Broadband hyperspectral imaging for breast tumor detection using spectral and spatial information

Esther Kho; Behdad Dashtbozorg; Lisanne L. de Boer; Koen K. van de Vijver; Henricus J.C.M. Sterenborg; Theo J.M. Ruers

doi:10.1364/BOE.10.004496

Broadband hyperspectral imaging for breast tumor detection using spectral and spatial information

Esther Kho^*, Behdad Dashtbozorg, Lisanne L. de Boer, Koen K. van de Vijver, Henricus J.C.M. Sterenborg, Theo J.M. Ruers

^*Corresponding author for this work

Nanobiophysics

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

48 Citations (Scopus)

160 Downloads (Pure)

Abstract

Complete tumor removal during breast-conserving surgery remains challenging due to the lack of optimal intraoperative margin assessment techniques. Here, we use hyperspectral imaging for tumor detection in fresh breast tissue. We evaluated different wavelength ranges and two classification algorithms; a pixel-wise classification algorithm and a convolutional neural network that combines spectral and spatial information. The highest classification performance was obtained using the full wavelength range (450-1650 nm). Adding spatial information mainly improved the differentiation of tissue classes within the malignant and healthy classes. High sensitivity and specificity were accomplished, which offers potential for hyperspectral imaging as a margin assessment technique to improve surgical outcome.

Original language	English
Pages (from-to)	4496-4515
Number of pages	20
Journal	Biomedical optics express
Volume	10
Issue number	9
DOIs	https://doi.org/10.1364/BOE.10.004496
Publication status	Published - 1 Sept 2019

UN SDGs

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

Access to Document

10.1364/BOE.10.004496Licence: CC BY

Kho2019broadbandFinal published version, 3.27 MBLicence: CC BY

Cite this

@article{f9a7bd3afcba425e8fdae6890e17688e,

title = "Broadband hyperspectral imaging for breast tumor detection using spectral and spatial information",

abstract = "Complete tumor removal during breast-conserving surgery remains challenging due to the lack of optimal intraoperative margin assessment techniques. Here, we use hyperspectral imaging for tumor detection in fresh breast tissue. We evaluated different wavelength ranges and two classification algorithms; a pixel-wise classification algorithm and a convolutional neural network that combines spectral and spatial information. The highest classification performance was obtained using the full wavelength range (450-1650 nm). Adding spatial information mainly improved the differentiation of tissue classes within the malignant and healthy classes. High sensitivity and specificity were accomplished, which offers potential for hyperspectral imaging as a margin assessment technique to improve surgical outcome.",

author = "Esther Kho and Behdad Dashtbozorg and {de Boer}, {Lisanne L.} and {van de Vijver}, {Koen K.} and Sterenborg, {Henricus J.C.M.} and Ruers, {Theo J.M.}",

year = "2019",

month = sep,

day = "1",

doi = "10.1364/BOE.10.004496",

language = "English",

volume = "10",

pages = "4496--4515",

journal = "Biomedical optics express",

issn = "2156-7085",

publisher = "The Optical Society",

number = "9",

}

TY - JOUR

T1 - Broadband hyperspectral imaging for breast tumor detection using spectral and spatial information

AU - Kho, Esther

AU - Dashtbozorg, Behdad

AU - de Boer, Lisanne L.

AU - van de Vijver, Koen K.

AU - Sterenborg, Henricus J.C.M.

AU - Ruers, Theo J.M.

PY - 2019/9/1

Y1 - 2019/9/1

N2 - Complete tumor removal during breast-conserving surgery remains challenging due to the lack of optimal intraoperative margin assessment techniques. Here, we use hyperspectral imaging for tumor detection in fresh breast tissue. We evaluated different wavelength ranges and two classification algorithms; a pixel-wise classification algorithm and a convolutional neural network that combines spectral and spatial information. The highest classification performance was obtained using the full wavelength range (450-1650 nm). Adding spatial information mainly improved the differentiation of tissue classes within the malignant and healthy classes. High sensitivity and specificity were accomplished, which offers potential for hyperspectral imaging as a margin assessment technique to improve surgical outcome.

AB - Complete tumor removal during breast-conserving surgery remains challenging due to the lack of optimal intraoperative margin assessment techniques. Here, we use hyperspectral imaging for tumor detection in fresh breast tissue. We evaluated different wavelength ranges and two classification algorithms; a pixel-wise classification algorithm and a convolutional neural network that combines spectral and spatial information. The highest classification performance was obtained using the full wavelength range (450-1650 nm). Adding spatial information mainly improved the differentiation of tissue classes within the malignant and healthy classes. High sensitivity and specificity were accomplished, which offers potential for hyperspectral imaging as a margin assessment technique to improve surgical outcome.

UR - http://www.scopus.com/inward/record.url?scp=85073975400&partnerID=8YFLogxK

U2 - 10.1364/BOE.10.004496

DO - 10.1364/BOE.10.004496

M3 - Article

AN - SCOPUS:85073975400

SN - 2156-7085

VL - 10

SP - 4496

EP - 4515

JO - Biomedical optics express

JF - Biomedical optics express

IS - 9

ER -

Broadband hyperspectral imaging for breast tumor detection using spectral and spatial information

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this