Applications

Have an application you’d like to contribute to this page? Please reach out!

Contents:

Timelapse of timelapses and z-stacks

This notebook shows how to repeatedly acquire a short time series and then a z stack at a set with a set delay in between.

Using multi_d_acquistion_events

This notebook shows how to use the multi_d_acquisition_events function to automatically create events to use for acquistion.

Fast 3D + time imaging with hardware triggering

This notebook acquires a fast TZYX data series. The camera is run at reduced ROI to achieve higher framerate (here 200 frames per second). Movement of the z stage is “sequenced” to speed up acquisition. The z stage advances to the next position in the sequence when a trigger from the camera is received. This eliminates delays due to software communication.

Convert Micromanager multidimensional tiff data into zarr format

This notebook explains how to use pycro-manger to readout the data saved by Micro-manager’s multi-dimensional acquisition and convert it into zarr format. This is useful when the data is large (more than hundreds of GBs). Currently there is no python reader that can directly readout the large multi-dimensional data saved by Micromanger. The Pycro-Manager Java Python Bridge makes this possible.

Interactive Microscopy Control with ImJoy

This tutorial notebook shows how you can combine pycro-manager with ImJoy which is a web framework for building rich and powerful interactive analysis tools. Step by step, it shows how you can use ImJoy plugins for acquiring and visualzing images using a dedicated ImJoy plugin for Pycro-Manager in the notebook with snap/live buttons, exposure/binning controls and a full-featured device property browser. The built plugin can be hosted on Github, and use independently outside the Jupyter notebook interface.

Denoising acquired images using deep learning

This tutorial demonstrates how to train a deep learning model for image denoising using data aquired by Pycro-Manager. This training is performed on Google Colab, which provides free usage of GPUs in the cloud, so no specialized hardware is required to implement it. The trained model will then be used to denoise images in real time using a Pycro-Manager image processor.

Computational quantitative phase imaging from focal stacks

This example shows how to compute 2D quantitative phase images from collected focal stacks, without the need for specialized optics, using computational imaging. Specifically, we will solve and inverse problem based on the Transport of Intensity Equation (TIE). The inverse problem is implemented in an image processor, to enable on-the-fly quantitative phase imaging during acquisition.

Targeted multi-contrast microscopy using attention-based multi-instance learning for tissue sections

This tutorial shows how to use pycro-manager to perform analysis driven targeted multimodal/multiscale acquisition for automated collagen fiber-based biomarker identification. We will acquire brightfield images of a H&E stained cancer histology slide at 4x magnification, identify pathology relevant ROIs using a deep learning model based on the 4x image, and zoom into these ROIs to perform the collagen fiber-specific method of second-harmonic generation (SHG) laser scanning at 20x magnification. This allows for disease-relevant, collagen-specific features to be collected automatically and correlated with the gold standard H&E pathology method. We use Pycro-manager to read/write hardware properties (e.g. camera exposure, lamp intensity, turret position, stage position, etc.), change Micro-Manager hardware property configuration groups, acquire images and access the image data as NumPy array, and perform Z-stack acquisition via multi-dimension acquisition events.

Auto CyCIF Project

This notebook shows an implementation of the CycIF Multiplex immunostaining method on several slides in parallel. It utilizes Micro-Magellan as a user interface to define the bounds of each tissue section, calculates the center of the section, and uses this coordinate to drive the stage underneath the robotic pipettor for staining cycles. It also executes autofocus routines at sub-sampled tiles to provide speed increases. It goes through typical 4 color images and sets an executes a simple auto-expose routine and calculates new exposure times. Finally, it takes the saved data and repackages it into a format that the alignment and stitching software, Ashlar, can accept.

Closed-loop acquisition and perturbation with pycro-manager

This tutorial shows how to use pycro-manager for a simple closed-loop (CL) experiment in python. CL perturbations are triggered by signals derived from data acquired from the sample itself during a live recording session. Well-designed CL experiments are crucial for systems neuroscience (our research domain), but this example can be adapted for any field. Specifically we perform some canonical image processing (template filtering with 2d gaussian kernel, thresholding, median filtering), then find local peaks, then take a window of pixel values around each peak. We use these pixel values to trigger our arbitrary “stimulus” function which can e.g. change optical settings on the microscope, call a separate program, send a serial port command etc.