Cross-Dataset Generalization of Deep Learning-Based Detectors for Intracranial Hemorrhage Subtype Localization on Noncontrast Head CT: A Comparative Study.

Overview

1. Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan
2. China Medical University Hospital, Taichung 404, Taiwan
3. Master’s Program of Biomedical Informatics and Biomedical Engineering, Feng Chia University, Taichung 407, Taiwan
4. Department of Management Science, National Yang Ming Chiao Tung University, Hsinchu 300, Taiwan
5. Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei 106, Taiwan
6. Show Chwan Memorial Hospital, Changhua 500, Taiwan
7. Department of Medical Imaging, China Medical University Hospital, Taichung 404, Taiwan
8. Department of Radiology, School of Medicine, College of Medicine, China Medical University, Taichung 406, Taiwan
9. Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu 300, Taiwan
10. Department of Computer Science and Information Engineering, National Taiwan University, Taipei 106, Taiwan

Abstract

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, 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 0.54–0.79 in the BHX-to-RSNA+ direction and 0.17–0.74 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

Reproduced under the paper's license (CC BY), from the paper cited above.

Code

The paper links to its data, not to its authors' code: see the Data section.

Tracing map

A tracing map links a paper to the code its authors published: this paper has none, so it has no map.

Data

Datasets cited

• physionet.org/content/bhx-brain-bounding-box, at PhysioNet; found in “Data Availability Statement”

Data Availability Statement

The datasets used in this study are publicly available. The Brain Hemorrhage Extended (BHX) dataset is available at PhysioNet (https://physionet.org/content/bhx-brain-bounding-box/1.1/) (accessed on 1 November 2025). The RSNA 2019+ dataset is available at (http://www.multi-lesions.top) as provided by Liu et al [14] (accessed on 25 November 2025). Additional data generated or analyzed during this study are available from the corresponding author upon reasonable request.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 9 authors, 4 keywords, 2 funders, 24 references.

Cite

This paper

Lee, C.-H., Muzakky, H., Juan, C.-E., Chang, C.-C., Li, Y.-H., Lee, T.-Y., Juan, C.-H., Tsai, M.-T., & Juan, C.-J. (2026). Cross-Dataset Generalization of Deep Learning-Based Detectors for Intracranial Hemorrhage Subtype Localization on Noncontrast Head CT: A Comparative Study. Diagnostics (Basel, Switzerland), 16(11), 1705. https://doi.org/10.3390/diagnostics16111705

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.3390/diagnostics16172776

A Resource-Efficient Hybrid Deep Learning Framework with External Validation for Intracranial Hemorrhage Detection in CT Scans.

Journal: Diagnostics (Basel, Switzerland)

In common: physionet.org/content/bhx-brain-bounding-box, other, 2 references

[2] doi:10.1016/j.patter.2026.101538 [code]

A multi-modal foundation model for brain disease diagnosis and medical imaging.

Journal: Patterns (New York, N.Y.)

In common: 2 references

[3] doi:10.1016/j.ynirp.2026.100375

Deep learning-based identification of causative genes in lissencephaly using 3D-MRI volumetric datasets.

Journal: Neuroimage. Reports

In common: 2 references

[4] doi:10.1155/ijbi/9929121 [code]

Automated Semisupervised Measurement of Optic Nerve Sheath Diameter From CT Following Traumatic Brain Injury.

Journal: International journal of biomedical imaging

In common: other, methods / tools, 1 reference

[5] doi:10.3389/fneur.2026.1849021

Early pharmacologic venous thromboembolism prophylaxis and in-hospital mortality in ICU patients with intracerebral hemorrhage: a landmark, propensity-weighted cohort study.

Journal: Frontiers in neurology

In common: stroke, 1 reference

[6] doi:10.1038/s41598-026-50240-8

High-precision brain tumor segmentation with switchable normalization in faster R-CNN architecture.

Journal: Scientific reports

In common: methods / tools, 1 reference

[7] doi:10.2196/84095

Detection of Interpretable and Fine-Grained Brain Tumor Magnetic Resonance Imaging Based on Progressive Pruning: Machine Learning Model Development and Validation Study.

Journal: JMIR medical informatics

In common: methods / tools, 1 reference

[8] doi:10.1002/mp.70419 [code]

Comparing deep learning stroke segmentation in NCCT, CTA, and CTP: Accuracy, domain transfer, and temporal sampling effect.

Journal: Medical physics

In common: stroke, other, methods / tools

[9] doi:10.1371/journal.pone.0350637 [code]

Deep learning-based arterial waveform analysis for predicting postoperative cerebrovascular events in pediatric patients with Moyamoya disease.

Journal: PloS one

In common: stroke, 1 reference

[10] doi:10.3390/s26144390 [code]

Neuro-Symbolic Class-Contrast Evidence Audit for Reliable Cross-Subject Wearable Activity Recognition.

Journal: Sensors (Basel, Switzerland)

In common: other, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.