Analysis of thermal camera measurements in patient with acute cholecystitis

The aim of this study is to analyze the infrared (IR) (thermal camera) images of patients admitted to our hospital with abdominal pain and diagnosed with acute cholecystitis and to evaluate their availability in diagnosis. Patients admitted to our hospital with abdominal pain, diagnosed with acute cholecystitis and healthy volunteers were included in the study. In the patient group, after hospitalization, thermal images including four quadrants of the abdomen were taken. The same procedures were repeated for the volunteers. As a result of statistical analysis, a significant difference was found between right upper quadrant temperature in the patient and control groups (p <0.001). First, a machine learning model was created with the community classifier vector with all thermal analysis data of 40 people both from the patient and control groups. Then, we uploaded images of 10 patients and 10 controls that the machine hadn't seen before, and examined how it distinguish patients and healthy ones, and found the accuracy, sensitivity and specificity rates. Then we made the same application with statistically significant parameters with a linear distinguishing vector and examined the results again. When the thermal camera analysis data of the right and left upper quadrant temperatures were compared statistically, it was found to be statistically significant in the patient and control groups (p <0.001). By using machine learning method with artificial intelligence, we found that artificial intelligence has 95% accuracy, 100% sensitivity and 90% specificity in terms of separating the patient and control group. It was concluded that our study, in which thermal camera images were analyzed in patients diagnosed with acute cholecystitis patients and normal people with high success by artificial intelligence and machine learning, could be one of the auxiliary methods in the diagnosis of Acute Cholecystitis, and it was concluded that this method could be developed with studies with broader participation.