Bedridden patient pain symptom real-time monitoring method and device based on expression recognition

Abstract

The invention discloses a bedridden patient pain symptom real-time monitoring method and device based on expression recognition. The method comprises the following steps: 1, establishing a pain expression training data set; 2, establishing a neural network model for analyzing pain expressions, and performing training to obtain a pain grading model; and 3, acquiring three real-time images of a human face at a same moment, preprocessing the images, and inputting the preprocessed images into the neural network model to obtain probabilities corresponding to A pain levels, and selecting a pain level corresponding to the maximum probability as the pain level of the detection images at the current moment, and performing alarm processing on the pain level exceeding a threshold so as to realize real-time monitoring. According to the invention, pain can be accurately evaluated in real time and automatically, so that effective monitoring of a bedridden patient is realized.

Description

technical field [0001] The invention relates to the fields of image recognition and medical technology, in particular to a method and device for real-time monitoring of a bedridden patient based on facial expression recognition. Background technique [0002] Related technologies in facial expression recognition have been widely used in security, entertainment, finance and other fields, and play an important role. Its application method and application scope are still improving and expanding, and it has broad development prospects. [0003] In medicine, how to realize the assessment of the patient's pain degree has always been an important issue. Pain is defined as a subjective feeling in medicine, and patients' self-assessment has high reference value. However, some special patients, such as patients who have lost active consciousness, high-risk or disabled patients, are often unable to describe their own pain level in words. At this time, manual assessment by medical obse...

AI Technical Summary

Problems solved by technology

However, this manual evaluation method has many disadvantages: such as low efficiency, limited persistence, and the influence of observer's subjectivity, etc.

There are few feature points selected on the face, and the division of facial conditions is relatively general, so it is difficult to accurately detect the pain degree of the patient's own subjective feeling

[0006] There are also some new methods, such as evaluating data such as the grip strength of the sensor when the patient is in pain, which requires the patient to be conscious and actively participate in the evaluation, which is greatly affected by physical conditions and individual differences

Images