Symptomatic Characteristics and Screening of Retinopathy of Prematurity: A Systematic Review


  • Elizabeth Ndunge Mutua School of Computing & Engineering Sciences, Strathmore university, Kenya
  • Bernard Shibwabo Kasamani School of Computing & Engineering Sciences, Strathmore university, Kenya.
  • Christoph Reich Institute for Data Science, Cloud Computing and IT Security, Furtwangen university, Germany.


ROP Classification, Screening Procedure, ROP Diagnosis, Birth Weight, Gestational Age


Background: Retinopathy of Prematurity (ROP), is a retina vascular disorder which affects infants. ROP occurs because of immature retina tissues which develop abnormally. The diagnosis of the disease involves an ophthalmologist conducting an examination of the retina to identify the presence of the abnormally growing vessels. Deep Learning applications have widely been developed for ROP disease diagnosis and their development suffers numerous challenges. Aim: To establish current global ROP statistics, ROP disease screening guidelines within Africa, the current rate of preterm births in Africa and the challenges associated with Deep Learning applications for the disease diagnosis. Methods: A database search (Ovid, PubMed, Science Direct, Embase, Web of Science, AJOL, MEDLINE, CINAHL, Google Scholar) was performed to identify relevant articles that were published up to October 2022. PRISMA guidelines of academic papers review were followed to ensure that all included articles had reported their study methodological elements as well as meeting their objectives where forty-five articles were included. Results and Conclusions: The results reveal that there are few studies published from the period 2018 to 2022. South Africa and Kenya are the only countries in Africa with national ROP screening guidelines. Deep Learning systems can effectively be applied to assist ophthalmologists in accurately diagnosing the disease however, proper development and testing of these systems is required. The information in this review may help to guide future studies on ROP disease screening in Africa as well as pointing out the challenges of Deep Learning systems for the disease diagnosis.