Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning

Qizhen Sun; Zhun Xie; Jiaqi Han; Yiqi Cai; Lijun Xu; Jianguo Ma

doi:10.1109/IUS51837.2023.10308082

Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning

Qizhen Sun^*
, Zhun Xie
, Jiaqi Han
, Yiqi Cai
, Lijun Xu
, Jianguo Ma

^*Corresponding author for this work

School of Instrumentation and Optoelectronic Engineering

Beihang University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Computer-aided diagnosis is important for breast ultrasound evaluation, as it can enhance diagnostic efficiency and objectivity. However, the performance of such systems, particularly deep learning methods, has been hindered by the limited resolution and interpretability of traditional B-mode ultrasound. We hypothesized that ultrasonic spectral information can serve as a representative feature for differentiating tissues. By leveraging machine learning to recognize tissue spectral patterns, we differentiated tissues in breast ultrasound with an AUCROC of 0.832 and enabled the color enhancement of the malignant probability in lesion areas. Moreover, we explored the integration of tissue spectral pattern recognition with deep learning vision models, resulting in a specificity of 0.896 and an AUCROC of 0.873 for lesion classification, representing a significant improvement.

Original language	English
Title of host publication	IUS 2023 - IEEE International Ultrasonics Symposium, Proceedings
Publisher	IEEE Computer Society
ISBN (Electronic)	9798350346459
DOIs	https://doi.org/10.1109/IUS51837.2023.10308082
State	Published - 2023
Event	2023 IEEE International Ultrasonics Symposium, IUS 2023 - Montreal, Canada Duration: 3 Sep 2023 → 8 Sep 2023

Publication series

Name	IEEE International Ultrasonics Symposium, IUS
ISSN (Print)	1948-5719
ISSN (Electronic)	1948-5727

Conference

Conference	2023 IEEE International Ultrasonics Symposium, IUS 2023
Country/Territory	Canada
City	Montreal
Period	3/09/23 → 8/09/23

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

RF signals
breast ultrasound
deep learning
spectral patterns

Access to Document

10.1109/IUS51837.2023.10308082

Cite this

@inproceedings{fc64398048924d10a6a14474ef66f215,

title = "Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning",

abstract = "Computer-aided diagnosis is important for breast ultrasound evaluation, as it can enhance diagnostic efficiency and objectivity. However, the performance of such systems, particularly deep learning methods, has been hindered by the limited resolution and interpretability of traditional B-mode ultrasound. We hypothesized that ultrasonic spectral information can serve as a representative feature for differentiating tissues. By leveraging machine learning to recognize tissue spectral patterns, we differentiated tissues in breast ultrasound with an AUCROC of 0.832 and enabled the color enhancement of the malignant probability in lesion areas. Moreover, we explored the integration of tissue spectral pattern recognition with deep learning vision models, resulting in a specificity of 0.896 and an AUCROC of 0.873 for lesion classification, representing a significant improvement.",

keywords = "RF signals, breast ultrasound, deep learning, spectral patterns",

author = "Qizhen Sun and Zhun Xie and Jiaqi Han and Yiqi Cai and Lijun Xu and Jianguo Ma",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 2023 IEEE International Ultrasonics Symposium, IUS 2023 ; Conference date: 03-09-2023 Through 08-09-2023",

year = "2023",

doi = "10.1109/IUS51837.2023.10308082",

language = "英语",

series = "IEEE International Ultrasonics Symposium, IUS",

publisher = "IEEE Computer Society",

booktitle = "IUS 2023 - IEEE International Ultrasonics Symposium, Proceedings",

address = "美国",

}

Sun, Q, Xie, Z, Han, J, Cai, Y, Xu, L & Ma, J 2023, Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning. in IUS 2023 - IEEE International Ultrasonics Symposium, Proceedings. IEEE International Ultrasonics Symposium, IUS, IEEE Computer Society, 2023 IEEE International Ultrasonics Symposium, IUS 2023, Montreal, Canada, 3/09/23. https://doi.org/10.1109/IUS51837.2023.10308082

Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning. / Sun, Qizhen; Xie, Zhun; Han, Jiaqi et al.
IUS 2023 - IEEE International Ultrasonics Symposium, Proceedings. IEEE Computer Society, 2023. (IEEE International Ultrasonics Symposium, IUS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning

AU - Sun, Qizhen

AU - Xie, Zhun

AU - Han, Jiaqi

AU - Cai, Yiqi

AU - Xu, Lijun

AU - Ma, Jianguo

PY - 2023

Y1 - 2023

N2 - Computer-aided diagnosis is important for breast ultrasound evaluation, as it can enhance diagnostic efficiency and objectivity. However, the performance of such systems, particularly deep learning methods, has been hindered by the limited resolution and interpretability of traditional B-mode ultrasound. We hypothesized that ultrasonic spectral information can serve as a representative feature for differentiating tissues. By leveraging machine learning to recognize tissue spectral patterns, we differentiated tissues in breast ultrasound with an AUCROC of 0.832 and enabled the color enhancement of the malignant probability in lesion areas. Moreover, we explored the integration of tissue spectral pattern recognition with deep learning vision models, resulting in a specificity of 0.896 and an AUCROC of 0.873 for lesion classification, representing a significant improvement.

AB - Computer-aided diagnosis is important for breast ultrasound evaluation, as it can enhance diagnostic efficiency and objectivity. However, the performance of such systems, particularly deep learning methods, has been hindered by the limited resolution and interpretability of traditional B-mode ultrasound. We hypothesized that ultrasonic spectral information can serve as a representative feature for differentiating tissues. By leveraging machine learning to recognize tissue spectral patterns, we differentiated tissues in breast ultrasound with an AUCROC of 0.832 and enabled the color enhancement of the malignant probability in lesion areas. Moreover, we explored the integration of tissue spectral pattern recognition with deep learning vision models, resulting in a specificity of 0.896 and an AUCROC of 0.873 for lesion classification, representing a significant improvement.

KW - RF signals

KW - breast ultrasound

KW - deep learning

KW - spectral patterns

UR - https://www.scopus.com/pages/publications/85178610552

U2 - 10.1109/IUS51837.2023.10308082

DO - 10.1109/IUS51837.2023.10308082

M3 - 会议稿件

AN - SCOPUS:85178610552

T3 - IEEE International Ultrasonics Symposium, IUS

BT - IUS 2023 - IEEE International Ultrasonics Symposium, Proceedings

PB - IEEE Computer Society

T2 - 2023 IEEE International Ultrasonics Symposium, IUS 2023

Y2 - 3 September 2023 through 8 September 2023

ER -

Ultrasonic Spectral Information Promotes Specificity and Visualization of Breast Cancer in Deep Learning

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this