Data

Tools

Indexes

Applications

Experiments

Open Source Science

ct-body-composition

An artificial intelligence model for body composition analysis in CT scans

https://github.com/rosenthal-lab-at-dana-farber/ct-body-composition

Science Score: 67.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ✓
    DOI references
    Found 2 DOI reference(s) in README
  • ✓
    Academic publication links
    Links to: arxiv.org, springer.com
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (14.1%) to scientific vocabulary

Keywords

ai cancer computervision imaging oncolocy research
Last synced: 6 months ago · JSON representation ·

Repository

An artificial intelligence model for body composition analysis in CT scans

Basic Info
  • Host: GitHub
  • Owner: Rosenthal-Lab-at-Dana-Farber
  • License: gpl-2.0
  • Language: Python
  • Default Branch: master
  • Homepage:
  • Size: 146 KB
Statistics
  • Stars: 9
  • Watchers: 1
  • Forks: 2
  • Open Issues: 0
  • Releases: 0
Topics
ai cancer computervision imaging oncolocy research
Created almost 2 years ago · Last pushed 12 months ago
Metadata Files
Readme License Citation

README.md

CT Body Composition

This repository provides code and models weights for training and running body composition estimation models on abdominal CT scans.

Note: Git-LFS

The model weights are stored in this repository using git-lfs. Please ensure you have git-lfs set up following the instructions here instructions before cloning the repository.

Getting Started

After cloning the repository, you have two options for setting up the environment. You may install all the necessary components directly on your system, or if you have docker on your machine, you may build a docker image that contains all the necessary requirements.

See the documentation pages for further details: * Installation - For installing directly on your system * Docker - For building and using the docker image * Training - For training new models * Inference - For running the model on new data

Publications

This code accompanies the following publication:

Population-Scale CT-Based Body Composition Analysis Of a Large Outpatient Population Using Deep Learning To Derive Age, Sex, and Race-Specific Reference Curves

K. Magudia, C.P. Bridge, C.P. Bay, A. Babic, F.J. Fintelmann, F. Troschel, N. Miskin, W. Wrobel, L.K. Brais, K.P. Andriole, B.M. Wolpin, and M.H. Rosenthal

Radiology

Article at RSNA

Furthermore, an earlier version of the same model was developed for the following publication:

Fully-Automated Analysis of Body Composition from CT in Cancer Patients Using Convolutional Neural Networks

C.P. Bridge, M. Rosenthal, B. Wright, G. Kotecha, F. Fintelmann, F. Troschel, N. Miskin, K. Desai, W. Wrobel, A. Babic, N. Khalaf, L. Brais, M. Welch, C. Zellers, N. Tenenholtz, M. Michalski, B. Wolpin, and K. Andriole

Workshop on Clinical Image-based Procedures, MICCAI, Granada 2018

Article at Springer Link, Article at Arvix

If you use this code in your publication, please cite these papers.

Acknowledgements

The Python code for body composition estimation was written by Christopher Bridge at MGH & BWH Center for Clinical Data Science (now MGB AI) and the Athinoula A. Martinos Center for Biomedical Imaging, Alex Chowdhury, and Sahil Nalawade both at the Department of Informatics & Analytics at the Dana Farber Cancer Institute. The z-score curve fitting R code in the stats directory was written by Camden Bay at Brigham and Women's Hospital. The project was conceived and led by Michael Rosenthal at the Dana Farber Cancer Institute and Florian Fintelmann at Massaschusetts Department of Radiology with assistance from Kirti Magudia at Brigham and Women's hospital.

  • Chris Bridge, Massachusetts General Hospital (cbridge at mgh dot harvard dot edu)
  • Alex Chowdhury, Dana Farber Cancer Institute (Alexander underscore Chowdhury at dfci dot harvard dot edu)
  • Kirti Magudia, Duke University (kirti dot magudia at duke dot edu)
  • Michael Rosenthal, Dana Farber Cancer Institute (Michael underscore Rosenthal at dfci dot harvard dot edu)
  • Florian Fintelmann, Massachusetts General Hospital (fintelmann at mgh dot harvard dot edu)
  • Camden Bay, Brigham and Women's Hospital (cpbay at bwh dot harvard dot edu)

See Also

The z-score fitting process associated with this work is available here.

Owner

  • Name: Rosenthal-Lab-at-Dana-Farber
  • Login: Rosenthal-Lab-at-Dana-Farber
  • Kind: organization

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "Citation for this repository"
authors:
  - family-names: Bridge
    given-names: Christopher
  - family-names: Chowdhury
    given-names: Alexander
  - family-names: Nalawade
    given-names: Sahil
  - family-names: Umeton
    given-names: Renato
  - family-names: Rosenthal
    given-names: Michael
title: "Population-Scale CT-based Body Composition Analysis of a Large Outpatient Population Using Deep Learning to Derive Age-, Sex-, and Race-specific Reference Curves"
date-released: 2020-11-24
doi: 10.1148/radiol.2020201640
url: https://github.com/Rosenthal-Lab-at-Dana-Farber/ct_body_composition
preferred-citation:
  type: article
  authors:
      - family-names: Magudia
        given-names: Kirti
      - family-names: Bridge
        given-names: Christopher
      - family-names: Bay
        given-names: Camden
      - family-names: Babic
        given-names: Ana
      - family-names: Fintelmann
        given-names: Florian
      - family-names: Troschel
        given-names: Fabian
      - family-names: Nityanand
        given-names: Miskin
      - family-names: Wrobel
        given-names: William
      - family-names: Brais
        given-names: Lauren
      - family-names: Andriole
        given-names: Katherine
      - family-names: Wolpin
        given-names: Brian
      - family-names: Rosenthal
        given-names: Michael
  doi: 10.1148/radiol.2020201640
  journal: "Radiology"
  publisher: Radiological Society of North America
  month: 11
  year: 2020
  issue: 2
  volume: 298
  start: 319
  title: "Fully-Automated Analysis of Body Composition from CT in Cancer Patients Using Convolutional Neural Networks"
  url: https://doi.org/10.1148/radiol.2020201640
title: "Fully-Automated Analysis of Body Composition from CT in Cancer Patients Using Convolutional Neural Networks"
date-released: 2018-08-11
doi: 10.1007/978-3-030-01201-4_22
preferred-citation:
  type: article
  authors:
      - family-names: Bridge
        given-names: Christopher
      - family-names: Rosenthal
        given-names: Michael
      - family-names: Wright
        given-names: Bradley
      - family-names: Kotecha
        given-names: Gopal
      - family-names: Fintelmann
        given-names: Florian
      - family-names: Troschel
        given-names: Fabian
      - family-names: Miskin
        given-names: Nityanand
      - family-names: Desai
        given-names: Khanant
      - family-names: Wrobel
        given-names: William
      - family-names: Babic
        given-names: Ana
      - family-names: Khalaf
        given-names: Natalia
      - family-names: Brais
        given-names: Lauren
      - family-names: Welch
        given-names: Marisa
      - family-names: Zellers
        given-names: Caitlin
      - family-names: Tenenholtz
        given-names: Neil
      - family-names: Michalski
        given-names: Mark
      - family-names: Wolpin
        given-names: Brian
      - family-names: Andriole
        given-names: Katherine
  journal: MICCAI
  publisher: Springer
  month: 08
  year: 2018
  start: 204
  title: "Fully-Automated Analysis of Body Composition from CT in Cancer Patients Using Convolutional Neural Networks"
  url: https://link.springer.com/chapter/10.1007/978-3-030-01201-4_22

GitHub Events

Total
  • Watch event: 2
  • Push event: 2
Last Year
  • Watch event: 2
  • Push event: 2