MRI detection of structural brain lesions in epilepsy is critical for diagnosis and surgical planning. Despite advances in MRI technology, lesions including focal cortical dysplasia (FCD) and hippocampal sclerosis (HS) remain undetected on visual radiological review in up to 40% of patients. This chapter reviews automated methods developed for lesion detection, spanning voxel-based analyses, surface-based approaches, and cutting-edge AI-driven algorithms. For FCD detection, we focus on three tools available for reuse, which have high sensitivity in detecting lesions that had previously been missed. For HS, we describe automated approaches for characterizing hippocampal morphology and detecting and lateralizing the pathology. While such tools show promise for improving our ability to diagnose and treat subtle epileptogenic lesions, challenges remain in integrating them into routine practice. Future directions include expanding models to incorporate additional modalities and a wider range of lesional pathologies as well as prospective validation of developed models. Nevertheless, automated lesion detection could significantly impact epilepsy management, facilitating earlier diagnosis, targeting surgical planning, and, ultimately, improving patient outcomes.