AUTOMATION OF POLYCYSTIC OVARY SYNDROME DIAGNOSTICS THROUGH MACHINE LEARNING ALGORITHMS IN ULTRASOUND IMAGING - research report

Aug 4, 2026

Upload by Maeve

Machine learning enables automation of ultrasound-image interpretation for diagnosing polycystic ovary syndrome (PCOS). Conventional laboratory and instrumental approaches rely on expert assessment of ultrasound images, which can be qualification-dependent and subjective. The study develops a software module using convolutional neural networks (CNNs) to improve diagnostic accuracy and objectivity for polycystic disease manifestations of PCOS. The work highlights CNN interpretability through Grad-CAM visualization to localize image regions driving each prediction.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

INTRODUCTION

Clinical context and prevalence of PCOS

Diagnostic challenges of ultrasound interpretation

Ultrasound criteria used for PCOS assessment

MACHINE LEARNING APPROACH (CNN) AND INTERPRETABILITY

Submitted: 2024-03-11 | Revised: 2024-06-11 | Accepted: 2024-06-12 Keywords: polycystic ovary syndrome, PCOS, ultrasound imaging, Neural Networks, grad-CAM, Python programming Roman GALAGAN [0000-0001-7470-8392]* , Serhiy ANDREIEV[0009-0007-4314-6017]* , Nataliia STELMAKH [0000-0003-1876-2794]**, Yaroslava RAFALSKA [0000-0002-1047-3114]***, Andrii MOMOT [0000-0001-9092-6699]* AUTOMATION OF POLYCYSTIC OVARY SYNDROME DIAGNOSTICS THROUGH MACHINE LEARNING ALGORITHMS IN ULTRASOUND IMAGING Abstract This article presents a study aimed at using machine learning to automate the analysis of ultrasound images in the diagnosis of polycystic ovary syndrome (PCOS). Today, various laboratory and instrumental methods are used to diagnose PCOS, includingthe analysis of ultrasound images performed by medical professionals. The peculiarity of such analysis is that it requires high qualification of medical professionals and can be subjective. The aim of this work is to develop a software module based on convolutional neural networks (CNN), which will improve the accuracy and objectivity of diagnosing polycystic disease as one of the clinical manifestations of PCOS. By using CNNs, which have proven to be effective in image processing and classification, it becomes possible to automate the analysis process and reduce the influence of the human factor on the diagnosis result. The article describes a machine learning model based on CNN architecture, which was proposed by the authors for analyzing ultrasound images in order to determine polycystic disease. In addition, the article emphasizes the importance of the interpretability of the CNN model. For this purpose, the Gradient-weighted Class Activation Mapping (Grad-CAM) visualization method was used, which allows to identify the image areas that most affect the model's decision and provides clear explanations for each individual prediction. 1. INTRODUCTION Polycystic ovary syndrome is one of the most common endocrine pathologies affecting women of reproductive age worldwide. Characterized by hormonal imbalance and the * Igor Sikorsky Kyiv Polytechnic Institute, Faculty of Instrumentation Engineering, Department of Automation and Non-Destructive Testing Systems, Ukraine, [r.galagan@kpi.ua](r.galagan@kpi.ua) ** Igor Sikorsky Kyiv Polytechnic Institute, Faculty of Instrumentation Engineering, Department of ComputerIntegrated Technologies of Device Production, Ukraine *** Bogomolets National Medical University, Pharmaceutical Faculty, Department of Organization and Economy of Pharmacy, Ukraine presence of cysts in the ovaries, this disease causes a variety of complications and health problems (Bulsara et al., 2021) . The prevalence of PCOS is striking, as it affects approximately 5-15% of women of reproductive age (Azziz, 2016; Liu et al., 2021) . Moreover, the range of 5-15% is a kind of averaged range, since different studies give different values. World Health Organization data indicate that 70% of PCOS cases worldwide remain undiagnosed (World Health Organization, 2023) . Despite numerous studies, the exact causes of PCOS are still unknown, but there is a hypothesis that genetic and environmental factors interact (Hoeger et al., 2021) . The impact of PCOS on women's fertility has important social and demographic implications. Women suffering from this syndrome often face problems with natural conception, and many of them need to resort to the use of assisted reproductive technologies, such as in vitro fertilization. In addition, PCOS is associated with the risk of developing long-term diseases such as type 2 diabetes and cardiovascular disease, which can negatively affect the quality and duration of life. Early diagnosis of PCOS plays an important role in further treatment and elimination of the disease's consequences (Garad & Teede, 2020) . PCOS is associated with a disorder of folliculogenesis (Rasquin et al., 2022) . PCOS has a multifactorial nature and is characteri

FAQs

What problem does the study address in PCOS diagnosis?

The study targets the subjectivity and specialist-dependence of ultrasound image interpretation used in PCOS diagnosis, aiming to make the analysis more objective and accurate.

What model is proposed for automated analysis of ultrasound images?

A convolutional neural network (CNN) architecture is used to analyze ultrasound images and determine polycystic disease related to PCOS.

How does the approach explain why the model makes a specific prediction?

Grad-CAM visualization is applied to identify the image areas most affecting the model’s decision, providing interpretable explanations for individual predictions.

Aug 4, 2026

Upload by Maeve