Project Abstract

Thyroid disease continues to be a significant global health issue, affecting millions of people around the world. Diagnosing thyroid disorders, particularly in subclinical cases, presents substantial challenges for physicians despite their common occurrence. A major obstacle to accurate diagnosis is the presence of endogenous interference in patient blood samples, which can greatly impact the interpretation of clinical data. Such interferences especially affect thyroid function tests, which are routinely used to evaluate thyroid status. This could lead to delays in diagnosing and treating the disease. This project investigates the use of machine learning (ML) algorithms to identify interferences in patient blood test results, facilitating the early detection of thyroid disorders. Previous studies have mostly concentrated on attaining a high level of accuracy in identifying thyroid diseases using laboratory data. However, the problem of interference in machine learning-based healthcare decision-making is overlooked. Therefore, this project aims to address this gap by exploring the potential combination of machine learning techniques to develop an ensemble model. This model will offer physicians a clear understanding of its results and, ideally, provide prescriptive analytics to suggest more testing when needed. The objective is to decrease errors in thyroid disease detection by implementing machine learning algorithms specifically designed to detect and minimise the effects of interferences. Hopefully, this project can make a valuable contribution towards better patient outcomes and more informed clinical decision-making in the treatment of thyroid diseases.

Keywords: Machine Learning, Healthcare Data Analysis, AI for Medical Application

 Conference Details

Session: Presentation Stream 11 at Presentation Slot 4

Location: College 127 at Tuesday 7th 13:30 – 17:00

Markers: Nicholas Micallef, Alec Critten (GTA)

Course: MSc Computer Science, Masters PG

Future Plans: I’m looking for work