Developing a Learning Software for the Annotation of Fundus-images in Medical Training in regard to Cognitive Load, Usability and E-Learning Factors

Thumbnail Image

Files

master_Bittner_Sara-Jane_2023.pdf (44.42 MB)

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2023-12-11

Department

Major/Subject

Human-Computer Interaction and Design

Mcode

SCI3020

Degree programme

Master's Programme in ICT Innovation

Language

en

Pages

118+11

Series

Abstract

This thesis develops a learning tool for junior doctors in ophthalmology to learn how to interpret fundus images. This skill is needed in diagnosing diseases of the eye like diabetic retinopathy. Currently, these doctors primarily rely on ophthalmologists which often lack time for teaching and feedback. This is especially crucial as available learning materials are not sufficient to learn independently yet. To bridge this gap, a machine learning ML-based learning tool was developed. The tool stems from the OphthalmoAI project by the BMBF and the DFKI. The tool utilizes an ML algorithm that can detect pathologies in fundus images. The development process followed the design thinking framework, integrating the experiences and needs of junior doctors. The design was informed by the cognitive load theory, usability principles, and important e-learning factors like independent learning, and variability of examples. The results indicate that the developed ML-based learning tool meets the general requirements set in the research questions: First, the tool successfully reduced cognitive load by dividing tasks and simplifying information. Further, usability was enhanced through minimalist design and clear system visibility. Lastly, the feedback supported the junior doctors and the learning tool shows potential to increase the independence in learning of junior doctors. However, limitations arose in the accuracy of classifying pathologies and feedback quality. This could compromise the learning tool’s reliability in a medical context. Despite these technical constraints, the study suggests that an ML-based learning tool is feasible for medical education, addressing time constraints and providing valuable learning independence for junior doctors. Future studies could focus on addressing the detected technical limitations and based on that re-evaluate the tool’s impact on cognitive load as well as usability.

Description

Supervisor

Nieminen, Mika P.

Thesis advisor

Barz, Michael

Keywords

learning tool, cognitive load, usability, e-learning, medical training

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202401071332

Collections

[dipl] Perustieteiden korkeakoulu / SCI