Can smart phones be used in a teleradiology setting for evaluating lung cancer therapy response?

Works to be done daily can be done with mobile technology in more practical, flexible times, in desired environments. The transformation of smart phones and tablets into daily life leads to changes in the personal activities of individuals. With mobile health applications developed, individuals can access personal health services at any time and place they want. Imaging methods are an indispensable tool in modern medicine. In radiology, clinical depression affects radiologists because of working in low working conditions. With this study, it is aimed to increase productivity for radiologists by the use of some techniques to improve the performance of mobile displays at normal conditions. It was chosen a mobile dicom viewer which is called Medfilm in the Iphone 6S model smartphone. As a DICOM viewer in the computer, Radiant Dicom Viewer Version 4.1.16 was executed to examine the images in the Viewsonic VA2410-mh model monitor. For case analysis, nodules in lung CT images were examined. Data was taken from TCIA collection. TCIA is a service which de-identifies and hosts a large archive of medical images of cancer accessible for public download. 271 cases were included in the study. In both the observer 1 and the observer 2, the nodule was detected in 266 cases (98.2%) and no nodules were found in 5 cases (1.8%). In the evaluation made by the telephone, the first observer detected nodule in 251 cases (92.6%), while the second observer detected nodules in 252 (93%) patients. It was used Bland-Altman analysis, which is the most accurate method used to describe the harmony between the two measurements. According to Bland-Altman test with 95 % confidence interval, the difference between observer 1 and 2 were calculated as 0,6740,39 mm and 0,660,387 mm respectively. If the differences are normally distributed, we would expect 95% of the nodule measurement differences to lie between -0.095 mm and 1.443.It can be said that nearly all pairs of measurements by the two methods will be closer together than these extreme values, which we call 95% limits of agreement. The difference between observer 1 and observer 2 for smartphone and monitor measurement were calculated as -0,190,662 mm and -0,1620,55 mm respectively. If the differences are normally distributed, we would expect 95% of the nodule measurement differences to lie between -1.5 mm and 1.12 mm for observer 1 ; -1.25 mm and 1.12 mm for observer 2. The encouraging results show that smart phones can potentially be used in tele-radiology for evaluating therapy response in lung cancer during clinical trials or actual therapy for large tumors.