Enhancing PCOS Detection with SRGAN-Generated Synthetic Images and CNN Models

Tambde, Ridima Chetan (2024) Enhancing PCOS Detection with SRGAN-Generated Synthetic Images and CNN Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Polycystic Ovary Syndrome (PCOS) is a common hormonal disorder affecting women throughout the world and is detected using ultrasound scans. Due to this accurate diagnosis crucial for effective treatment, is often hindered due to the limited availability of data. For building reliable models, high-quality data is needed and with recent advancements in artificial intelligence, a concept called data generation where synthetic images are generated, has shown promising results. This study attempts the use of SuperResolution Generative Adversarial Networks (SRGAN) to create synthetic images from existing ones. These synthetic images, along with the original ones, are used to train and test various convolutional neural network (CNN) models, including NasNetMobile, Resnet-152, and Xception. Additionally, hybrid models combining all 3 CNN models with CatBoost are developed and evaluated. The SRGAN architecture is fine-tuned here till good images are obtained and the effectiveness is analyzed to determine their impact on diagnostic performance. Therefore, this research involves a comparison of the classification results from both original images and generated images, thus helping to understand if synthetic data influences the accuracy and reliability of diagnostic models with evaluation metrics like accuracy, precision, recall, F1 score, and the confusion matrix for understanding if any misclassifications. Thus, the study concludes by identifying the most effective model combinations and providing valuable insights for future research in medical imaging.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Yaqoob, Abid UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > RG Gynecology and obstetrics Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	26 Aug 2025 11:41
Last Modified:	26 Aug 2025 11:41
URI:	https://norma.ncirl.ie/id/eprint/8642

Actions (login required)

View Item