Beyond Boring Vision Therapy: How AI Empowers Students to Create Engaging Treatment Tools

Beyond Boring Vision Therapy: How AI Empowers Students to Create Engaging Treatment Tools

Imagine a child with lazy eye, traditionally forced to endure boring, repetitive exercises that feel more like punishment than treatment. Now picture the same child engaged with innovative, AI-powered therapeutic tools designed by PolyU’s School of Optometry students who have never written a line of code. This transformation is happening right now at PolyU, where successful implementation of an artificial intelligence (AI) integration programme is revolutionising how students approach vision therapy development.

Under the leadership of Professor Jeffrey Leung, Assistant Professor of the School of Optometry, this initiative demonstrates measurable improvements in student innovation capacity and educational outcomes.

Addressing the Challenge of Patient Compliance

Vision therapy treats critical binocular vision disorders affecting substantial patient populations. As Professor Leung explains, “Some of the very common eye problems include strabismus or squint, which means one eye is not aligned properly, turning in, turning out, up or down. Lazy eye is also the most common vision impairment in children.”

Traditional therapeutic tools, whilst clinically effective, face significant compliance challenges due to their tedious nature. Professor Leung identifies this critical gap, “I always tell my students that even though the training is very effective, if our compliance is bad, then the training is not very useful.”

The programme evolved through two transformative phases. During the initial phase spanning 2021 to 2023, students focused on hardware-based innovation through collaborative partnership with the PolyU Industrial Centre, producing 16 functional prototypes incorporating cutting-edge technologies including 3D printing for customised therapeutic devices. A student team achieved the remarkable feat of winning a design competition as “the only team without any engineering background” competing against biomedical engineering and computer science participants.

However, this phase revealed a fundamental challenge. Professor Leung observed, “Students spent a lot of time just solving technical issues, and they have less time available for their clinical innovation.”

The strategic deployment of AI as a “technical coach” in 2024 transformed this limitation into an opportunity. Professor Leung articulated the breakthrough, saying, “AI can actually bridge the gap. The students can offload the technical debt to the AI, and therefore students can retain more time on their creativity and problem-solving skills.” This phase marked a watershed moment—the first time students successfully developed software-based solutions without prior programming experience.

 

From Technical Barriers to Creative Freedom

This educational structure seamlessly integrates traditional clinical training with cutting-edge AI-powered development methodologies. Students progress through comprehensive foundation knowledge transfer, conduct patient needs assessment through direct clinic observation, and participate in mid-semester consultation sessions with practising clinicians. Professor Leung describes the innovative AI integration approach, “I actually treat it as a technical coach to teach our students how to do something that they are not very familiar with—that is coding.” Beyond therapy development, the AI integration extends to clinical education itself. Professor Leung has introduced an AI clinical tutor system that transforms traditional examination training. Students perform vision examination steps whilst being recorded and analysed by AI. The AI tutor provides feedback, informing students how to perform examinations more effectively. This dual application of AI technology—both as a development tool and educational mentor—creates a comprehensive learning ecosystem that enhances both creative innovation and clinical competency.

The initiative reveals complementary roles between educators and artificial intelligence. Professor Leung observes, “In these two years, we also have had debate about whether AI can replace teachers. But from this initiative, I found that teachers and AI can have different roles.” Teachers retain responsibility for clinical knowledge foundation and professional experience integration, whilst AI handles technical problem resolution and programming implementation.

Performance improvements span multiple dimensions. Professor Leung reports significantly faster innovation speeds compared to previous years, with enhanced technical capability extending across the entire student cohort. “Student feedback demonstrates substantial confidence gains. Students also reported that they felt more technically capable because they did not have any computer science background, but with AI coding, they can make something based on their original idea and design.”

Success factors include comprehensive instructor preparation and structured problem decomposition teaching. Professor Leung emphasises preparatory importance, “As a teacher, we must understand the limitations of AI.” He extensively tested various AI agents and platforms before implementation, ensuring thorough understanding of technological capabilities and constraints.

Strategic Vision for Healthcare Innovation

Immediate objectives centre on enhanced interdisciplinary collaboration. Professor Leung actively pursues partnerships, collaborating with colleagues with computer science backgrounds, hoping one day students’ designs can be used in clinical practice. The long-term strategic vision encompasses immersive technology integration, with Professor Leung envisioning expanded capabilities. “Maybe one day, we can move into VR technology that can be even more engaging,” he suggests, describing virtual reality and augmented reality applications for future therapeutic innovations. This vision therapy AI integration positions PolyU at the forefront of healthcare education innovation. Students develop enhanced technical capabilities and creative problem-solving skills, whilst staff experience improved teaching efficiency and engagement. The healthcare industry gains access to a pipeline of AI-literate professionals, and patients benefit from more engaging therapeutic tools.

The programme represents a successful model for strategic technology adoption in healthcare education, maintaining focus on human expertise whilst leveraging AI capabilities. This approach achieves measurable educational improvements whilst preparing students for an evolving healthcare technology landscape where AI tutoring and innovative manufacturing technologies work together to transform both learning and patient care.

 

 

 

 

 

 

This article is part of PolyU’s Education 4.0 (E4.0) series, which showcases how the University is responding to rapid AI advancements and a changing educational landscape. Through E4.0, PolyU is transforming learning and teaching by integrating AI and smart technologies into a student-centred approach that fosters innovation, expands the use of educational technologies, and prepares students for the future.