The doctoral dissertation in the field of Computer Science will be examined at the Faculty of Science, Forestry and Technology on the Joensuu Campus.
What is the topic of your doctoral research? Why is it important to study the topic?
My doctoral research studies how eye movements can help detect the early signs of mild cognitive impairment (MCI) and Alzheimer's disease (AD). Using eye-tracking during simple reading and number-naming tasks, I measure small changes in how people look and read, and combine these signals with machine learning to separate healthy aging from early cognitive decline.
This topic is important because dementia is rising worldwide, yet many cases are found too late. Current tests can be costly, invasive, or insufficiently sensitive to detect the earliest changes. Eye-tracking is fast, low-cost, and non-invasive, so it offers an objective and affordable way to identify people at risk sooner. Earlier detection gives patients and families more time to plan, find support, and benefit from care and future treatments.
What are the key findings or observations of your doctoral research?
My research shows that eye movements and speech recorded during simple reading and King Devick tasks can reveal early, subtle signs of cognitive decline. People with mild cognitive impairment and Alzheimer's disease made shorter and less stable fixations, skipped and re-read more words, and showed different gaze and pupil patterns than healthy older adults. MCI was the hardest group to identify, but combining eye-tracking with speech clearly improved accuracy, especially in separating MCI from healthy aging. What is new is the use of a multimodal deep-learning model that fuses gaze and speech, rather than relying on a single signal. The tasks are short, low-cost, and non-invasive, so they could fit into everyday clinical settings. I also used explainable AI to show which features drive each decision, making the results easier for clinicians to understand and trust. For the public and the scientific community, this offers a practical and objective step toward detecting MCI and AD earlier, when care and future treatments can help most
How can the results of your doctoral research be utilised in practice?
The results can support faster, lower-cost screening for early cognitive decline. Because the tests use only short reading and number-naming tasks with an eye-tracker and a microphone, they could be added to routine check-ups in memory clinics, primary care, or even mobile settings, without invasive procedures.
Clinicians could use the tools to flag people who may need closer assessment, helping detect mild cognitive impairment and Alzheimer's disease earlier. Earlier detection means patients and families can plan ahead, access support sooner, and take part in treatments or trials when they are most effective. The explainable AI methods also show which eye and speech features matter most, which can guide clinical judgment and future research.
Beyond healthcare, the same approach could help monitor cognitive changes over time, measure how well treatments work, and support the design of accessible digital health tools for aging populations.
What are the key research methods and materials used in your doctoral research?
This was a cross-sectional study carried out with older adults grouped as cognitively healthy, with MCI, and AD, recruited and clinically assessed with neuropsychological tests. We first reviewed the existing literature to map how eye-tracking has been used in this field. Participants then completed short, naturalistic tasks, mainly a reading task and the King-Devick number-naming test, while their eye movements were recorded using a screen-based eye-tracker and their speech was recorded using a microphone.
From the eye data, we extracted features such as fixations, saccades, regressions, skipping, and pupil responses, and I added acoustic speech features. These signals were analyzed with statistical tests and machine learning, including a multimodal deep-learning model that fuses gaze and speech. We also applied explainable AI to identify which features contributed most to distinguishing the groups.
The doctoral dissertation of Hasnain Ali Shah, MSc, entitled Eye-tracking based behavioral analysis for early detection of mild cognitive impairment and Alzheimer’s Disease will be examined at the Faculty of Science, Forestry and Technology, Joensuu campus. The opponent will be Professor Oleg Komogortsev, Texas State University, USA, and the custos will be Professor Roman Bednarik, University of Eastern Finland. Language of the public defence is English.
