Recent advances in artificial intelligence (AI) offer many opportunities to implement it in a broad range of industries. One of the main ambitious application of AI is in healthcare and patient monitoring. In healthcare industry, unlike the most commercial applications of AI, a missed detection/prediction of a clinical event may result in the death of a patient. Moreover, a high false alarm rate may lead to misdiagnosis causing extra effort and cost for care providers. Thus, in healthcare applications, it is required to strictly minimize the number of false alarms without sacrificing the sensitivity rate of the detection system. This dissertation focuses on epileptic seizure detection and classification in long-term electroencephalogram (EEG) records. More specifically, the two main challenges in supervised EEG seizure detection, the curse of dimensionality and the curse of variability are tackled.<br /><br />First, a signal decomposition technique, applicable to physiological signals, is devised which can be used as a preliminary step for feature extraction. This is performed by proposing a novel time-frequency transform based on rational functions, namely, rational short time Fourier transform (RSTFT). A sparse decomposition method is then proposed by reconstructing the input signal into several components using non-overlapping sub-sets of the RSTFT coefficients. Sparse representation of signal components is then obtained by inducing L1 regularization penalty on the RSTFT coefficients during the reconstruction phase. The effectiveness of the proposed sparse decomposition method is evaluated in the classification of long-term EEG records for purposes of epileptic seizure detection and sleep stage scoring.<br /><br />Another part of the thesis investigates the detection of seizures in textural images constructed using a proposed scheme for mapping of EEG signals into gray-scale image domain. The proposed mapping strategy makes it plausible to correlate textural changes of the obtained images with seizure activities in EEG records. The textural analysis is then carried out using the well-known gray-level cooccurrence matrix (GLCM) and Haralick’s feature extraction method resulting in a compact representation of EEG epochs. All the methods proposed in this thesis are evaluated using public EEG data-sets freely available on-line. The results obtained by the proposed methods are comprehensively compared with the other conventional dedicated feature extraction techniques using several classifiers.<br /><br />The main contribution of the thesis is in the adaptation of conventional feature extraction techniques, commonly used in the textural analysis of images, to be applicable in EEG signal analysis. Additionally, the discriminatory power of feature descriptors is improved by representing EEG signals using their sparse rational components. The proposed rational local Gabor binary pattern (LGBP)-width feature outperforms competing methods in both seizure detection and classification problems. Moreover, its perform consistency in patient/non-patient specific scenarios demonstrates its ability to tackle the curse of variability in seizures.