pathology-roi-selector

What This Skill Does

The pathology-roi-selector is a specialized OpenClaw agent skill designed for high-precision Region of Interest (ROI) detection within Whole Slide Imaging (WSI) datasets. At its core, the skill provides a structured execution environment for processing large-scale pathology images, ensuring that every analysis step is reproducible, documented, and bounded by clear logical constraints. It serves as an essential utility for data scientists and pathologists who need to convert raw WSI files into clean, usable training data for AI/ML pipelines.

The skill operates by utilizing a hardened entry point (scripts/main.py) which enforces strict adherence to parameter validation. By mandating explicit configuration before execution, it prevents common errors associated with arbitrary data processing. It excels at parsing image scopes, managing threshold configurations, and producing consistent output formats, making it a reliable building block for complex digital pathology workflows.

Installation

You can integrate this skill into your OpenClaw environment using the following command:

clawhub install openclaw/skills/skills/aipoch-ai/pathology-roi-selector

Ensure that your system environment meets the prerequisite of Python 3.10+ to maintain compatibility with the underlying repository baseline. Before running, it is recommended to verify the environment using python -m py_compile scripts/main.py.

Use Cases

• AI Model Training: Isolate specific tissue regions from WSIs to create ground-truth datasets for deep learning models.
• Quality Control: Automatically flag regions that fail defined quality thresholds, ensuring only high-fidelity data enters the analysis pipeline.
• Standardized Research Reporting: Use the tool to generate consistent, audit-ready outputs when performing multicenter clinical studies.
• Workflow Automation: Integrate ROI extraction into larger batch-processing scripts where consistent output boundaries are non-negotiable.

Example Prompts

1. "Run the pathology-roi-selector on the WSI file located at /data/samples/patient_001.svs using a threshold of 0.85 for tissue detection."
2. "Extract all ROIs from the directory /data/batch_05/ and save the coordinate metadata to /outputs/json/ while keeping the original image references intact."
3. "Validate the configuration for the ROI selector and perform a test run on the provided sample image to check for execution errors."

Tips & Limitations

To maximize the utility of this skill, always define your scope filters clearly. Because this tool is designed for structured workflows, it does not support ad-hoc or 'guess-heavy' processing. Always review the CONFIG block in scripts/main.py before execution. If your input files are exceptionally large, ensure your environment has sufficient memory allocated for image processing buffers. Finally, because this script performs local file writes, maintain a clean directory structure to avoid overwriting existing data. If you encounter errors, verify that your input image format is supported by the underlying processing libraries.