rtm-pymmcore

Real-time feedback control microscopy using pymmcore as an interface.

Overview

This repository enables real-time cell segmentation and feature extraction during microscopy experiments. Instead of relying on post-processing, this approach provides immediate feedback control. For example, if spatial stimulation is available, stimulation areas can be dynamically defined based on cell morphology.

Another key application is real-time segmentation and feature extraction to streamline experiments and minimize post-processing requirements.

Information

This repository is a work in progress and leverages pymmcore-plus to control the microscope and acquire images. The workflow follows these main steps:

1. Defining Acquisition Positions & Stimulation Parameters

   • A dataframe is generated to store acquisition positions, which can be selected using pymmcore-plus GUI widgets.
   • Stimulation parameters (e.g., pulse duration, intensity) are defined and configured within the Jupyter notebook files.

2. Image Acquisition & Processing

   • The dataframe is passed to the core of rtm-pymmcore, which handles image acquisition, segmentation, and feature extraction.
   • Optionally, an external segmentation engine (running on another computer or instance via imaging-server-kit) can be used.
   • Single-cell results are stored in a dataframe.

3. Data Visualization & Analysis

   • The viewer script allows visualization of individual fields of view (FOVs) along with their corresponding stimulation conditions.
   • The data_analysis_plotting scripts facilitate report generation.

Microscope Setup

The workflow relies on a uManager configuration file for microscope setup. The configuration file must include:

• A setup group with a Startup preset (executed at script initialization).
• Additional presets for each fluorophore used in the experiment (e.g., filter or laser selection).

Running the Script

• The workflow is executed step by step via Jupyter notebooks.
• Currently, only full field of view (FOV) stimulation is implemented, but the modular code structure allows easy adaptation to other stimulation patterns.
• Two versions of the 01_ERK-KTR_full_fov_stimulation scripts exist for different microscope setups:
  • 01_ERK-KTR_full_fov_stimulation_Jungfrau (recommended for a general overview).
  • 01_ERK-KTR_full_fov_stimulation_Niesen, which includes an additional routine to prevent laser sleep.