Medical examination data identification and analysis method based on NRS-LDA

Abstract

The invention relates to a medical examination data identification and analysis method based on NRS-LDA. The method comprises the following steps: (1) establishing a data acquisition system to acquirerequired original medical examination data; (2) eliminating redundant information by using NRS; (3) carrying out feature extraction work on the reduced data by adopting LDA; (4) dividing the processed data set into a training set and a test set in proportion, and establishing an RBF neural network recognition model based on K-means clustering on the training set; and (5) taking the data on the test set as the input of the RBF neural network identification model, and testing the classification performance of the RBF identification model. According to the method, the NRS-LDA is combined with the RBF neural network identification model to perform medical auxiliary diagnosis, so that the cost is saved, the accurate identification of medical examination data is realized, and the method has great research significance and practical value for medical diagnosis.

Description

technical field [0001] The invention relates to the medical and health field, in particular to an NRS-LDA-based medical examination data identification and analysis method. Background technique [0002] Traditional medical diagnosis relies too much on the clinical experience and subjective judgment of doctors, which is prone to misdiagnosis and missed diagnosis. The content of medical examination data is extremely rich, including many important attributes, such as blood oxygen parameters, heart rate, weight, height, etc., but it still has many useless attributes. Medical examination data itself has some characteristics, such as data pattern polymorphism and data redundancy, etc., and the pattern polymorphism and redundancy of this data increase the space of data storage, increase the cost, and increase the The difficulty of medical diagnosis and disease prediction will also make the diagnosis cycle longer. How to save medical costs, reduce the difficulty of medical diagnos...

AI Technical Summary

Problems solved by technology

Data mining includes steps such as data cleaning and dimensionality reduction. At present, methods for dimensionality reduction of medical examination data mainly include algorithms such as artificial neural networks and factor analysis. However, the training time of artificial neural networks is too long and prone to overfitting and underfitting. Fitting and other issues, factor analysis may sometimes fail when the least squares method is used to calculate the score

Images