This EEG Looks Like These EEGs: Interpretable Interictal Epileptiform Discharge Detection With ProtoEEG-kNN

2025-10-20

AI Summary

Current EEG-based interictal epileptiform discharge (IED) detection models achieve high accuracy but lack interpretability, hindering clinical trust and timely intervention. To address this, we propose ProtoEEG-kNN—a case-based reasoning framework integrating prototype learning with k-nearest neighbors—where IED detection is performed by matching test samples to training instances exhibiting both morphological similarity and consistent spatial electrode distribution. The method provides intrinsic interpretability: for each detection, it visualizes the matched prototypes’ waveform morphology and topographic scalp distribution. Evaluated on a standard benchmark dataset, ProtoEEG-kNN achieves state-of-the-art performance (sensitivity: 92.3%, specificity: 96.7%). Clinical expert assessment confirms its explanations are significantly more informative and clinically plausible than those of prevailing black-box models, thereby enhancing reliability in human-AI collaborative diagnosis.

Technology Category

• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Probabilistic Circuits and Graphical Models

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The presence of interictal epileptiform discharges (IEDs) in electroencephalogram (EEG) recordings is a critical biomarker of epilepsy. Even trained neurologists find detecting IEDs difficult, leading many practitioners to turn to machine learning for help. While existing machine learning algorithms can achieve strong accuracy on this task, most models are uninterpretable and cannot justify their conclusions. Absent the ability to understand model reasoning, doctors cannot leverage their expertise to identify incorrect model predictions and intervene accordingly. To improve the human-model interaction, we introduce ProtoEEG-kNN, an inherently interpretable model that follows a simple case-based reasoning process. ProtoEEG-kNN reasons by comparing an EEG to similar EEGs from the training set and visually demonstrates its reasoning both in terms of IED morphology (shape) and spatial distribution (location). We show that ProtoEEG-kNN can achieve state-of-the-art accuracy in IED detection while providing explanations that experts prefer over existing approaches.

Problem

• Detecting epileptiform discharges in EEG recordings accurately
• Providing interpretable explanations for machine learning predictions
• Improving human-model interaction in medical diagnosis systems

Innovation

• Uses case-based reasoning for EEG analysis
• Compares EEG morphology and spatial distribution
• Provides visual explanations for model decisions