Retinopathy of prematurity (ROP) is a sight threatening proliferative retinal vascular disease affecting premature infants. Vision loss in ROP is preventable through the early identification and treatment of severe disease. Timely screening and accurate diagnosis is therefore crucial for the diagnosis of ROP, however, multiple challenges exist in current screening processes including limited access to expert ophthalmologists required for ROP screening, subjectivity of diagnosis and cost and time burdens involved in transporting infants to tertiary hospitals. Artificial intelligence (AI) has the potential to overcome current challenges in ROP diagnosis and may transform the way ROP is screened for and managed. Through innovative deep learning technology, a well-designed, well-validated detection algorithm may provide accessible, objective analysis of retinal images to assist expert ophthalmologists in detecting referrable ROP. This thesis will introduce readers to the pathophysiology and grading of ROP, evidence for current treatment guidelines and AI applications in ophthalmology. The systematic review will provide the background evidence into requirements for an accurate, reliable AI algorithm in ROP diagnosis and the validation of our AI algorithm, ROP.AI, will provide insight into the revolutionary diagnostic potential of this deep learning program. Finally, we will discuss future plans for ROP.AI including a methodology proposal to implement the algorithm into a prospective clinical trial for the diagnosis of ROP.