HiTMicTools Documentation

Welcome to the HiTMicTools documentation. HiTMicTools is a comprehensive toolkit for High-Throughput Microscopy Analysis developed by the Boeck Lab at University Hospital of Basel. It provides deep learning-based image processing pipelines for automated microscopy analysis, including cell segmentation, focus restoration, classification, and tracking.

What is HiTMicTools?

HiTMicTools streamlines the analysis of high-throughput microscopy data through:

• Automated Image Processing: Focus restoration, alignment, and background correction for brightfield and fluorescence channels

• Deep Learning Models: Integrated neural network models for cell segmentation, classification, and quality control

• Flexible Pipelines: Multiple analysis workflows (ASCT_focusrestore, ASCT_scsegm, ASCT_zaslavier, etc.)

• Cell Tracking: Optional btrack-based trajectory reconstruction for lineage analysis

• Scalable Processing: Built-in support for parallel processing and SLURM cluster deployment

• Resource Management: Smart GPU/CPU memory management for multi-process environments

Key Features

• Model Collections: Simplified deployment with bundled model packages (.zip files)

• Command-Line Interface: Easy automation via hitmictools CLI

• Configuration-Based: YAML configuration files for reproducible analyses

• Multiple Input Formats: Support for ND2, TIFF, OME-TIFF, and Jetraw-compressed images

• Comprehensive Outputs: CSV measurements, labeled masks, aligned images, and tracking data

Getting Started

This documentation guides you through:

1. Basic Analysis: Running standard pipelines without tracking

2. Advanced Tracking: Enabling btrack-based cell trajectory analysis

3. Cluster Computing: Deploying large-scale analyses on SLURM clusters

4. Model Management: Working with model collections and individual checkpoints

User Guide

• Launching the Analysis Pipeline
  • Overview
  • 1. Prerequisites
  • 2. Project Structure
  • 3. Configuration File
  • 4. Running the Analysis
  • 5. Understanding the Output
  • 6. Monitoring and Performance
  • 7. Troubleshooting
  • 8. Advanced Configuration
• Launching Analysis with Cell Tracking
  • Overview
  • About btrack
  • 1. Prerequisites
  • 2. Understanding Tracking Configuration
  • 3. Pipeline Configuration
  • 4. Running Analysis with Tracking
  • 5. Understanding Tracking Output
  • 6. Validating Tracking Results
  • 7. Optimizing Tracking Parameters
  • 8. Troubleshooting
  • 9. Advanced Tracking Features
  • 10. Downstream Analysis
  • Summary
• Using SLURM for Large-Scale Analysis
  • Overview
  • Workflow Summary
  • 1. Prerequisites
  • 2. Splitting Files into Batches
  • 3. Generating SLURM Scripts
  • 4. Customizing SLURM Scripts
  • 5. Submitting and Managing Jobs
  • 6. Configuration for SLURM
  • 7. Best Practices
  • 8. Example Workflows
  • 9. Troubleshooting
  • 10. Advanced Topics
  • Summary
• Model Management and Resource Control
  • Overview
  • 1. Model Collections (Recommended Approach)
  • 2. Individual Model Specification (Advanced)
  • 3. Model Architectures
  • 4. Resource Management
  • 5. Model Performance and Optimization
  • 6. Troubleshooting Models
  • 7. Model Storage Best Practices
  • Summary

Quick Start

Install HiTMicTools and run your first analysis:

# Create conda environment
conda create -n hitmictools python=3.9
conda activate hitmictools

# Install HiTMicTools
pip install git+https://github.com/phisanti/HiTMicTools

# Run analysis with a configuration file
hitmictools run --config your_config.yml

For detailed installation instructions and requirements, see the README.