This project aims to develop an automated screening platform for early identification of malnutrition in cancer patients by creating a machine-learning-based model that combines CT imaging input and questionnaire-based tools. The proposed solution will incorporate novel biomedical image segmentation tools based on artificial intelligence to assess body composition from CT scans. Accurate skeletal muscle and adipose tissue segmentation will be used with current questionnaire-based parameters such as BMI changes and food aversion to develop thresholds for detecting malnutrition.