AI-Based Cervical Cytology Image Processing System
WinCyto
AI‑assisted analysis software designed to support cervical cytology workflows. Provides intuitive visualization, intelligent cell detection, and structured result presentation to help cytopathologists improve screening efficiency and consistency.
• Pretrained on 1M+ annotations
• Intuitive interface presents key findings
• Precise TBS‑Oriented Classification
• Flags suspicious findings
Product Overview
Trained on a curated dataset of over 1 million annotated cytological specimens, WinCyto identifies morphologically suspicious regions and organizes findings aligned with the TBS reporting framework. Designed as a pathologist-assistive tool rather than an autonomous diagnostic engine, the system keeps clinical decision-making firmly under human oversight throughout the screening and review pipeline.
The platform is intended to improve laboratory efficiency by supporting case triage, highlighting critical regions, and reducing repetitive visual review — while maintaining full pathologist control over final interpretation.
Core Features
At-a-glance Interface
Pretrained on 1M+ professionally annotated cytology samples
Highlights suspicious findings and critical regions
Aligned with TBS classification standards
Supports faster case review and triage
Professional Cytology Workflows
Optimized for high-throughput screening environments
Reduces repetitive manual review tasks
Consistent AI-assisted pre-screening support
Pathologist-in-the-loop review architecture
Multiple Image Format Support
SVS whole-slide image compatibility
TIFF multi-layer image support
Native support for ZYP slide format
Broad Sample Preparation Compatibility
Membrane-based preparation
Filter-based LBC slide preparation
Centrifugation-based preparation
Cell concentration by centrifugation
Gravity-based sedimentation
Natural sedimentation workflow support
Performance Metrics
| Metric | Value |
|---|---|
| Average Analysis Time / Slide | ≤ 60 seconds |
| Sensitivity | 99% |
| Specificity | 90% |
| Negative Predictive Value (NPV) | 99% |
Performance data derived from internal validation studies. Results may vary depending on specimen quality and workflow configuration.
