Project Overview
A Japanese medical device company approached BCT Global to develop an AI-powered assistant that could pre-screen chest X-rays for anomalies, helping radiologists prioritize urgent cases.
The Challenge
Radiologists reviewing 150+ scans daily with rising burnout. AI system needed >90% sensitivity for clinical usefulness.
- 8-12% of urgent findings being delayed
- PMDA approval required extensive documentation
- Patient data must comply with APPI
- PACS integration via DICOM standard essential
Our Solution
- Data Preparation & Model Research (Months 1-3): Collected and anonymized 50,000+ chest X-ray images for training.
- Model Development & Training (Months 3-6): Trained anomaly detection model using PyTorch; achieved 94% sensitivity, 91% specificity.
- Integration & Validation (Months 6-8): Built FastAPI inference server on Azure, DICOM-compliant PACS integration.
- Regulatory & Deployment (Months 8-10): PMDA submission and approval as Class II medical device software. Deployed in 3 hospitals.
System Architecture
Technology Stack
Results & Impact
35%Reduction in Review Time
94%Detection Sensitivity
100%PMDA Compliance
10 moDevelopment to Approval
"BCT Global's team not only built an accurate model but also prepared documentation that our regulatory affairs team described as 'the most thorough they had ever reviewed from an external partner.' Their understanding of both the technical and regulatory aspects of healthcare AI was exceptional."
Chief Medical OfficerJapanese Medical Device Company — PMDA-regulated, Tokyo