Semi-Supervised Anomaly Detection Pipeline for SOZ Localization Using Ictal-Related Chirp

2025-08-18

AI Summary

This study evaluates the spatial concordance between statistically identified abnormal channels—based on time-frequency features of chirp events—and clinically defined seizure onset zones (SOZs). We propose a semi-supervised anomaly detection framework: (1) extracting chirp event characteristics—including onset/offset frequencies and duration—via time-frequency analysis; (2) detecting abnormal channels using Local Outlier Factor (LOF) with adaptive neighborhood selection; and (3) quantifying spatial overlap between statistical and clinical SOZs via an electrode proximity-weighted similarity metric. In seizure-free patients, the weighted spatial similarity reaches 0.903; in surgically successful cases, it is 0.865—significantly higher than in unsuccessful cases (0.460), demonstrating strong correlation with clinical outcomes. The key contributions are the integration of chirp-specific time-frequency dynamics, adaptive LOF-based anomaly detection, and anatomy-informed spatial weighting—enhancing both reliability and interpretability of individualized SOZ localization.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Semi-Supervised Learning

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion
• Security and Privacy: Large-scale security measurements

Abstract

This study presents a quantitative framework for evaluating the spatial concordance between clinically defined seizure onset zones (SOZs) and statistically anomalous channels identified through time-frequency analysis of chirp events. The proposed pipeline employs a two-step methodology: (1) Unsupervised Outlier Detection, where Local Outlier Factor (LOF) analysis with adaptive neighborhood selection identifies anomalous channels based on spectro-temporal features of chirp (Onset frequency, offset frequency, and temporal duration); and (2) Spatial Correlation Analysis, which computes both exact co-occurrence metrics and weighted index similarity, incorporating hemispheric congruence and electrode proximity. Key findings demonstrate that the LOF-based approach (N neighbors=20, contamination=0.2) effectively detects outliers, with index matching (weighted by channel proximity) outperforming exact matching in SOZ localization. Performance metrics (precision, recall, F1) were highest for seizure-free patients (Index Precision mean: 0.903) and those with successful surgical outcomes (Index Precision mean: 0.865), whereas failure cases exhibited lower concordance (Index Precision mean: 0.460). The key takeaway is that chirp-based outlier detection, combined with weighted spatial metrics, provides a complementary method for SOZ localization, particularly in patients with successful surgical outcomes.

Problem

• Localizing seizure onset zones using anomaly detection
• Evaluating spatial concordance between clinical and statistical SOZs
• Combining chirp-based outlier detection with weighted spatial metrics

Innovation

• Unsupervised outlier detection with LOF
• Spatial correlation analysis with weighted metrics
• Chirp-based spectro-temporal feature analysis