Design and Implementation of Mathematical Medical Model to Detect Benign and Malignant Lung Tumors

This article belongs to the field of medical digital image processing technology, specifically a lung malignant tumor and benign nodule detection system based on PET/CT image texture features. The main function of this article is to find useful texture features in PET/CT images to better distinguish between lung dense tissue and benign tissue. Firstly, the ROI of the region of interest is segmented in the PET/CT image, and then five texture features of the ROI are extracted: roughness, contrast, busyness, complexity, and intensity. Then, distance calculation and feature classifier are used to classify and distinguish features, and a combination of multiple features is used to effectively distinguish malignant tumors and benign nodules.With the popularization of people's awareness of prevention and screening, and the development of imaging examination, the discovery of lung tumors has also increased. The accurate differentiation of benign and malignant lung tumors is critical for timely and appropriate clinical intervention. At present, the accuracy of the benign and malignant evaluation of lung tumors depends to a great extent on the doctors' personal clinical experience and diagnosis and treatment level. There are inevitably problems such as missed diagnosis and excessive intervention. Therefore, the mathematical prediction model came into being. The quantitative malignant probability value can more objectively guide the follow-up, diagnosis and treatment of lung tumors, and overcome the shortcomings of empirical medicine. According to the clinical characteristics and some imaging characteristics of the collected patients with lung tumors, the independent risk factors of lung malignant tumors were analyzed, and the accuracy of two foreign mathematical models in predicting benign and malignant tumors was evaluated and compared, further demonstrating its clinical application value in Chinese people.