Author:

Keywords:

SISTA, BIOTENSORS - 339804;info:eu-repo/grantAgreement/EC/FP7/339804

Abstract:

Within this thesis the automated algorithms for the EEG-based assessment of the brain f unctioning of asphyxiated infants have been develo ped. Their goal is to assess the severit y of the hypoxic brain injuries in the asphyxiated infants. This estimate will assist ¨clinicians to promptly diagnose and to guide furt her treatment decisions. Three main contribut ions are developed. First, an algor ithm that detects dynamic interburs t intervals has been extended with a post pro cessing step that removes dubious and uncerta in detections. In this way, a trustworthy alg orithm has been developed. Second, we explored qua ntification of long-range temporal behavior o f the neonatal EEG. We explored (Multifractal ) Detrended Fluctuation Analysis and, further on, we proposed four metrics that show¨ potential to differentiate the EEG background grad es. Third, an automated method for the backgr ound EEG classification has been developed. As the first step, it maps shorter, segme nted, EEG segments’ features into segments’ feature space, thereby creati ng a 3D distribution. Next, this 3D structure ¨is represented as a data tensor that is used ¨for further dimensionality reduction and rob ust classification. The algorithms and their¨ performances have been verified by expert EEG readers, demonstrating its potential. In add ition, the efficient visualization devel oped within our project NeoGuard will enable fast insight into the algorithms’ output and, hopefully, very soon be implement ed in the NICUs.