This article sets out to retrace the course of a neuro-phenomenological project initiated by Francisco Varela, the purpose of which is anticipating epileptic seizure, while trying to evaluate the relevance of the neuro-phenomenological approach from the methodological, therapeutic and epistemological viewpoints. New mathematical methods for analysing the neuro-electric activity of the brain have recently enabled researchers to detect subtle modifications of the cerebral activity a few minutes before the onset of an epileptic seizure. Do these neuro-electric changes correspond to modifications in the patients' subjective experience, and if that is the case, what are they? In a first part, after having recalled the context of the project, I will describe the methods I used for trying to detect the dynamic micro-structure of preictal experience, the difficulties I met and the results I obtained. Then I will show how this "pheno-dynamic" analysis and neuro-dynamic analysis have guided, determined and mutually enriched each other throughout this project. In a third part, I will show that this genetic approach to epileptic seizure opens a new line of research into a cognitive and non pharmacological therapy of epilepsy. Finally, I will argue through this example that neuro-phenomenological co-determination could shed new light on the difficult problem of the "gap" which supposedly separates subjective experience from neurophysiological activity.