← 返回 JSSC 论文列表JSSC 2023第1期Digital Circuits
A 962-nJclass Neural Signal Processor With Adaptable Intelligence for Seizure Pr
首款用于癫痫预测的神经信号处理器，集成预处理、特征提取、可重构SVM核和后处理单元，提升预测性能。
92.0%灵敏度, 0.57/h误报率, 8.44ms训练延迟, 2.31mW功耗, 6.05MHz频率
癫痫预测神经信号处理器支持向量机能量优化矩阵乘法
▸近似能量算子（AEO）减少特征提取器面积28%
▸基于缩放的Newton-Raphson除法器减少迭代次数62.5%
▸基于指针的矩阵乘法减少ADMM-SVM训练计算复杂度99.9%
Abstract
This work presents the world’s first neural signal
processor for seizure prediction, which includes a preprocessing
unit, a feature extractor, a reconfigurable support vector machine
(SVM) kernel, and a postprocessing unit. Seizure prediction per-
formance is enhanced by on-chip training for model adaptation.
Design optimization is applied across the layers of abstraction
to minimize the area and energy. The area of the feature
extractor is reduced by 28% with an approximated energy oper-
ator (AE