hermionewarr/Transfer_learning_sarcopenia_C3

A machine learning model was trained, on a cohort of head and neck cancer patients, to segment skeletal muscle in the neck at the level of the C3 vertebra. The model was a convolutional neural network and had a Dice score of 0.89 (89% accuracy). A value of skeletal muscle area (SMA) and density (SMD) was found for each patient in a cohort of around a hundred head and neck squamous cell carcinoma patients in the Cancer Imaging Archive data collection. The median value of SMD was 26 HU in women and 33 HU in men. The median value of SMA was 38 cm^2 and 25 cm^2 for men and women respectively.

Jupyter Notebook