A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors From Ultrasound Images

Jignesh Chowdary; Pratheepan Yogarajah; Priyanka Chaurasiaa; Velmathi Guriviah

doi:10.1177%2F01617346221075769

A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors From Ultrasound Images

Jignesh Chowdary
, Pratheepan Yogarajah
, Priyanka Chaurasiaa
, Velmathi Guriviah

VIT University

Research output: Contribution to journal › Article › peer-review

59 Citations (Scopus)

228 Downloads (Pure)

Abstract

Breast cancer is one of the most fatal diseases leading to the death of several women across the world. But early diagnosis of breast cancer can help to reduce the mortality rate. So an efficient multi-task learning approach is proposed in this work for the automatic segmentation and classification of breast tumors from ultrasound images. The proposed learning approach consists of an encoder, decoder, and bridge blocks for segmentation and a dense branch for the classification of tumors. For efficient classification, multi-scale features from different levels of the network are used. Experimental results show that the proposed approach is able to enhance the accuracy and recall of segmentation by 1.08%, 4.13%, and classification by 1.16%, 2.34%, respectively than the methods available in the literature.

Original language	English
Pages (from-to)	3-12
Number of pages	10
Journal	Ultrasonic Imaging
Volume	44
Issue number	1
Early online date	30 Jan 2022
DOIs	https://doi.org/10.1177%2F01617346221075769
Publication status	Published (in print/issue) - 7 Feb 2022

Bibliographical note

Publisher Copyright:
© The Author(s) 2022.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Multi-task learning
breast cancer
malignant
classification
segmentation
benign
multi-task learning

Access to Document

10.1177%2F01617346221075769Licence: CC BY-NC

01617346221075769Final published version, 1.32 MBLicence: CC BY-NC

Cite this

@article{a3a9b91e66924d33b751b0d7fbff83ac,

title = "A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors From Ultrasound Images",

abstract = "Breast cancer is one of the most fatal diseases leading to the death of several women across the world. But early diagnosis of breast cancer can help to reduce the mortality rate. So an efficient multi-task learning approach is proposed in this work for the automatic segmentation and classification of breast tumors from ultrasound images. The proposed learning approach consists of an encoder, decoder, and bridge blocks for segmentation and a dense branch for the classification of tumors. For efficient classification, multi-scale features from different levels of the network are used. Experimental results show that the proposed approach is able to enhance the accuracy and recall of segmentation by 1.08\%, 4.13\%, and classification by 1.16\%, 2.34\%, respectively than the methods available in the literature.",

keywords = "Multi-task learning, breast cancer, malignant, classification, segmentation, benign, multi-task learning",

author = "Jignesh Chowdary and Pratheepan Yogarajah and Priyanka Chaurasiaa and Velmathi Guriviah",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2022.",

year = "2022",

month = feb,

day = "7",

doi = "10.1177\%2F01617346221075769",

language = "English",

volume = "44",

pages = "3--12",

journal = "Ultrasonic Imaging",

issn = "0161-7346",

publisher = "SAGE Publications Inc.",

number = "1",

}

TY - JOUR

T1 - A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors From Ultrasound Images

AU - Chowdary, Jignesh

AU - Yogarajah, Pratheepan

AU - Chaurasiaa, Priyanka

AU - Guriviah, Velmathi

N1 - Publisher Copyright: © The Author(s) 2022.

PY - 2022/2/7

Y1 - 2022/2/7

N2 - Breast cancer is one of the most fatal diseases leading to the death of several women across the world. But early diagnosis of breast cancer can help to reduce the mortality rate. So an efficient multi-task learning approach is proposed in this work for the automatic segmentation and classification of breast tumors from ultrasound images. The proposed learning approach consists of an encoder, decoder, and bridge blocks for segmentation and a dense branch for the classification of tumors. For efficient classification, multi-scale features from different levels of the network are used. Experimental results show that the proposed approach is able to enhance the accuracy and recall of segmentation by 1.08%, 4.13%, and classification by 1.16%, 2.34%, respectively than the methods available in the literature.

AB - Breast cancer is one of the most fatal diseases leading to the death of several women across the world. But early diagnosis of breast cancer can help to reduce the mortality rate. So an efficient multi-task learning approach is proposed in this work for the automatic segmentation and classification of breast tumors from ultrasound images. The proposed learning approach consists of an encoder, decoder, and bridge blocks for segmentation and a dense branch for the classification of tumors. For efficient classification, multi-scale features from different levels of the network are used. Experimental results show that the proposed approach is able to enhance the accuracy and recall of segmentation by 1.08%, 4.13%, and classification by 1.16%, 2.34%, respectively than the methods available in the literature.

KW - Multi-task learning

KW - breast cancer

KW - malignant

KW - classification

KW - segmentation

KW - benign

KW - multi-task learning

UR - https://www.scopus.com/pages/publications/85124896439

U2 - 10.1177%2F01617346221075769

DO - 10.1177%2F01617346221075769

M3 - Article

SN - 0161-7346

VL - 44

SP - 3

EP - 12

JO - Ultrasonic Imaging

JF - Ultrasonic Imaging

IS - 1

ER -

A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors From Ultrasound Images

Abstract

Bibliographical note

Funding

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this