Cancer is characterized by the proliferation of aberrant cells originating from any organ inside the human body. Fundamentally, the proliferation of cells within these organs is reached a point of saturation. Deep learning (DL) is a subfield within the realm of machine learning and artificial intelligence that has found extensive application across various disciplines, including but not limited to health care and medication creation. The study of cancer prognosis involves the estimation of the eventual outcome for individuals affected by cancer, as well as the estimation of their survival rates. The primary aim of this study is to investigate the advancements and challenges associated with enhancing cancer diagnosis using deep learning models based on green artificial intelligence. The research approach utilised in this study is qualitative in nature. The present review study primarily examined the period spanning from 2018 to 2024. Based on the findings of this study, DL emerges as a versatile model that necessitates few data alterations and demonstrates superior performance when applied to vast quantities of data.