Design of automatic follicle detection and ovarian classification system for ultrasound ovarian images

Title

Design of automatic follicle detection and ovarian classification system for ultrasound ovarian images

Creator

Saranya R.; Sridevi R.

Description

Polycystic Ovary Syndrome (PCOS) is a common reproductive and metabolic disorder characterized by an increased number of ovarian follicles. Accurate diagnosis of PCOS requires detailed ultrasound imaging to assess follicles size, number, and position. However, noise often needs to be improved on these images, complicating manual detection for radiologists and leading to potential misidentification. This paper introduces an automated diagnostic system for integration with ultrasound imaging equipment to enhance follicle identification accuracy. The system consists of two main stages: preprocessing and follicle segmentation. Preprocessing employs an adaptive Frost filter to reduce noise, while follicle segmentation utilizes a region-based active contour combined with a modified Otsu method. Unlike the conventional Otsu method, where the threshold value is selected manually, the modified Otsu method automatically selects initial threshold values using an iterative approach. After segmentation, features are extracted from the segmented results. An SVM classifier then categorizes the ovarian image as normal, cystic, or polycystic. Experimental results demonstrate that the proposed methods Follicle Identification Rate is 96.3% and the False Acceptance Rate is 2%, which significantly improves classification accuracy, highlighting its potential advantages for clinical application. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

Source

Multimedia Tools and Applications

Date

2024-01-01

Publisher

Springer

Subject

Accuracy; Active contour; Biomedical instrumentation; Detection of PCOS; Follicle; Modified Otsu; Sensing systems; Ultrasound image

Coverage

Saranya R., Department of Computer Science with Data Analytics, PSG College of Arts & Science, Tamil Nadu, Coimbatore, India; Sridevi R., Department of Computer Science, CHRIST University, Karnataka, Bangalore, India

Rights

Restricted Access

Relation

ISSN: 13807501; CODEN: MTAPF

Format

Online

Language

English

Type

Article

Identifier

https://doi.org/10.1007/s11042-024-20370-0

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85212391699&doi=10.1007%2fs11042-024-20370-0&partnerID=40&md5=0003457aec33606840def110694640fd