Cerebral apoplexy risk screening method based on abnormal sample detection and multi-dimensional information output

Abstract

The invention belongs to the technical field of cerebral apoplexy risk screening, and particularly relates to a cerebral apoplexy risk screening method based on abnormal sample detection and multi-dimensional information output, and the method comprises the steps: constructing a cerebral apoplexy risk factor data set of a screening object, inputting the data set into a trained cerebral apoplexy risk screening prediction evaluation model, acquiring risk factor importance indexes and cerebral apoplexy risk grade assessment results, and marking data of abnormal assessment results; wherein the cerebral apoplexy risk screening prediction evaluation model comprises a risk factor importance detection evaluation model and a cerebral apoplexy risk level prediction evaluation model; Before risk level classification prediction evaluation is carried out, an Isolation Forest algorithm is adopted to check and eliminate abnormal samples in the screening data, so that the influence of the abnormal samples on a prediction classification evaluation link is effectively reduced, and accuracy of prediction evaluation is effectively improved.

Description

technical field [0001] The invention belongs to the technical field of stroke risk screening, and in particular relates to a stroke risk screening method based on abnormal sample detection and multidimensional information output. Background technique [0002] Stroke has become a major disease that threatens the health of residents all over the world, and it is a major public health problem that countries all over the world need to solve urgently. Stroke has the characteristics of high morbidity, high mortality and high recurrence rate, which seriously affects the quality of life and health of human beings. According to incomplete statistics, it is estimated that the incidence of stroke patients in China is about 13 million, and the stroke mortality rate accounts for about 30% of the global total. Therefore, how to screen high-risk groups for stroke and focus on high-risk groups according to the screening results Monitoring and detection is a method to reduce the mortality r...

AI Technical Summary

Problems solved by technology

At present, there are problems such as poor recognition accuracy and insufficient comprehensive recognition of stroke risk level in the method of screening stroke based on electronic medical records. Occasional errors, false positives or missed negatives of the collected personnel, and strong subjectivity of evaluation indicators have led to a large number of abnormal samples in the screening data set. The existence of abnormal samples has caused large deviations in the model training process and overall testing. low accuracy

Images