Lines and Tubes eLearning Materials

Start Learning
Understand the radiographic assessment of correct placement and potential complications of commonly used lines and tubes.
eLearning

Scope

We have developed a series of  eLearning materials to facilitate student learning of radiographic assessment of lines and tubes and their potential complications.In this project, lines and tubes covered include: percutaneous central venous catheters, Port-a-Caths, Hickman lines, peripherally inserted central venous catheters (PICC lines), endotracheal tubes(ETT), amongst others.
Video Component
The animated videos with voiceover cover relevant anatomy for lines and tubes radiographic assessments, a step-by-step walkthrough of the radiological interpretation of line positioning that highlights the important aspects, potential pitfalls, and complications in each step of the assessment. The video content suits students of varying degrees of proficiency on the topic and supplements conventional lectures by acting as bite-sized refreshers on the topic.

Radiology Labyrinth - A 2D mobile maze game

Students enter the game as a miniaturized radiologist navigating a doctor’s office maze. To“return to normal”, they must solve different radiological image interpretation challenges.
Minigame 1
Minigame 2
Minigame 3
Minigame 4
Minigame 5
Minigame 6
Different mini games include:
1
Labeling key anatomical landmarks for lines and tubes interpretation on chest X-rays
2
A flipping card game to associate the purposes and appearances of different lines and tube type
3
Identifying the type of line or tube used in a given CXR
4
Recognizing normal/abnormal positioning of lines and tubes and their tips to teach accurate image interpretation
5
Recognizing normal/abnormal positioning of lines and tubes and their tips with answers derived from the AI attention heat map of a deep learning model
6
Multiple-choice questions on the theory (indications, radiographic positions) of endotracheal tubes and central venous catheters.
A reward system enables students to earn points based on their performance and purchase“power-ups” to assist their progression. Students’ scores from each mini game are recorded and analyzed, facilitating efficient evaluation of learning performance.

XRLiA - X-ray Line Assistant Large Language Model (LLM) Chatbot

Details of the development of the Chatbot can be found here: XRLiA
Models
AI Model Performance
Central Venous Catheter (CVC)
Endotracheal Tube (ETT)
We have developed deep learning AI models trained on chest X-ray images focused on2 stages: (i) segmentation and (ii) classification of lines and tubes. A segmentation model generates masks for any lines and tubes present on a CXR image, which will then be fed to the classification model for identification of the type of line and the position of the tip.
Performance results for central venous catheter (CVC) and endotracheal tube (ETT) tip localisation:
CVC ETT
AUC Accuracy AUC Accuracy
0.902 ± 0.008 0.920 ± 0.004 0.959 ± 0.022 0.991 ± 0.001
A5 - fold cross-validation of the model was conducted using 5,867 images. The results showed that the classification model can accurately identify the position of the  tip as normal or abnormal, with an area-under-curve (AUC) of 0.902 for central venous catheters and 0.959 for endotracheal tubes.
Models
Pilot Testing
We conducted pilot testing sessions of our eLearning materials, where we invited student users for app testing, assessment, and focused group interviews.
In these sessions, each participant completed:
A pre-module test that consisted of multiple-choice questions (MCQs) related to lines and tubes, evaluating their preexisting domain knowledge
A pre-module survey on theirexisting confidence in the topic
Then, participants completed each level of the developed game and materials under the supervision of our project team members. Questions arising from the game were answered by the Project Supervisor.
A post-module test that consisted of domain knowledge MCQs, to determine any learning improvement
A post-module survey on their confidence and satisfaction after using the modules
A group focused interview session to gather feedback
Afterwards, participants completed:
A post-module test that consisted of domain knowledge MCQs, to determine any learning improvement
A post-module survey on their confidence and satisfaction after using the modules
A group focused interview session to gather feedback
Results:
Results of the pre-post surveys showed that the eLearning materials met our goals of enhancing domain knowledge skills and student confidence.
The mean domain knowledge test score showed a relative increase of 127% from a mean of 37.37% in the pre-game survey to 84.85% in the post-game survey.
Participants also showed a significant relative increase (mean 107%)in their overall confidence level
The participants gave an average rating of 6.48 on a 7 point Likert scale for their learning satisfaction and experience.

Acknowledgments

This project was supported by the University Grants Committee (UGC) of Hong Kong under the Teaching Development and Language Enhancement Grant 2022–25.
Get started
Have questions or ideas to share?
Have questions or ideas to share?
focusedradiology@gmail.com
Contact Us
Terms of use
© 2025 Focused Radiology. All Rights Reserved.
Privacy Policy