eLearning
We aimed to develop an integrated eLearning package to facilitate student learning of radiographic assessment of lines and tubes and their potential complications.
The eLearning package consists of micromodules (videos, minigames, integrated mobile apps, and trained AI deep learning models).
Artificial Intelligence
We incorporated a deep learning model trained and validated for lines and tubes detection and classification. The tube is segmented and then classified into abnormal, indeterminate (incomplete), or normal. Attention maps are generated after each attempt, and is used to evaluate the accuracy of the students' answer.
We have also developed a separate chatbot app with 4 case simulation games in addition to the mobile game app to further bridge the gap between theory and practice in radiology. The app features clinical cases based on common medical scenarios involving lines and tubes, with associated questions on image interpretation and patient management. Students can respond to the questions using natural text, and the chatbot will give a structured evaluation of the student’s performance and ask personalized follow-up questions on the student’s areas of improvement.
Please stay tuned for more information.