NeuroHD-RA: Neural-distilled Hyperdimensional Model with Rhythm Alignment

2025-07-12

AI Summary

Traditional high-dimensional computing (HDC) for ECG-based disease detection suffers from reliance on static random projections, limited task adaptability, and poor interpretability. To address these limitations, this paper proposes a neuro-enhanced, interpretable HDC framework. Its key contributions are: (1) an RR-interval-guided heart-rate-aligned segmentation strategy ensuring physiologically consistent signal modeling; (2) a trainable RR-block encoder coupled with a binarized linear projection layer, preserving symbol-level interpretability while enabling end-to-end optimization; and (3) integration of neural distillation with a joint loss combining cross-entropy and surrogate metrics. Experiments demonstrate state-of-the-art performance: 73.09% precision and 0.626 F1-score on Apnea-ECG—significantly outperforming baseline HDC and deep learning methods—and strong generalization and robustness on PTB-XL. The framework bridges the gap between HDC’s efficiency and neural networks’ adaptability, offering both high accuracy and physiological interpretability.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

We present a novel and interpretable framework for electrocardiogram (ECG)-based disease detection that combines hyperdimensional computing (HDC) with learnable neural encoding. Unlike conventional HDC approaches that rely on static, random projections, our method introduces a rhythm-aware and trainable encoding pipeline based on RR intervals, a physiological signal segmentation strategy that aligns with cardiac cycles. The core of our design is a neural-distilled HDC architecture, featuring a learnable RR-block encoder and a BinaryLinear hyperdimensional projection layer, optimized jointly with cross-entropy and proxy-based metric loss. This hybrid framework preserves the symbolic interpretability of HDC while enabling task-adaptive representation learning. Experiments on Apnea-ECG and PTB-XL demonstrate that our model significantly outperforms traditional HDC and classical ML baselines, achieving 73.09% precision and an F1 score of 0.626 on Apnea-ECG, with comparable robustness on PTB-XL. Our framework offers an efficient and scalable solution for edge-compatible ECG classification, with strong potential for interpretable and personalized health monitoring.

Problem

• Develops interpretable ECG disease detection using neural-distilled HDC
• Improves rhythm-aware encoding via trainable RR interval segmentation
• Enables edge-compatible ECG classification with adaptive representation learning

Innovation

• Neural-distilled HDC architecture with RR-block encoder
• Rhythm-aware trainable encoding using RR intervals
• Hybrid framework combining interpretability and adaptive learning