Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jun 13, 2026

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

Cross-Dataset Generalization of Deep Learning-Based Detectors for Intracranial Hemorrhage Subtype Localization on

Chiao-Hua Lee¹, Hikam Muzakky¹, Cheng-En Juan^2,3

¹Department of Medical Imaging, China Medical University Hsinchu Hospital, Hsinchu 302, Taiwan.

Diagnostics (Basel, Switzerland)

|June 12, 2026

Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Transcriptomics, lifestyle, and autoimmune thyroiditis: an integrated association study.

Frontiers in immunology·2026

Same author

Strontium-crosslinked potassium-alginate beads as coordination-driven ion-exchange biomaterials for regenerative microenvironment engineering.

International journal of biological macromolecules·2026

Same author

Nervous necrosis virus B2 non-structural protein shut off host translation via sequestration of polyadenylate-binding protein.

Biochemical and biophysical research communications·2026

Same author

Advances in nanotechnology for addressing biofilm-related challenges in medical devices.

Nanotechnology·2026

Same author

An Adaptive Deep Learning Framework for Multi-Label Chest X-Ray Diagnosis Using a Hybrid CNN-Transformer Architecture and Class-Wise Ensemble Fusion.

Diagnostics (Basel, Switzerland)·2026

Same author

scDIAGRAM: detecting chromatin compartments from individual single-cell Hi-C matrix without imputation or reference features.

Briefings in bioinformatics·2026

Deep learning models for intracranial hemorrhage (ICH) subtype localization show varied performance. Dataset characteristics significantly impact model generalization, highlighting the need for external validation before clinical use.

Area of Science:

Medical Imaging
Artificial Intelligence in Medicine
Radiology

Background:

Intracranial hemorrhage (ICH) subtype localization on noncontrast head CT is crucial for diagnosis and treatment.
Evaluating deep learning detector performance across different datasets and architectures is essential for clinical translation.

Purpose of the Study:

To assess the impact of detector architecture and dataset characteristics on ICH subtype localization.
To evaluate the bidirectional cross-dataset generalization capabilities of deep learning models.

Main Methods:

Retrospective analysis of two public datasets (BHX and RSNA 2019+) for training and external validation.
Evaluation of six deep learning detectors, including CNN-based and Swin Transformer-based models.
Performance assessment using mAP@50, BB-DSC, and BB-IoU, with image-level and patient-level analyses.

Keywords:

deep learning intracranial hemorrhage noncontrast head CT object localization

Related Experiment Videos

Last Updated: Jun 13, 2026

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

Main Results:

Swin-RT-DETR showed subtype-dependent advantages in internal validation, while Faster R-CNN performed comparably for IVH and better for IPH.
Substantial performance degradation was observed during external validation across all architectures and subtypes.
Dataset characterization revealed significant differences in subtype prevalence, geometry, and annotation between datasets.

Conclusions:

ICH subtype localization performance is highly sensitive to dataset characteristics, annotation variations, and domain shift.
The advantages of advanced architectures like Swin-RT-DETR diminish under external validation.
Rigorous external validation, dataset characterization, and patient-level evaluation are critical for clinical integration of automated ICH localization models.