OcupathIF

OcupathIF is a desktop application developed by biostate.AI that automates the full analysis pipeline for multi-channel fluorescence microscopy images used in pathology research. Designed for researchers who need to move quickly from raw image data to actionable insight, OcupathIF handles cell segmentation, phenotyping, and spatial analysis — all without writing a single line of code. The application runs on both macOS (Apple Silicon and Intel) and Windows, and is available as a free, one-click install.

Built around a guided five-step workflow — View, Analyze, Segment, Count, and Export — OcupathIF takes qptiff images from Akoya Vectra/Polaris systems and delivers CSV-ready results in minutes. Its combination of AI-powered segmentation, spectral deblending, and interactive visualization makes it suitable for pathology research teams working with complex multi-channel immunofluorescence data.

Automated Cell Analysis Pipeline

• Accepts qptiff format from Akoya Vectra/Polaris imaging systems
• Automatically extracts excitation/emission wavelength metadata and fluorophore names from qptiff files
• AI-powered nuclear segmentation using Cellpose 3.x
• Optional SAM 2.1 boundary refinement for improved cell delineation
• Spectral deblending to separate overlapping fluorophore signals for accurate phenotyping
• Binary phenotype classification based on channel expression
• Spatial proximity analysis with configurable distance thresholds and visual mask overlays

Interactive Visualization and Review

• Interactive whole-slide viewer allowing researchers to inspect and adjust results at every step before export
• False-color channel overlays for intuitive multi-channel image review
• Adjustable expression gates with GMM-based automatic gating and customizable per-channel thresholds
• Cell boundary visualization directly within the viewer
• Spatial conditioning queries enabling questions such as "How many CD8+ cells are within 50μm of PD-L1+ cells?"

Performance and Speed

• Thumbnail-first progressive loading delivers an image preview in approximately 9 seconds, compared to around 3 minutes without this approach
• In-memory caching enables same-image reloads in approximately 40 milliseconds
• DZI tile generation at 0.37 seconds per 5000×5000 channel using pyvips
• Parallel tile processing delivers a 10x pipeline speedup over serial methods, making large whole-slide images tractable

Key Features

• Automatic channel detection: reads excitation/emission wavelengths and fluorophore names directly from qptiff metadata
• AI cell segmentation: Cellpose 3.x nuclear detection with optional SAM 2.1 boundary refinement
• Spectral deblending: separates overlapping fluorophore signals for accurate phenotyping
• Expression gating: customizable per-channel thresholds with GMM-based automatic gating
• Spatial analysis: proximity queries with configurable distance and visual mask overlays
• Multi-language support: English, Chinese, Arabic, and Hindi
• Cross-platform: compatible with macOS (Apple Silicon and Intel) and Windows
• One-click install: standalone application with automatic Python environment setup, no coding required

OcupathIF is available as a free standalone desktop application for macOS and Windows. It requires no coding expertise and is designed for rapid deployment in pathology research environments. For teams working with Akoya Vectra/Polaris imaging systems, OcupathIF provides a complete, self-contained solution from raw fluorescence image to exportable cell-level data. A demo can be requested via biostate.AI at [email protected].