Artificial intelligence kidney tumor prediction system based on knowledge graph

Abstract

The invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence kidney tumor prediction system based on a knowledge graph. The system comprises a basic number library unit, a training learning unit, an auxiliary diagnosis unit and a function application unit. The basic number library unit is used for processing and analyzing massive instance data and constructing a knowledge base; the training learning unit is used for building a knowledge graph and a training model and carrying out deep learning on the training model; the auxiliary diagnosis unit is used for identifying and predicting the tumor examination image; and the function application unit is used for improving the functionality of the system by adding a plurality of application services. According to the method, a full-spectrum kidney tumor association diagram is established based on a knowledge graph and an artificial intelligence algorithm, tumor classification and feature extraction are performed according to the knowledge graph, and tumor positions are detected and labeled for assisting image diagnosis, so that tumor prediction can be effective and accurate, clinical diagnosis and treatment decisions are guided, unnecessary operations are avoided, and the patient is benefited.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular, to an artificial intelligence kidney tumor prediction system based on a knowledge map. Background technique [0002] Renal cell carcinoma (RCC) accounts for 85% of renal malignancies, with 65,000 new cases worldwide each year. In the absence of histological diagnosis, renal tumors diagnosed as RCC by imaging have reached the indications for surgical resection, which has led to the overdiagnosis and treatment of many benign renal tumors. Approximately 20% of surgically resected renal tumors have been reported to be benign postoperatively, challenging the necessity of surgery for all suspicious lesions as patients are exposed to both risks and morbidity. Noninvasive preoperative imaging techniques, such as ultrasound, CT, and MRI, are widely used for the characterization of renal tumors. However, CT and MRI have limited sensitivity and specificity in differe...

AI Technical Summary

Problems solved by technology

However, in the application of knowledge graphs, most of them are used for simple association analysis, and the scope of use is limited. It has not been included in the Bayesian network directed graph model and the Markov random field undirected graph model, namely The breadth and depth of existing knowledge graphs need to be improved

In addition, there is currently no artificial intelligence kidney tumor prediction system and detection method based on knowledge graphs

Images