The review presented by Moreno-Castelblanco et al. focuses on how artificial intelligence models, specifically deep and machine learning, process EEG for lower limb motor imagery in lower limbs to support brain-computer interface applications in neurohabilitation. A systematic review used a sample of 35 articles published between 2020 and 2025 that focused on lower-limb motor imagery and EEG recording techniques in addition to artificial intelligence models for rehabilitation. From these articles, the study design, participants, preprocessing, classification models, and outcomes were recorded. A majority of the studies utilized machine learning or deep learning to classify motor imagery; classification and interpretation (30% of the models) with a mix of machine learning and artificial intelligence models had an accuracy range from 70% to 94.3%. Furthermore, motor imagery based brain-computer interfaces using EEGs have shown, utilizing a deep learning model (CNN), 89.04% offline accuracy and 57.28% real time accuracy. The authors highlight advances in signal, AI-based, and multimodal acquisition, however the study requires larger data sets and a standardized method across trials.