorrhages by using a ghost convolution-based YOLO network","volume":"80","author":"Kothala","year":"2023","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1016\/j.imed.2024.11.002","article-title":"Deep learning-based identification and localization of intracranial hemorrhage in patients using a large annotated head computed tomography dataset: A retrospective multicenter study","volume":"5","author":"Liu","year":"2025","journal-title":"Intell. Med."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"188084","DOI":"10.1109\/ACCESS.2024.3510517","article-title":"Benchmarking YOLO Models for Intracranial Hemorrhage Detection Using Varied CT Data Sources","volume":"12","author":"Tapia","year":"2024","journal-title":"IEEE Access"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Huang, C.C., Chiang, H.F., Hsieh, C.C., Zhu, B.R., Wu, W.J., and Shaw, J.S. (2025). Impact of Dataset Size on 3D CNN Performance in Intracranial Hemorrhage Classification. Diagnostics, 15.","DOI":"10.3390\/diagnostics15020216"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Hoang, Q.T., Pham, X.H., Trinh, X.T., Le, A.V., Bui, M.V., and Bui, T.T. (2024). An Efficient CNN-Based Method for Intracranial Hemorrhage Segmentation from Computerized Tomography Imaging. J. Imaging, 10.","DOI":"10.3390\/jimaging10040077"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Sato, S., Oura, D., and Sugimori, H. (2025). Application of 9-Channel Pseudo-Color Maps in Deep Learning for Intracranial Hemorrhage Detection. Multimodal Technol. Interact., 9.","DOI":"10.3390\/mti9020017"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Mabit, L., Lepoittevin, M., Valls, M., Thomas, C., Guillevin, R., and Herpe, G. (2025). Real-Life Performance of a Commercially Available AI Tool for Post-Traumatic Intracranial Hemorrhage Detection on CT Scans: A Supportive Tool. J. Clin. Med., 14.","DOI":"10.3390\/jcm14134403"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Koyun, M., Cevval, Z.K., Reis, B., and Ece, B. (2025). Detection of Intracranial Hemorrhage from Computed Tomography Images: Diagnostic Role and Efficacy of ChatGPT-4o. Diagnostics, 15.","DOI":"10.3390\/diagnostics15020143"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"9975","DOI":"10.1038\/s41598-023-37114-z","article-title":"A comparison of performance between a deep learning model with residents for localization and classification of intracranial hemorrhage","volume":"13","author":"Angkurawaranon","year":"2023","journal-title":"Sci. Rep."},{"key":"ref_22","unstructured":"Reis, E.P., Nascimento, F., Aranha, M., Secol, F.M., Machado, B., Felix, M., Stein, A., and Amaro, E. (2025, December 01). Brain Hemorrhage Extended (BHX): Bounding Box Extrapolation from Thick to Thin Slice CT Images. Available online: https:\/\/physionet.org\/content\/bhx-brain-bounding-box\/1.1\/."},{"key":"ref_23","unstructured":"(2025, December 01). Scope of Human Research Cases Exempt from Review by the Research Ethics Committee: Wei-Shu-Yi-Zi No. 1010265075, Available online: https:\/\/gazette.nat.gov.tw\/EG_FileManager\/eguploadpub\/eg018127\/ch08\/type1\/gov70\/num35\/Eg.htm."},{"key":"ref_24","unstructured":"(2025, December 01). Scope of Human Research Cases Exempt from Obtaining the Research Subject\u2019s Consent: Wei-Shu-Yi-Zi No. 1010265083, Available online: https:\/\/gazette.nat.gov.tw\/EG_FileManager\/eguploadpub\/eg018127\/ch08\/type1\/gov70\/num36\/Eg.pdf."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"e190211","DOI":"10.1148\/ryai.2020190211","article-title":"Construction of a Machine Learning Dataset through Collaboration: The RSNA 2019 Brain CT Hemorrhage Challenge","volume":"2","author":"Flanders","year":"2020","journal-title":"Radiol. Artif. Intell."},{"key":"ref_26","unstructured":"(2025, December 01). Ultralytics YOLOv8. Ultralytics. Available online: https:\/\/github.com\/ultralytics\/ultralytics."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Lin, T.Y., Goyal, P., Girshick, R., He, K., and Doll\u00e1r, P. (2017, January 22\u201329). Focal Loss for Dense Object Detection. Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy.","DOI":"10.1109\/ICCV.2017.324"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"1137","DOI":"10.1109\/TPAMI.2016.2577031","article-title":"Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks","volume":"39","author":"Ren","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_29","unstructured":"(2015). IBM SPSS Statistics for Windows, IBM Corp."},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27\u201330). Deep Residual Learning for Image Recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Xie, S., Girshick, R., Doll\u00e1r, P., Tu, Z., and He, K. (2017, January 21\u201326). Aggregated Residual Transformations for Deep Neural Networks. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.634"}],"container-title":["Diagnostics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/11\/1705\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T10:36:48Z","timestamp":1780396608000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/11\/1705"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,2]]},"references-count":31,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2026,6]]}},"alternative-id":["diagnostics16111705"],"URL":"https:\/\/doi.org\/10.3390\/diagnostics16111705","relation":{},"ISSN":["2075-4418"],"issn-type":[{"value":"2075-4418","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,2]]}}}