A stroke is the loss of brain cells caused by either inadequate blood flow or bleeding that affect brain function. An obstruction in the cerebrovascular system that directly restricts blood vessels and halts blood flow to specific areas of the brain causes an ischemic stroke, a frequent kind of stroke. Computed tomography (CT) scanning is often used to evaluate stroke, and the timely and accurate identification of ischemic brain stroke (IBS) using CT scans is crucial for determining the appropriate course of treatment. However, a number of circumstances, such as the huge volume of patients admitted to medical facilities and the hectic schedules of experts, might render manual diagnosis of ischemic stroke prone to mistakes. Concerning such issues, artificial intelligence (AI) frameworks like Machine Learning (ML) and Deep Learning (DL) models have been used to construct automated systems for stroke diagnosis. Efficient patient outcomes and timely medical treatment depend on the early and accurate identification of stroke, which is made possible by these technologies. DL models for automatically detecting ischemic stroke and segmenting brain CT images have been developed in recent years. This article outlines the evolution of different methods used for segmenting and classifying IBS aiming to encourage further study in this field. DL models developed specifically for the classification and segmentation of ischemic brain stroke will be examined first in the study. Each framework’s efficacy is then assessed in terms of performance via an analysis of its benefits and drawbacks. Finally, certain improvements are recommended to increase the effectiveness of ischemic stroke detection and segmentation. By examining and assessing the difficulties encountered in the literature, readers may clearly identify problems and provide creative solutions in DL-based models for ischemic brain stroke diagnosis.