Neutrophils Maturation Classifier App

A minimal 3D neutrophil maturation classification app with convolution neural network.

Overview

This application allows for the automated classification of neutrophils at different maturation stages using a 3D convolutional neural network. It's designed to assist researchers and clinicians in analyzing 3D neutrophil morphology with geometric and CNN embedded features.

Note: GPU runtime only available on Linux.

Features

• 3D visualization of neutrophil isosurfaces
• Real-time classification of neutrophil maturation stages
• Support for TIFF image input
• Batch processing capabilities

Installation

Installation with Conda (Mamba)

1. Download Miniforge and install it if you don't have conda/mamba.
2. Create and activate a new environment:

   mamba create -n neutrophils python=3.10
   mamba activate neutrophils
   

3. Install the package:

   pip install neutrophils-classifier-app
   

Note: This package now uses the published neutrophils-core library from PyPI, which provides the core machine learning functionality.

Run the App

neutrophil-classifier

Note: The first launch may take up to 10 minutes as the application performs initial setup and caching. Subsequent startups will be significantly faster.

For Developers

Development Setup

1. Set up the conda environment:

git clone git@github.com:bpi-oxford/Neutrophils-Classifier-App.git
cd Neutrophils-Classifier-App
git submodule update --init --recursive
mamba env create -f env.yaml
mamba activate neutrophils

2. Install in editable mode:

pip install -e .

3. Run the application:

neutrophil-classifier

Packaging Setup

1. Set up the conda environment:

mamba env create -f env-minimal.yaml
mamba activate neutrophils-minimal

2. See BUILD_GUIDE.md for detailed build and publish instructions.