LBP-GLZM Based Hybrid Model for Classification of Breast Cancer

Title

LBP-GLZM Based Hybrid Model for Classification of Breast Cancer

Creator

Mahapatra D.; Umme Salma M.

Description

Classifying mammogram images is difficult because of their complex backgrounds and the differences in resolutions across the images. One of the toughest parts is telling the difference between harmless (benign) and harmful (malignant) tissue. This is hard because the differences between them are incredibly subtle. As a consequence, the distinctive features embedded within tissue patches become not just relevant but critical for the accurate and automatic classification of these images. Traditionally, efforts to automate this classification process have encountered limitations when relying on a singular feature or a restricted set of characteristics. The subtle variations in texture within these images often render such approaches insufficient in achieving high-quality categorization results. Recognizing this, the present investigation undertakes a more comprehensive approach by incorporating distinct feature extraction techniques - specifically, the utilization of Local Binary Pattern (LBP) and Gray Level Zone Matrix (GLZM). These techniques are adept at capturing and delineating the nuanced texture features inherent in mammogram images. By extracting and analyzing these textural nuances, the aim is to construct a hybrid model capable of classifying mammograms into three distinct categories: malignant, benign, and without the necessity for further examination or follow-up. This proposed hybrid model holds significant promise in the field of mammography classification by leveraging the strengths and complementary attributes of multiple feature extraction methods. The integration of LBP and GLZM aims not only to enhance the accuracy of classification but also to improve the robustness of the system in identifying subtle yet crucial differences in tissue textures. Ultimately, the goal is to create a hybrid feature extraction framework that augments the diagnostic capabilities of mammography, providing more precise and reliable categorization of breast tissue for effective medical decision-making and patient care. 2024 IEEE.

Source

Proceedings of ICWITE 2024: IEEE International Conference for Women in Innovation, Technology and Entrepreneurship, pp. 305-310.

Date

2024-01-01

Publisher

Institute of Electrical and Electronics Engineers Inc.

Subject

Augmentation; Breast Cancer; CNN; DDSM; GLZM; LBP

Coverage

Mahapatra D., Christ (Deemed to Be University), Department of Statistics and Data Science, Karnataka, Bengaluru, India; Umme Salma M., Christ (Deemed to Be University), Department of Statistics and Data Science, Karnataka, Bengaluru, India

Rights

Restricted Access

Relation

ISBN: 979-835038328-7

Format

Online

Language

English

Type

Conference paper

Identifier

https://doi.org/10.1109/ICWITE59797.2024.10503412

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85192560889&doi=10.1109%2fICWITE59797.2024.10503412&partnerID=40&md5=4ef67843eb115bce4bb66789fa35b195