<p>Integrating Deep Learning (DL) algorithms into cancer research has significantly advanced early detection, diagnosis, prognosis and therapeutic decision-making. This review provides a comprehensive analysis of DL applications across multiple cancer types, including liver, cervical, colorectal, lung, breast, leukaemia and prostate cancers. Among various applications of DL, some important ones include the early detection of malignancies, enabling precise tumour and cell characterisation and optimising personalised treatment strategies. A tremendous amount of complex biomedical data in the form of medical imaging, multi-omics, clinical reports, etc., is generated in oncology. AI-driven tools can be trained on these wide ranges of datasets to identify patterns that conventional methods might overlook. While the potential of deep learning (DL) and machine learning (ML) in oncology is substantial, challenges such as the availability of large datasets, model generalisation and maintaining diagnostic consistency emphasise the need for continued innovation and refinement to maximise the impact of AI in clinical settings. To assist researchers in identifying relevant models and techniques, we have organised data in tabular format, presenting the types of datasets used, models employed, techniques used and their respective accuracy levels. This review provides a clear outlook of the most frequently employed DL techniques in oncology, along with a comparative analysis of various tools in terms of accuracy and sensitivity. Besides, we have also outlined prospective directions for future research, emphasising areas with significant potential for impact. This review is an essential resource for researchers seeking to understand the present landscape of AI in oncology and use it in advancing cancer care.</p>