Tools
Tools
Sathiadhas Puvaneswari, Nithya (2022) Detection of Polycystic Ovarian Syndrome using Convolutional Neural Network in conjunction with Transfer Learning Models. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (8MB) | Preview |
Preview |
PDF (Configuration manual)
Download (3MB) | Preview |
Abstract
Polycystic Ovarian Syndrome, commonly abbreviated as PCOS, is a medical ailment found in women aged between 17 to 40 years. It is a reproductive disorder that causes infertility in women during their childbearing age, may cause an irregular menstrual cycle, and may lead to multiple ovarian diseases. This medical disorder is one of the most concerning diseases in their reproductive age and may lead to long-term complications. Considering the disorder’s uncertainty, the treatment remains undiagnosed for extended periods. Hence, a method is needed to diagnose and detect the presence of a growing number of follicles in the ovary so that the disease can be detected at an early stage. With the evolution of technology, deep learning methods have been immensely accepted for medical diagnosis, and varying networks with multiple layers have been adopted to generate desired results using image classification. Despite the fact that the implementation of deep learning algorithms leads to the generation of image classification with high levels of accuracy, it bears the limitation of consuming large network bandwidths, including computational complexity and execution time. Hence, for the implementation of the same, the thesis puts forward a model that could automate diagnosing the disease using ultrasound images of the ovaries. For this purpose, the conceptual theory of Convolutional Neural Network (CNN) is used in conjunction with transfer learning models: EfficientNet-B2 and ResNet-50. In addition to the usage of the dataset, the data augmentation process has also been implemented due to the limited number of available data. On conduction of the experiment, it was observed that the CNN model generated the highest accuracy of 91% compared to EfficientNet-B2 and ResNet-50.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Agarwal, Bharat UNSPECIFIED |
| Uncontrolled Keywords: | CNN; Data Augmentation; Deep learning; EfficientNet-B2; PCOS; ResNet-50 |
| 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: | Tamara Malone |
| Date Deposited: | 25 May 2023 16:10 |
| Last Modified: | 25 May 2023 16:10 |
| URI: | https://norma.ncirl.ie/id/eprint/6650 |
Actions (login required)
 |
View Item |