Research And Implementation Of The Pre-hospital Emergency Treatment Remote Assist System

In modern medicine, when a patient has a sudden illness and can’t be moved, he has to call for help so the ambulance will be sent to the patient to carry him back to the hospital. This situation is particularly important because of the condition of the patients, the relative limitation of treatment conditions. In the first aid situation, some illness cannot be treated the first time due to the lack of knowledge of nurses and the lack of equipment. Therefore, remote help from the doctors in the hospital is needed. By observing the situation of the patient and the query of the past illness history of the patient, the doctor can give some suggestions for the nurses and be well prepared for the treatment after the patient arrives the hospital.This project is a model project of the project started by a hospital in Guangdong province which aimed at the research and building of the system of pre-hospital care. And this project mainly aimed at implement an efficient remote assist system for the ambulance; it enables doctors to assist the nurse for the treatment through remote connection. And by using image classifier, it makes the images taken from the patient more optimized. And it makes the remote help more efficient by building the record system and the management system.In this paper, the main work is divided into the following sections: The first one does image characteristics extraction by using the color histogram and the improved HOG descriptors, and by using SVM, it implements the medical image classification. The second is the implementation of the remote assist system using the image classification and processing, and the elimination of the adverse effects for the images in the situation. The third is the hospital ambulance dispatch and management visualization.In the medicine image learning part, the hand, feet, head, electrocardiogram and oral images are collected from hospitals. By using HOG and the color histogram, the system get the image’s descriptors. The system firstly classifies the electrocardiogram and other images using color histogram. And then use the HOG and SVM to classify the other images. Since HOG is not rotation invariant, this system has to improve the HOG for the process of some images. The second part is the implementation of the remote assist system. In this part, the doctor observes the patient through this system, and gets the images of the patient. The system classifies the images into oral image, ECG image, and skin image, and uses different methods to process these images to help the doctor get a better understanding of the illness. In the meantime the doctor can query the record database and get more information of the patient. In the third management part, it implements the Web interface to manage the information of the ambulances, doctors and the nurses and implements illness record collection and information extraction for the second part.This system uses image classification, image optimization and information management into the ambulance first aid situation. The management system makes the management easier and more efficient. The assist system implements the function of remote assist and record query. This system makes the remote first aid much more efficient.