Title:
Sig-Lime: An Enhancement Of Lime For Explainable Cardiac Arrhythmia Classification From ECG Signals

Author:

Supervisor:

Department:

Computing

Year:

2024

URL:

Abstract:

Cardiac arrhythmia represents a significant health concern, necessitating early and accurate classification for effective treatment and enhanced patient outcomes. The high accuracy of deep learning mo...
View more

Multimedia Content:

01

Methodology

The thesis adopts a multi-phase experimental methodology that combines data preparation, deep learning model development, and enhancement of interpretability (LIME). Phase 1: Preliminary Study •Revi...
View more
Methodology Image
02

Citation Analysis

1. Diagnostics (MDPI) Scope: Medical diagnostics, including AI-based ECG analysis and arrhythmia classification. Relevance: Recently published works on interpretable deep learning for ECG signals ...
View more
03

Impact

1. Societal Impact •Better Healthcare Outcomes: More accurate arrhythmia detection enables earlier diagnosis and treatment, reducing risks of stroke, heart failure, and sudden cardiac death. •Impr...
View more
04

Benefits to Society

1. Healthcare Benefits Enables early and accurate detection of cardiac arrhythmias, reducing risks of stroke, heart failure, and sudden cardiac death. Provides transparent and explainable AI outpu...
View more
05

Applications

1. Healthcare Diagnostics Systems Application: Automated cardiac arrhythmia detection from ECG signals. Use Case: Hospitals and clinics use the hybrid CNN–GRU model to classify arrhythmias such as...
View more
06

Infographic

Summarize of this research
Infographic Image
07

Patents / Products

1. Explainable ECG Diagnostic Engine Product Type: Software module or medical API Use Case: Integrates with hospital information systems to automatically classify arrhythmias and provide interpret...
View more
08

Experts

1. Assoc. Prof. Ts Dr Mohd Soperi bin Mohd Zahid Institution: Universiti Teknologi PETRONAS (UTP), Malaysia Contribution: Main supervisor for the SIG-LIME research; guided development of hybrid CN...
View more