Detection of anaemia using medical images: A comparative study of machine learning algorithms – A systematic literature review

Abstract

BackgroundAnaemia is a global public health challenge that affects children and pregnant women. Anaemia develops when the body's supply of red blood cells declines or when the structure of the cells is weakened. The clinical diagnosis of anaemia has several challenges in practice, including insufficient funding for medical tests, inadequate medical personnel and resources in remote areas, and client reluctance that results in abstinence. Several machine learning techniques for anaemia detection have been developed due to their affordability, simplicity of use, and non-invasive nature as compared to the invasive approach. MethodsWe perform a systematic review and examine current trends and concepts of machine learning in healthcare services to identify viable approaches to detect anaemia. We compare the most successful machine learning algorithms currently in use regarding machine learning, evaluation metrics, image augmentation, and the origin and size of the dataset used. ResultsThe result of this study is a clear indication that non-invasive methods such as the use of machine learning algorithms to detect anaemia are affordable, and provides results on time. ConclusionsThis systematic review provide scientific evidence, with the results describing how effective machine learning could make anaemia detection feasible. Machine and deep learning algorithms are introduced and used to make a wide-ranging analysis of images, diagnosis, and clinical analysis in the disciplines of medical fields such as anaemia detection.