Malnutrition in gastric cancer

In this retrospective study of 312 gastric cancer patients, researchers developed and tested a technique based on deep learning to assess the risk of malnutrition prior to surgery. Clinical factors and characteristics of the psoas muscle at the level of the third lumbar vertebra were taken into account. Analyses showed that BMI, lymphocytes and albumin were clinically independent of malnutrition risk. The model developed by the researchers proved relevant for assessing the risk of malnutrition prior to surgery in gastric cancer. It allowed patients to be stratified from low to high risk. Overall survival time was lower in the high-risk group than in the low-risk group.