ABSTRACT
The seizure detection algorithm (SDA) has become an increasingly common tool in epileptology since the advent of automated medical signal analysis. SDAs –nd utility beyond clinical epilepsy management in basic science as an analysis tool, and more recently in the development of devices for automated seizure warning and treatment. This chapter provides an introduction to the formulation, evaluation, and optimization of SDAs, with examples chosen from the literature. It is directed more at clinicians than at those whose primary expertise is mathematics, engineering, or physics-but will hopefully also motivate practicing engineers and mathematicians to revisit the ways in which they design and implement these algorithms. It is emphasized that seizure detection is important not only for timely intervention but for understanding seizure dynamics and evaluating treatment as well. To this end, approaches for seizure quanti–cation are discussed. The goal is to provide readers with the tools for understanding the workings of any SDA within a generic framework, and to pinpoint critical considerations for correct SDA use and performance assessment.
