Improving Abnormality Detection on Chest Radiography Using Game-Like Reinforcement Mechanics

Area of Science:

• Medical Education
• Radiology
• Digital Learning

Background:

• Digital learning resources often prioritize explicit knowledge over implicit skills like abnormality detection.
• Implicit skills in radiology, such as interpreting chest radiographs, traditionally require extensive clinical practice.
• There is a need for effective digital tools to train implicit diagnostic skills.

Purpose of the Study:

• To evaluate the effectiveness of a gamified online module in teaching abnormality detection in chest radiographs.
• To assess the impact of game mechanics (instant feedback, timed cases) on learning implicit skills.
• To compare the performance of novice and experienced radiography interpreters using the module.

Main Methods:

• Developed a free, online gamified module with 74 chest radiograph cases (normal and abnormal).
• Implemented game mechanics: instantaneous feedback, rapid-fire presentation, and 15-second case timers.
• Recruited 60 volunteers, divided into 'never trained' and 'previously trained' groups, and measured performance on training and testing sets.

Main Results:

• The 'never trained' group showed significant improvement, scoring 59.1% during training and 64.9% during testing (P<.001).
• The 'previously trained' group showed minimal improvement (75.6% training vs. 76.4% testing, P=.56), indicating less benefit for experienced interpreters.
• 87% of users found the module efficient, and 83% deemed it valuable for medical students.

Conclusions:

• A gamified online module can enhance abnormality detection skills in novice chest radiograph interpreters.
• Experienced interpreters may not benefit as significantly from this specific type of training.
• User feedback indicates a positive reception and perceived value of the gamified learning approach.