By exploring how AI models could enhance annotation processes, colleagues in Medical Sciences sought to reduce the manual workload for educators while enriching the learning experience for students by providing clearly annotated resources that encourage autonomous, interactive, and engaging study.
As part of the University’s AI in Teaching and Learning Exploratory Fund, the ‘AI-Assisted Annotations for Histology’ project set out to explore how artificial intelligence could support and enhance histology education - specifically by easing the complex task of annotating digital tissue slides.
The project aimed to build and deploy an AI-driven pipeline capable of automating the identification, segmentation, clustering and labelling of cells in histological images. Rather than developing new classification models or enabling real-time lab-based annotation - both out of scope due to time and technical constraints - the team focused on building a robust system to streamline existing annotation processes. This approach promised to reduce the workload for teaching staff while offering students a more engaging and independent learning experience.
From the AI Competency Centre, Haseeb Ahmad, Senior Research Software Engineer, led the technical development of the machine learning pipeline. He adapted it to handle a wide variety of tissue types, despite the challenge of limited ground truth data.
Initial results have shown promising potential in using AI to support histology teaching. However, the project also highlighted limitations - particularly the difficulty of applying a single model across the wide diversity of cell and tissue structures. Some configurations performed well on certain slides but poorly on others, underlining the need for ongoing refinement.
Looking ahead, the project will focus on scaling the approach and deepening understanding of how AI behaves across complex biological datasets. Future iterations will explore embedding interactive features - such as quizzes and enhanced annotation tools - within the CSlide platform to further enrich the teaching and learning experience in histology.
AI-assisted Simulated Patient for Clinical Communication Skills Training
AI in Education