Özet:

Epidural hematom (EAH) is the accumulation of blood in the potential gap between the external membrane of the brain (dura mater) and the bone. Acute subdural and epidural hematom appears as a hyper-intensive collection that is often found in the brain convexite in the IT scan. These types of bleeding can become fatal by increasing the inherent pressure and creating mass effects. Therefore, it is very important to immediately recognize these bleeding in an emergency trauma environment. Therefore, early diagnosis is required to reduce mortality and morbidity in these cases. There has been a growing interest in recent times in artificial intelligence (AI) and machine learning (ML) algorithms for diagnosis in medical fields. In this study, a controlled learning method was used in which the decision tree ML algorithm was trained with the patient’s conditions (EAH or Normal). This study suggests a Early Warning System (EWS) that scans all the chronic ITs obtained in the trauma center. In this study, EWS is trained with CT scans taken from about 100 patients, able to predict EAH with 100% accuracy with image recognition and controlled learning algorithms. Each MR cut obtained for each patient is passed by one-on-one image processing analysis and EAH detection is made. The decision tree method, which is a controlled learning algorithm for this, has been trained and used to determine EAH in the MR cuts. The algorithm has been developed in a way that, when EAH is detected in more than 10 patients, the emergency physician and the consultant brain surgeon will be immediately warned by e-mail.

Anahtar Kelimeler: