royba84 / detection-of-pneumonia-from-x-ray-images Goto Github PK

Pneumonia is an infectious disease of the lungs that mainly affects the pulmonary vasculature and causes the oxygen not to pass into the blood. Symptoms usually include a combination of cough, chest pain, high fever and difficulty breathing. Pneumonia is most often caused by a bacterial or viral infection and sometimes by autoimmune diseases. Diagnosis of inflammation is based on performing an X-ray of the chest and blood tests. During this period of the corona virus (Covid-19) whose patients are in a higher risk group to develop pneumonia the importance of accuracy and decoding time of the test increases. In light of this period there is a high increase in patients and therefore a quick and accurate solution is needed to decipher the X-rays. This project presents a solution for deciphering photographs using Deep Learning technology. This solution brings results almost in real time and with an accuracy of over 90% and thus will be used to reduce the load of radiologists who decode a lot of photographs of patients. The mechanism will "learn" from a pre-existing X-ray database of healthy and sick people (the training set), using the CNN (Convolutional Neural Network) algorithm we "taught". He will know how to classify new x-rays that he has never encountered (test set) and determine if there is a presence of pneumonia in the photo.

License: GNU General Public License v3.0