Towards Trustworthy Healthcare AI: Multi-Method Explainable AI for Transparent Adverse Drug Reaction Detection from Patient-Generated Text

Sunny, Alex (2025) Towards Trustworthy Healthcare AI: Multi-Method Explainable AI for Transparent Adverse Drug Reaction Detection from Patient-Generated Text. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (769kB) | Preview
Preview	PDF (Configuration Manual) Download (631kB) | Preview

Abstract

Adverse Drug Reactions (ADR) are a critical challenge in pharmacovigilance. While there are automated ADR detection ML systems, their opaque decision-making process stalls clinical usage. This research addresses how these ADR detection systems can be combined with explainable AI methods to enhance transparency. Extensive and comprehensive review of the state of the art and previous works and contributions have also been discussed. A framework integrating ML and DL models with five distinct XAI techniques (LIME, SHAP, Integrated Gradient, Tree Interpreter and PDP) is implemented for ADR detection from patient provided drug reviews. The LSTM model achieved optimal performance (91% precision) for ADR detection. The XAI methods are evaluated using quantitative metrics like faithfulness, fidelity and complexity. LIME and SHAP have the best balance between accuracy and interpretability for clinical applications and the Tree Interpreter showcased the highest faithfulness. This research contributes an evaluation framework for XAI methods in healthcare, showcases the practical integration of XAI in pharmacovigilance and provides evidence-based guidance for developing transparent ADR detection systems which improve clinical decision-making and maintain model performance.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Chikkankod, Arjun UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence R Medicine > Healthcare Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Chemical Industry > Pharmaceutical industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	26 Aug 2026 12:22
Last Modified:	26 Aug 2026 12:22
URI:	https://norma.ncirl.ie/id/eprint/9672

Actions (login required)

View Item