Skip to content

yoavfreund/shapeology_code

Repository files navigation

Guidelines for Shapeology

About

This project is meant to analyze brain images based on cell shape features. The functions are:

  • Extracting cells from brain images
  • Learning cell shape features
  • Training decision trees with cell shape features
  • Generating detection score maps
  • Finding statistically significant regions
  • To be continued

Developed with Python 3.7.

Installation

Please read the Installation.md for the details of the installation process.

Usage

Step 1: Cell Extraction

This step is to extract cells from brain section images and permute the generated cell patches in a random order.

Cell_extractor_local.md offers a guided example showing how to complete this step on your computer locally.

To speed up the process, you can refer to (Cell_extractor_aws.md) for the method to run on multiple AWS instances. Essential credential files of AWS and datajoint (VaultBrain in Installation.md) are required.

Step 2: Diffusion Map Training

This step is to find representative cell patches via K-means++ and then use them to calculate diffusion map.

Diffusionmap.md provides a guided example showing how to complete this step on your computer.

About

Analysis of shape distribution using graph theory

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages