QUMIA

Quantitative Ultrasound Muscle Imaging Analysis

Photo credit – Shutterstock

There are over 600 neuromuscular diseases affecting nerves and muscles in children and adults. Neuromuscular diseases severely impact functioning and quality of life. Muscle ultrasound is an important screening tool for neuromuscular diseases. Muscle ultrasound has the advantages that it is painless and easy to perform. However, the current technological approach to muscle ultrasound is time-inefficient and has a suboptimal performance. In the current project, we developed software based on novel deep learning techniques to improve speed and performance of muscle ultrasound image analysis. We found that our software is fast and has a high performance to predict whether a patient with a suspected neuromuscular disease truly has a neuromuscular disease.

We expect that our new software will accelerate use and value of muscle ultrasound in research and the clinical setting. Therefore, we call other scientific and clinical groups working on muscle ultrasound to join our efforts to further develop and test our software. An important next step is to incorporate our approach in clinical practice, thereby improving care for patients with a neuromuscular disease.

Have a look at our software on github: https://github.com/QUMIA

Participating organisations

Netherlands eScience Center
Radboud University Medical Center
Life Sciences
Life Sciences

Team

JD
Jonne Doorduin
Peter Kok
Peter Kok
Lead RSE
Netherlands eScience Center
Sonja Georgievska
Sonja Georgievska
Candace Makeda  Moore
Candace Makeda Moore

Related software

qumia-train-scripts

QU

Scripts and files related to ML training for the QUMIA project

Updated 1 month ago
1 3