Recognition method for infant pain level based on crying

Abstract

A recognition method for the infant pain level based on crying. The method includes the following steps that 1, crying data of an infant feeling pain is collected and pain level labeling is conducted;2, effective crying fragments are intercepted, crying data is pretreated, and a continuous Y-frame audio data set is obtained; 3, 20 acoustic features are extracted from effective crying fragments toform 20-dimensional feature vectors; 4, based on the 20-dimensional features extracted in the step 3, support vector machine classifiers are trained, the parameter of the classifier which gives the optimal result is determined; 5, the feature selection algorithm is used for dimensionality reduction of a multi-dimensional feature set, and an n-dimensional optimal feature vector combination is selected; 6, the crying data of the infant is recorded, the effective fragments are intercepted, the n-dimensional feature vectors selected from step 5 are extracted and input into the classifiers trainedin the step 4, and the pain grade result is predicted. The method is suitable for evaluating the pain level of the infant in an neonatal ward according to their crying, and provides a reference for clinical doctors and nurses to take comfort measures.

Description

technical field [0001] The invention relates to a method for classifying and identifying infant pain, in particular to a method for identifying infant pain levels based on crying. Background technique [0002] Pain management during infant medical care is very important. Frequent pain-inducing operations in the neonatal ward may even lead to a series of short-term or long-term adverse effects such as developmental delay and permanent damage to the central nervous system. Therefore, good pain management is very important for a baby's future healthy growth and development. However, for infants who cannot complain of pain, it can only be judged by medical staff through observation and evaluation. [0003] Many physiologists believe that pain can be assessed by behavioral manifestations such as facial expressions, cries of pain, heart rate, breathing rate, blood pressure, nerve tones, and limb movements. At present, nurses mainly use the commonly used neonatal facial coding sy...

AI Technical Summary

Problems solved by technology

However, this evaluation method requires a lot of manpower and material resources, and it is difficult to guarantee real-time performance. It is also greatly affected by personal subjective factors, and cannot objectively reflect the degree of pain of the baby.

At the same time, the scores of the scale tools cannot be intuitively correlated with the comfort measures to be taken, and cannot provide direct reference suggestions for clinical pain management

Images