Brain disease judgment system based on intracranial multi-modal information fusion of machine learning

Abstract

The invention provides a brain disease judgment system based on intracranial multi-modal information fusion of machine learning. The judgment system comprises an input layer and an output layer, wherein intracranial information is input into the input layer, multi-level processing of data is carried out, various intracranial injury types are automatically judged and output. An intervention treatment combination is specified according to the type combination of the output layer, wherein the intracranial information comprises intracranial pressure, intracranial oxygen partial pressure, intracranial temperature, intracranial electroneurographic signals, intracranial sodium ion concentration and intracranial potassium ion concentration.

Description

technical field [0001] The invention relates to the technical field of physiological signal monitoring, in particular to a brain disease judging system based on machine learning-based multimodal information fusion of the brain. Background technique [0002] Traumatic brain injury is a global health problem with a high mortality rate. In China, about 600,000 people suffer from traumatic brain injury every year, of which about 100,000 people die, causing direct and indirect economic losses of more than 10 billion yuan. Preliminary statistics from the China Craniocerebral Trauma Database show that among more than 13,000 inpatients with acute craniocerebral trauma in 47 hospitals in China, the fatality rate of severe craniocerebral trauma is >20%, and the severe disability rate is >50%. Foreign data show that for patients with severe traumatic brain injury with Glasgow Coma Scale (Glasgow Coma Scale, GCS) ≤ 8 points, the mortality rate is as high as 35% to 45%. [0003] T...

AI Technical Summary

Problems solved by technology

[0012] To summarize the above statements, the current monitoring of intracranial physiological indicators mostly uses implantable sensor probes, which bring about problems such as large wound area, susceptibility to infection, and inconvenient wire connection, which significantly limit its clinical application.

At present, there is no highly integrated instrument and method at home and abroad that can monitor multi-modal intracranial cerebral cortex signals simultaneously in real time.

Images