AVA stenosis positioning and stenosis degree multistage classification system based on MPSF extraction

Abstract

The invention discloses an AVA stenosis positioning and stenosis degree multistage classification system based on MPSF extraction. An acoustic signal collector collects and transmits acoustic signalsof a tester AVA relative to upstream and downstream positions; after the processor receives the acoustic signals of the tester AVA relative to upstream and downstream positions, the following steps are executed: the received acoustic signals are segmented according to a heart rate period to obtain acoustics sub-signals; performing singular spectrum analysis and welch power spectrum estimation on the acoustics sub-signals to obtain frequency spectrum energy of a high-frequency part, and forming a group of MPSF by the frequency spectrum energy of the acoustics signals from the same AVA relativeupstream and downstream positions; and inputting each group of MPSF into an AVA stenosis positioning and stenosis degree multistage classification model constructed based on a long-term and short-termmemory network, and outputting a classification result of the AVA stenosis position and the stenosis degree of the tester through calculation. According to the system, the accuracy of AVA stenosis area positioning and stenosis degree classification is improved.

Description

technical field [0001] The invention belongs to medical classifiers, in particular to a multi-level classification system for AVA stenosis location and stenosis degree based on MPSF extraction. Background technique [0002] Hemodialysis is the most common treatment for kidney replacement in patients with chronic renal failure. In order to achieve blood drainage, a vascular access (VA) is usually surgically created in the patient's forearm. Generally, such a VA is achieved by connecting arteries and veins in the arm. This vascular access is called Autologous fistula (AVF). However, for some patients, their physical condition is not suitable for the construction of AVF, and artificial blood vessels need to be implanted to connect the arteries and veins. Such access is called artificial fistula (AVG). [0003] The detection method of arteriovenous access (AVA) stenosis put into use in the hospital is usually physical examination combined with color Doppler ultrasonography (Du...

AI Technical Summary

Problems solved by technology

[0005] There are many problems in the above acoustic AVA stenosis discrimination methods: (1) The inconsistency in the judgment of characteristic frequency bands reflects that various studies have not found the true relationship between power spectrum and AVA stenosis (Degree Of Stenosis, DOS), and there is serious over-fitting. (2) The acquisition position is single and based on the premise of the existence of stenosis, it is not position-sensitive, and it is not usable in real cases where the presence or absence of stenosis is unknown; (3) The binary classification method cannot respond Severity of AVG stenosis

This method can only locate the stenosis area, but cannot determine the severity of the stenosis

Images