Method for monitoring intracranial pressure based on convolutional neural network algorithm

Abstract

The invention discloses a method for monitoring intracranial pressure based on a convolutional neural network algorithm. The method for monitoring the intracranial pressure based on the convolutional neural network algorithm comprises S1: detecting the actual intracranial pressure value by means of an FVEP noninvasive intracranial pressure monitoring technology, and forming an FVEP waveform; S2: constructing a convolutional neural network, taking the intracranial pressure value actually measured in the step S1 as the result, constantly training the convolutional neural network to learn actual FVEP waveforms corresponding to different intracranial pressure values, comprehensively considering various external factors which affect the intracranial pressure, and serving the external factors as input of the convolutional neural network system in common; and S3: carrying out matching, identification and analysis on the FVEP waveform graph by means of the convolutional neural network, and achieving effective identification of the FVEP waveform. According to the invention, the convolutional neural network algorithm is employed to carry out pattern identification on the measured FVEP waveform, and continuously learn and optimize, and effective intracranial pressure prediction is achieved.

Description

technical field [0001] The invention belongs to the technical field of neurosurgery, and in particular relates to a method for monitoring intracranial pressure based on a convolutional neural network algorithm. Background technique [0002] The pressure generated by the contents of the cranial cavity on the wall of the cranial cavity is the intracranial pressure (Intracranial Pressure) ICP. Since the cerebrospinal fluid in the brain is between the wall of the cranial cavity and the brain tissue, the hydrostatic pressure of the cerebrospinal fluid is generally used to represent the intracranial pressure. It is 0.7-2.0kPa, and the normal intracranial pressure of children is 0.5-1.0kPa. [0003] Increased intracranial pressure is a common clinicopathological syndrome in neurosurgery, and it is a common symptom of craniocerebral injury, brain tumor, cerebral hemorrhage, hydrocephalus, and intracranial inflammation. Sustained above 2.0kPa, which caused the corresponding syndrom...

AI Technical Summary

Problems solved by technology

Identifying N2 waves from a series of positive and negative waves is the key to FVEP monitoring of intracranial pressure, but so far, there is no unified standard for the identification of N2 waves and the judgment of N2 wave latency

Inconsistent identification of N2 waves may lead to different results for the same patient measured by different physicians

This is detrimental to the accuracy of the measurement

[0010] 2) The intracranial pressure value changes with the conditions (cause, individual), and the accuracy, universality, and simplification of the FVEP method identification model cannot be taken into account, and because there are many influencing parameters and the relationship is complicated, the relationship between FVEP and intracranial pressure is a complicated one. Due to the nonlinear system, a single identification model cannot accurately match all situations, taking into account various factors, and it is impossible to use a function to give the exact relationship between various waves of FVEP and intracranial pressure, as well as the patient's age, physiological conditions, etc.

This limits the accuracy of FVEP

For example, clinical studies have shown that the incubation period of FVEP is different in different age groups. For example, different diseases have different effects on the metabolism of nerve cells. For neurological diseases such as hypoxic encephalopathy and encephalitis, it will also lead to N2 waves. Prolonged incubation period, leading to inaccurate judgment

Therefore, for different patients and different etiologies, the same regression function is used for the corresponding relationship between FVEP curve and intracranial pressure, and there will inevitably be errors

Images