Published March 22, 2026 | Version v0.3.2
Software Open

PaulRitsche/DeepACSA: v0.3.2

Authors/Creators

  • 1. University of Basel - Department of Sport, Exercise and Health
  • 2. University of Michigan

Description

DeepACSA

 

Automatic analysis of human lower limb ultrasonography images

 

DeepACSA is an open-source tool to evaluate the anatomical cross-sectional area of muscles in ultrasound images using deep learning.
More information about the installtion and usage of DeepACSA can be found in the online documentation. You can find information about contributing, issues and bug reports there as well.
If you find this work useful, please remember to cite the corresponding paper, where more information about the model architecture and performance can be found as well.

 

V0.3.2

 

With version 0.3.2, we included new models for the

 

- patella tendon (taken from Guzzi et al. 2026)
- vastus medialis (taken from Tayfur et al. 2025)

 

Below you can find some overview tables. All models and the newest installer can

 

Patella Tendon Models (Unet3+ best performing)

We provide a model for the automatic segmentation of the patellar tendon anatomical cross-sectional area at 25%, 50%, and 75% of tendon length in healthy subjects (UNet 3+).
We evaluated two model architectures for patellar tendon segmentation: UNet-VGG16 and UNet 3+. Their performance was assessed by comparing automated predictions with manual segmentations. Overall, both models demonstrated good agreement with manual analysis, with UNet 3+ showing the most consistent performance. Detailed methodology and results are reported in our publication Guzzi et al. 2026.

 

Vatus medialis model (only VGG16Unet trained)

We provide a model for the automatic segmentation of the vastus medialis cross-sectional area (ACSA) in healthy participants as well participants with ACL injuries.
A UNet-VGG16 model was evaluated and compared to manual analysis. Comparability calculations and detailled methodology can be found at Tayfur et al. 2025.

Files

DeepACSA_v0.3.2_example.zip

Files (2.3 GB)

Name Size
md5:d22ccca74fc3ecc27d4de3c5375738ab
348.6 MB Download
md5:cd3ab55a73392a681050fd4ff7bb2c47
4.1 MB Preview Download
md5:a59194fb883dfa500835a7237dca4b8b
12.6 MB Preview Download
md5:8bc9831d73d9d9ad2873862855415645
362.5 MB Download
md5:2eabf8f19ce8256c15c3f14f29bb99ee
270.7 MB Download
md5:1cbe8774eb6c6f2f0df9ee3c08b38025
362.5 MB Download
md5:54eb943b31a7d6768c876de80d177810
362.5 MB Download
md5:5e295b251f2336ab6d4d64b550e1869c
272.9 MB Download
md5:c8bcefa64d3f5eb45e1dd5707184a71d
272.9 MB Download

Additional details

Related works

Is published in
Publication: https://www.springermedizin.de/ultrasonography-based-patellar-tendon-area-measurement-comparabi/52053032 (URL)
Publication: https://www.sciencedirect.com/science/article/pii/S0301562924004319 (URL)
Publication: https://journals.lww.com/acsm-msse/Fulltext/2022/12000/DeepACSA__Automatic_Segmentation_of.21.aspx (URL)
Is supplement to
Software: https://github.com/PaulRitsche/DeepACSA/tree/v0.3.2 (URL)

Software

Repository URL
https://github.com/PaulRitsche/DeepACSA
Programming language
Python
Development Status
Active