Plant ailments can threaten farmers’ livelihoods and their ability to produce enough food. Early detection and diagnosis are crucial for effective management and control of plant diseases. Leaf analysis is a promising approach for disease stage prediction, as it can be used to spot minute changes in leaf physiology and appearance that might happen even before the start of obvious symptoms. The algorithms of machine learning and deep learning can be used to extract characteristics from pictures of leaves and categorize them into various disease stages. These algorithms have shown promising results in recent studies, achieving high accuracy in predicting disease stage even with limited training data. Machine learning and deep learning combined with leaf analysis have the potential to transform plant disease management by providing early detection, precise diagnosis, and tailored treatment. To develop effective disease management strategies, it is important to accurately determine the stage of plant disease development. Researchers are developing new methods for plant disease stage detection using several methods, such as image processing, machine learning, and spectroscopy. This can lead to significant economic and environmental benefits for farmers and can also help to improve food security. As a primary study, the evaluation includes a variety of articles from 2014 to 2023. When search tactics were considered, 117 research publications were found, of which 36 were deemed to be relevant. The paper discusses various deep learning research advances. It will also aid researchers in determining the current and future scenario of plant disease research utilizing deep learning technologies.