Comparative Analysis of Machine Learning Algorithms to Diagnose Polycystic Ovary Syndrome

Abstract

Ovarian problems have become the most common disease among young women, especially polycystic ovarian syndromes which are mainly caused due to the formation of a cyst inside the ovary. It has been estimated that PCOS is found in one in every ten women who are of child-bearing age. So, it has now become a challenge for medical practitioners to effectively diagnose PCOS at an early stage. This has becomes possible to some extent with the introduction of specially designed artificial intelligence which works using deep learning. It is helpful for emerging medical practitioners to detect PCOS in a better way than previously. It is also seen in many cases that artificial intelligence in the field of detection is being used extensively for the early diagnosis of PCOS. The use of specially designed machine learning programs such as extreme learning machines and the introduction of unique optimization techniques like Bayesian optimization enables early prediction and also reduces the random oversampling problems which occur during diagnosis. To enhance the machine learning approaches for the early diagnosis of PCOS, there are unique forms of non-invasive screening tests that prove to be effective in testing a wide range of specially designed novel machine learning approaches for the screening of PCOS-affected patients which show more effective results than early invasive diagnostic tests by processing the images which play a vital role in PCOS detection.

Topics & Concepts