AI Medical Diagnostic SaMD
Medical device software powered by cutting-edge deep learning. Analyzes medical images with high precision, delivering real-time insights and explainable results that physicians can trust.
Solutions
We serve healthcare needs through five pillars.
Business Pillars
Cloud Radiology Services
Remote image interpretation by specialist physicians. Delivering high-quality radiology reports in a timely manner regardless of location, helping bridge regional gaps in healthcare access. Our platform architecture, built for the future implementation of an AI inference engine, enables a smooth expansion to advanced AI-powered cloud radiology services.
Medical AI Managed Services
A managed service providing high-quality data annotation by specialist physicians and domain experts. Through a multi-stage quality assurance process, we build expertly curated datasets that support pharmaceutical R&D, drug discovery, and AI development. We plan to expand the service to specialized domains beyond healthcare in the future.
Medical Data Services
Generating rich structured datasets from medical images. Partnering with pharmaceutical companies to accelerate drug discovery and R&D with high-quality AI-analyzed clinical data and quantitative biomarkers.
Custom Medical AI Development
Custom AI model and system development. Building tailored solutions from concept to deployment for your unique diagnostic and imaging challenges.
Clinical Domains
Our AI Medical Diagnostic SaMD covers four clinical domains:
ophthalmology, gastroenterology, orthopedics, and neurology.
Grounded in deep clinical expertise across each area, we deliver solutions that address real clinical needs.
Ophthalmology
Overview
AI-powered diagnostic support systems for retinal diseases using OCT (Optical Coherence Tomography) and fundus photographs. Our system detects major retinal conditions that can lead to vision loss with high accuracy, supporting ophthalmologists in their diagnostic workflow.
Target Modalities
OCT (Optical Coherence Tomography)Fundus Camera
Target Diseases
Diabetic Retinopathy (DR)
Age-related Macular Degeneration (AMD)
Glaucoma
Retinal Vein Occlusion (RVO)
Macular Hole (MH)
Epiretinal Membrane (ERM)
Central Serous Chorioretinopathy (CSC)
Retinal Detachment (RD), etc.
Solutions
OCT Retinal Disease Classification AI
Automated analysis of retinal layer structures and disease classification. Detects over 10 types of retinal findings with high accuracy.
Diabetic Retinopathy Diagnostic AI
Fundus photograph-based diabetic retinopathy severity classification (Davis classification) and lesion segmentation. Automatically detects hemorrhages, exudates, neovascularization, and other pathological findings.
Severity / Staging Assessment
Beyond the presence or absence of disease, we also address severity and progression assessment—for example, staging of diabetic retinopathy (Stage 1–4), among other staged evaluations of major conditions.
Automated Finding Extraction & Visualization
Automatically extracts a range of OCT findings (such as subretinal fluid (SRF) and layer-structure abnormalities), with per-finding color coding and tagging.
Technical Features
- Novel AI model for OCT retinal disease classification (patent filed)
- XAI (Explainable AI) technology for visualizing diagnostic rationale
- Versatile AI model compatible with OCT devices from multiple manufacturers
Gastroenterology
Overview
An integrated AI diagnostic support system for abdominal gastrointestinal CT imaging, spanning the hepatobiliary-pancreatic organs (liver, biliary tract, pancreas) through the gastrointestinal tract (colon, small intestine, duodenum). Our goal is early detection of difficult-to-diagnose cancers with poor prognosis, working toward integrated imaging diagnosis across the abdominal digestive organs.
Target Modalities
CT (Contrast-enhanced / Non-contrast)
Target Diseases
Hepatocellular Carcinoma (HCC)
Metastatic Liver Cancer
Intrahepatic Cholangiocarcinoma (iCCA)
Gallbladder Cancer (GBC)
Bile Duct Cancer (CCA)
Pancreatic Cancer (PDAC)
Pancreatic Cystic Neoplasms (PCN)
Colorectal Cancer (CRC)
Small Bowel Tumors
Duodenal Tumors, etc.
Solutions
Integrated HBP CT Diagnostic AI
AI system for integrated analysis of liver, biliary tract, and pancreas. Combines organ segmentation with tumor detection for efficient diagnostic support.
GI Tract CT Tumor Detection AI
AI that detects and segments tumor lesions from CT images across the colon, small intestine, and duodenum. Tumor segmentation for the colon and small intestine is already in place, working toward integrated abdominal-GI diagnosis together with the hepatobiliary-pancreatic system.
Technical Features
- High-precision model based on state-of-the-art medical image segmentation technology
- Integrated analysis of multi-phase contrast CT (arterial, portal venous, equilibrium phases)
- Training with precisely annotated data from specialist physicians
- Novel AI model for hepatobiliary-pancreatic micro-tumor segmentation (patent application in preparation)
Orthopedics
Overview
AI diagnostic support systems for fractures and bone tumors using X-ray imaging. Our solutions support rapid diagnosis in emergency and primary care settings, streamlining triage to orthopedic specialists.
Target Modalities
X-ray
Target Diseases
Proximal Femur Fractures
Pediatric Supracondylar Humerus Fractures
Primary Bone Tumors, etc.
Solutions
X-ray Fracture Detection AI
Automatic detection of high-risk fractures including proximal femur fractures (common in elderly) and pediatric supracondylar humerus fractures.
X-ray Bone Tumor Screening AI
AI system that detects bone tumors from X-ray images and supports benign/malignant assessment, helping prevent missed lesions and enabling early detection and appropriate triage.
Technical Features
- Automated fracture and bone tumor detection from X-ray images
- Designed for rapid diagnostic support in emergency and primary care settings
Priority Areas
- Elderly Proximal Femur Fractures: Major cause of bedridden status with significant social impact
- Pediatric Supracondylar Humerus Fractures: High risk of missed diagnosis in emergency settings, early treatment critical
- X-ray Bone Tumors: Early differentiation between benign and malignant is critical, supporting appropriate triage
Neurology
Overview
In neurology, we work on two solutions: supporting thrombolysis (t-PA) eligibility decisions in acute ischemic stroke, and interpretation support for findings identified in brain dock (screening) exams. From acute care to preventive screening, we broadly support the interpretation of brain MRI and CT.
Target Modalities
CTMRI
Acute Care Targets
Acute Ischemic Stroke
t-PA (Alteplase) Eligibility Assessment
Key Brain Dock Findings
Cerebral white matter lesions
Cerebral microbleeds
Silent cerebral infarction
Carotid / major cerebral artery stenosis, occlusion, hypoplasia
Cerebral aneurysm
Brain tumors, etc.
Solutions
Acute Stroke t-PA Decision Support
End-to-end application supporting t-PA eligibility decisions for acute ischemic stroke patients. Integrates clinical checklists through to decision support, aiming to contribute to reducing Door-to-Needle Time.
Brain Dock Interpretation AI
AI solution supporting everything from detection of findings identified in brain dock exams through to patient explanation. Assists brain-MRI interpretation and enables early detection of asymptomatic abnormalities.
Technical Features
- The acute-care solution is designed as an end-to-end integrated application (checklist + decision support)
- The brain dock solution consistently supports the workflow from finding detection to patient explanation
- UI and workflow designed for the respective emergency and screening settings
Clinical Significance
- For t-PA, rapid decision-making within 4.5 hours of onset determines patient prognosis, driving strong demand for decision support
- In brain dock, it is important not to miss asymptomatic findings and to connect them to appropriate follow-up and treatment
- Supporting neurology from both acute and preventive perspectives with AI