Method for grouping disease diagnosis on basis of decision tree algorithms

Abstract

The invention provides a method for grouping disease diagnosis on the basis of decision tree algorithms. The method includes acquiring basic data of disease kinds; dividing the basic data of the disease kinds into a plurality of disease diagnosis grouping anatomy main categories according to human anatomy systems; further dividing each disease diagnosis grouping anatomy main category into a plurality of disease diagnosis operation child categories according to preset surgical types and operational difficulty degree data; selecting factors for each disease diagnosis operation child category by the aid of the decision tree algorithms and sub-dividing case samples into diagnosis related groups; determining that first complications in segmental characteristic variables selected by decision trees are general complications, determining that second complications in the segmental characteristic variables selected by the decision trees are important complications and dividing the diagnosis related groups on the basis of the general complications and the important complications. Preset conditions are met by the factors selected by the aid of the decision tree algorithms. Average values of expense of segmented nodes of the first complications are lower than preset thresholds, and average values of expense of segmented nodes of the second complications are higher than the preset thresholds. The method has the advantage that cases can be automatically grouped on the basis of clinical case data by the aid of medical diagnosis and the decision tree algorithms.

Description

technical field [0001] The invention relates to the field of computer application technology, in particular to a method for realizing disease diagnosis grouping based on a decision tree algorithm. Background technique [0002] The calculation of medical insurance payment is an important part of the medical security system, which is related to the interests of all parties in the medical insurance. The current medical expenses are increasing rapidly. In order to prevent medical insurance from making ends meet, it is necessary to control medical insurance expenses. [0003] Traditional methods of medical insurance cost control include total prepayment and project-based payment. "Pay by item" is reimbursed according to the fee bill according to the cost of the service received by the patient in the hospital. The disadvantage is that it cannot restrain the medical behavior of the hospital, which can easily lead to over-service, and the hospital lacks cost control awareness, ten...

AI Technical Summary

Problems solved by technology

Although the excessive growth of medical expenses has been controlled on the whole, the disadvantage is that it is

Images