Explore the research
Anatomy-Preserving Augmentation via Generative AI
A workflow for augmenting medical training images with anatomical controls and generated-sample selection.
Mobile Extraction of Carotid Ultrasound IMT Regions
A conference abstract combining MobileNetV3 image classification with SegNet segmentation to examine efficient carotid ultrasound analysis.
Image Translation for Polyp Detection Training
A study augmenting polyp-detector training with NBI-like images generated from white-light endoscopy.
We study lesion detection, segmentation, classification, and quantitative diagnostic support using endoscopic images, CT images, cellular images, and pathological images.
The laboratory combines deep learning with optics-, geometry-, and physics-based image understanding so that medical AI can not only detect lesions, but also estimate shape, depth, and size.
3D reconstruction of endoscopic polyps from monocular or sparse-view images is one of our characteristic research topics.
Why It Is Difficult
Medical images are affected by institution-specific acquisition protocols, device differences, patient variability, and lesion diversity. Small datasets and class imbalance are common.
Endoscopy further involves non-rigid deformation, specular highlights, illumination changes, and camera motion, making shape reconstruction and absolute-scale estimation from a single view intrinsically difficult.
Practical systems must also satisfy clinical constraints such as acceptable false positives, measurement stability, explainability, and workflow compatibility.
Approach
We connect data-driven deep learning with optical, geometric, and physical constraints, treating detection and 3D reconstruction as a connected problem.
1. Lesion detection, classification, and segmentation
We use transfer learning, semi-supervised learning, generative models, and data augmentation to build robust methods for endoscopic, CT, cellular, and pathological images.
2. 3D shape reconstruction and quantitative measurement
We estimate relative depth, 3D shape, and absolute size from monocular or sparse-view images using cues such as vascular structure, shading, illumination, and projection models.
3. Domain shift handling
We design medical image AI that remains stable under institution, device, and imaging-condition shifts through domain adaptation and data augmentation.
4. Integration with clinical workflows
We evaluate whether the methods reduce missed lesions, improve measurement reproducibility, and shorten diagnostic workflows.
Evaluation
Evaluation separates detection, classification, segmentation, depth error, surface error, size-estimation error, and reconstruction stability.
We also consider clinical usefulness, reduction in diagnostic time, missed-lesion reduction, and measurement reproducibility.
Current Questions
- 13D shape and absolute-size estimation of endoscopic polyps using vascular structure and optical cues
- 2Robust lesion detection, classification, and segmentation across small labels, institutions, and devices
- 3Multitask diagnostic support integrating detection, reconstruction, and malignancy estimation
- 4Learning and validation pipelines using generative models and explainability
Demo
A demo for reconstructing 3D shape from a single image is available.
DemoRelated publications (8)
Lesion-Wise Selection of Clinical-Attribute Fusion Stages for Dermoscopic Image Classification
Studies lesion-wise selection of the stage at which clinical attributes are combined with dermoscopic image features.
Izuhara, Keima, Usami, Hiroyasu
Proceedings of the 60th Tokai Fuzzy Workshop in Toyohashi (Gamaken 2026), 2026 (8)
S2-03, pp. 12-13, August 19, 2026
AI Arteriosclerosis Diagnosis System Using RLHF and Adversarial Attack: Quantitative Evaluation of IMT Extraction Robustness by Adversarial Learning
Studies carotid ultrasound wall-region extraction using human-feedback learning and robustness evaluation under adversarial inputs.
Usami, Hiroyasu
Journal of Comprehensive Engineering, 2026 (3)
Vol. 38, pp. 1-11
Anatomy-Preserving Diffusion-Based Data Augmentation for Medical Image Segmentation
Studies medical segmentation data augmentation with anatomical controls and selection of generated images.
Tominari, Hiroki, Nakayama, Natsuki, Arakawa, Naoko, Usami, Hiroyasu
Proceedings of the 19th International Joint Conference on Biomedical Engineering Systems and Technologies -- BIOIMAGING, 2026
BIOIMAGING 2026
Data Augmentation for Retinal Image Segmentation Based on Ground-truth Mask Images
Studies augmentation for retinal image segmentation using ground-truth masks as a guide.
Tominari, Hiroki, Usami, Hiroyasu
Proceedings of the 2025 Tokai-Section Joint Conference on Electrical, Electronics, Information, and Related Engineering, 2025 (9)
2025-09-18; program title variant: Data augmentation for retinal image segmentation based on ground-truth masks
Efficient Automatic IMT Region Extraction Using MobileNetV3 and SegNet for Mobile Carotid Ultrasound Imaging
A conference abstract combining MobileNetV3 image classification with SegNet segmentation to examine efficient carotid ultrasound analysis.
Usami, Hiroyasu, Nakayama, Natsuki, Arakawa, Naoko
BIOIMAGING 2025, 2025
BIOIMAGING25-367: Abstract, 20 min oral presentation
SegNet-based IMT Region Segmentation in Carotid Ultrasound Images
Studies segmentation of carotid wall regions in ultrasound images using SegNet.
Doi, Mizuki, Usami, Hiroyasu, Nakayama, Natsuki, Arakawa, Naoko
Proceedings of the 56th Tokai Fuzzy Workshop, 2024 (8)
2024-08-05
Polyp Shape Recovery from Single Endoscope Image using Medical Suture
A study using a known-diameter suture and a lighting model to estimate approximate polyp shape and absolute scale from one endoscope image.
Usami, Hiroyasu, Iwahori, Yuji, Wang, Aili, Bhuyan, M. K., Ogasawara, Naotaka, Kasugai, Kunio
The Open Bioinformatics Journal, 2019 (1)
Vol. 12, No. 1, pp. 1–17
Recovering Polyp Shape from an Endoscope Image Using Two Light Sources
A study combining a two-light photometric model with learned normal correction to estimate depth and shape from endoscope imagery.
Usami, Hiroyasu, Iwahori, Yuji, Hanai, Yuki, Kijsirikul, Boonserm, Kasugai, Kunio
International Journal of Software Innovation, 2017 (4)
Vol. 5, No. 2, pp. 33–54