| At present,patients with functional gastrointestinal diseases are very common in the population,and the number of patients with irritable bowel syndrome and functionald yspepsia accounts for 16% to 26% of the global population.Gastroenterology clinicians have observed that patients with functional gastrointestinal diseases are usually accompanied by anxiety and depression,and after targeted medication,they can not only improve the disease of patients with functional gastrointestinal diseases,but also enhance This type of patient’s compliance with medication.Appropriately taking symptomatic mood-relief drugs is beneficial to the physical and mental health of patients of this type.However,drugs for the treatment of anxiety and depression are often controlled drugs.Gastroenterologists relatively lack the experience and knowledge of psychopsych ological diagnosis.Therefore,the diagnosis of symptoms of such mental state depends on personal experience in the interrogation process.Moreover,the accurate evaluation of the patient’s mental state by professional psychologists takes a long time and careful interview evaluation,which has defects such as long time-consuming and cumbersom e medical procedures,which is not suitable for general inspection.According to the research of scientists,there is a certain correlation between psycho logical abnormalities and abnormal gastrointestinal tract function.Most of the performance of people’s mental and emotional states is expressed through facial expressions.Combined with some work that the scientific community has already done in facial expression recognition,this paper has designed a low-disturbance evaluation system for the symptoms of "anxiety disorder" in patients with functional gastrointestinal disease.With the help of a webcam,it is in the outpatient clinic with low interference Help the digestive physician to evaluate the patient’s "anxiety disorder" symptoms,save time,and provide a third-party objective evaluation results for the physician and the patient.The experimental scheme of this paper is designed to obtain the video of the face area of the patient during the outpatient consultation through the camera,and the video of the face area and the evaluation score of the scale when the evaluation of the "Hamilton Anxiety Scale" is completed.Feature extraction was performed on the video of the two scenes,respectively,to retain the "anxiety disorder" symptom information contained in the subject’s face.One method of feature extraction is to use facial expression recognition algorithm to read and calculate the probability of seven basic emotions frame by frame,and then obtain a set of 7-dimensional feature space after normalization.The other is to use the facial motion unit recognition algorithm to identify the occurrence probabilities of 35 facial motion units in the video sequence,and obtain a 35-dimensional feature space set after normalization.Finally,the support vector machine was used to train the binary classification model.The test results using the test set were: facial expression recognition algorithm: accuracy 55%,precision rate 44.44%,recall rate 50%;action unit recognition algorithm: accurate The rate is 84.38%,the accuracy rate is 100%,and the recall rate is 54.55%.Comparing the two results,the action unit recognition algorithm performs slightly better.According to the doctor’s more commonly used Hamilton anxiety scale score of more than 7 in clinical diagnosis,it is necessary to treat anxiety symptoms.According to the new standard,the subjects are divided into three types: no anxiety,possible anxiety,and anxiety,using facial action unit recognition Algorithm,the accuracy of the test set test results is 93.75%,the accuracy rate is 92.86%,and the recall rate is 100%.The performance of the classifier has been greatly improved.The "anxiety disorder" symptom evaluation system of functional gastrointestinal disease patients developed by us is deployed in the outpatient room.At the same time as the doctor’s consultation,the patient’s "anxiety disorder" symptom evaluation is referenced in parallel for doctor’s diagnosis and treatment process. |