Skip to main content
Fig. 4 | BMC Medicine

Fig. 4

From: Interpretable artificial intelligence-based app assists inexperienced radiologists in diagnosing biliary atresia from sonographic gallbladder images

Fig. 4

Example of a case where radiologists obtained the correct diagnosis with the assistance of the new smartphone app. a An original sonographic gallbladder image selected from the video of a 65-day-old female infant with biliary atresia. Seven radiologists considered the gallbladder to be normal. However, the diagnosis provided by the model obtained from the photos taken by six of these seven radiologists indicated biliary atresia, and the heatmaps all focused on the gallbladder. Finally, five radiologists revised the diagnosis of this infant to biliary atresia. b The smartphone photo taken by radiologist J. c The output interface of the test result in the new smartphone app for smartphone photo shown in b, displaying a probability value of 99.82%. With the assistance of the new smartphone app, radiologist J finally revised the diagnosis of this infant to biliary atresia

Back to article page