營養狀態評估方法及營養狀態評估系統

Method And System For Nutrition Status Evaluation

林宇旌醫師、郭昶甫醫師、馬誠佑博士
Email: caic@cgmh.org.tw (林口長庚醫院產學合作中心)

​

The present disclosure provides an automatic method for predicting prognostic nutrition index（PNI）based on abdominal computer tomographic（CT﹚ images. The laboratory blood test is the gold standard for assessing the nutrition of cancer patients, but it is painful. The present disclosure avoids invasive blood tests for cancer patients by predicting the status of the PNI using machine-learning techniques. The above method automatically segmented fat and muscle area from abdominal CT images and combines height, age, and sex information to predict the status of the PNI.