Dual-Tree Complex Wavelet Input Transform for Cyst Segmentation in OCT Images Based on a Deep Learning Framework

Reza Darooei; Milad Nazari; Rahele Kafieh; Hossein Rabbani

doi:10.3390/photonics10010011

Dual-Tree Complex Wavelet Input Transform for Cyst Segmentation in OCT Images Based on a Deep Learning Framework

Reza Darooei, Milad Nazari, Rahele Kafieh^*, Hossein Rabbani^*

^*Corresponding author af dette arbejde

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avis › Tidsskriftartikel › Forskning › peer review

1 Citationer (Scopus)

Abstract

Optical coherence tomography (OCT) represents a non-invasive, high-resolution cross-sectional imaging modality. Macular edema is the swelling of the macular region. Segmentation of fluid or cyst regions in OCT images is essential, to provide useful information for clinicians and prevent visual impairment. However, manual segmentation of fluid regions is a time-consuming and subjective procedure. Traditional and off-the-shelf deep learning methods fail to extract the exact location of the boundaries under complicated conditions, such as with high noise levels and blurred edges. Therefore, developing a tailored automatic image segmentation method that exhibits good numerical and visual performance is essential for clinical application. The dual-tree complex wavelet transform (DTCWT) can extract rich information from different orientations of image boundaries and extract details that improve OCT fluid semantic segmentation results in difficult conditions. This paper presents a comparative study of using DTCWT subbands in the segmentation of fluids. To the best of our knowledge, no previous studies have focused on the various combinations of wavelet transforms and the role of each subband in OCT cyst segmentation. In this paper, we propose a semantic segmentation composite architecture based on a novel U-net and information from DTCWT subbands. We compare different combination schemes, to take advantage of hidden information in the subbands, and demonstrate the performance of the methods under original and noise-added conditions. Dice score, Jaccard index, and qualitative results are used to assess the performance of the subbands. The combination of subbands yielded high Dice and Jaccard values, outperforming the other methods, especially in the presence of a high level of noise.

Originalsprog	Engelsk
Artikelnummer	11
Tidsskrift	Photonics
Vol/bind	10
Nummer	1
Antal sider	20
ISSN	2304-6732
DOI	https://doi.org/10.3390/photonics10010011
Status	Udgivet - jan. 2023

Adgang til dokumentet

10.3390/photonics10010011Licens: CC BY

Andre filer og links

Link to publication in Scopus

Citationsformater

@article{557ff26ad68a46a786d2578c759fd167,

title = "Dual-Tree Complex Wavelet Input Transform for Cyst Segmentation in OCT Images Based on a Deep Learning Framework",

abstract = "Optical coherence tomography (OCT) represents a non-invasive, high-resolution cross-sectional imaging modality. Macular edema is the swelling of the macular region. Segmentation of fluid or cyst regions in OCT images is essential, to provide useful information for clinicians and prevent visual impairment. However, manual segmentation of fluid regions is a time-consuming and subjective procedure. Traditional and off-the-shelf deep learning methods fail to extract the exact location of the boundaries under complicated conditions, such as with high noise levels and blurred edges. Therefore, developing a tailored automatic image segmentation method that exhibits good numerical and visual performance is essential for clinical application. The dual-tree complex wavelet transform (DTCWT) can extract rich information from different orientations of image boundaries and extract details that improve OCT fluid semantic segmentation results in difficult conditions. This paper presents a comparative study of using DTCWT subbands in the segmentation of fluids. To the best of our knowledge, no previous studies have focused on the various combinations of wavelet transforms and the role of each subband in OCT cyst segmentation. In this paper, we propose a semantic segmentation composite architecture based on a novel U-net and information from DTCWT subbands. We compare different combination schemes, to take advantage of hidden information in the subbands, and demonstrate the performance of the methods under original and noise-added conditions. Dice score, Jaccard index, and qualitative results are used to assess the performance of the subbands. The combination of subbands yielded high Dice and Jaccard values, outperforming the other methods, especially in the presence of a high level of noise.",

keywords = "deep learning, dual-tree complex wavelet transform (DTCWT), fluid accumulation, optical coherence tomography (OCT), segmentation, subband",

author = "Reza Darooei and Milad Nazari and Rahele Kafieh and Hossein Rabbani",

note = "Publisher Copyright: {\textcopyright} 2022 by the authors.",

year = "2023",

month = jan,

doi = "10.3390/photonics10010011",

language = "English",

volume = "10",

journal = "Photonics",

issn = "2304-6732",

publisher = "MDPI AG",

number = "1",

}

TY - JOUR

T1 - Dual-Tree Complex Wavelet Input Transform for Cyst Segmentation in OCT Images Based on a Deep Learning Framework

AU - Darooei, Reza

AU - Nazari, Milad

AU - Kafieh, Rahele

AU - Rabbani, Hossein

PY - 2023/1

Y1 - 2023/1

N2 - Optical coherence tomography (OCT) represents a non-invasive, high-resolution cross-sectional imaging modality. Macular edema is the swelling of the macular region. Segmentation of fluid or cyst regions in OCT images is essential, to provide useful information for clinicians and prevent visual impairment. However, manual segmentation of fluid regions is a time-consuming and subjective procedure. Traditional and off-the-shelf deep learning methods fail to extract the exact location of the boundaries under complicated conditions, such as with high noise levels and blurred edges. Therefore, developing a tailored automatic image segmentation method that exhibits good numerical and visual performance is essential for clinical application. The dual-tree complex wavelet transform (DTCWT) can extract rich information from different orientations of image boundaries and extract details that improve OCT fluid semantic segmentation results in difficult conditions. This paper presents a comparative study of using DTCWT subbands in the segmentation of fluids. To the best of our knowledge, no previous studies have focused on the various combinations of wavelet transforms and the role of each subband in OCT cyst segmentation. In this paper, we propose a semantic segmentation composite architecture based on a novel U-net and information from DTCWT subbands. We compare different combination schemes, to take advantage of hidden information in the subbands, and demonstrate the performance of the methods under original and noise-added conditions. Dice score, Jaccard index, and qualitative results are used to assess the performance of the subbands. The combination of subbands yielded high Dice and Jaccard values, outperforming the other methods, especially in the presence of a high level of noise.

AB - Optical coherence tomography (OCT) represents a non-invasive, high-resolution cross-sectional imaging modality. Macular edema is the swelling of the macular region. Segmentation of fluid or cyst regions in OCT images is essential, to provide useful information for clinicians and prevent visual impairment. However, manual segmentation of fluid regions is a time-consuming and subjective procedure. Traditional and off-the-shelf deep learning methods fail to extract the exact location of the boundaries under complicated conditions, such as with high noise levels and blurred edges. Therefore, developing a tailored automatic image segmentation method that exhibits good numerical and visual performance is essential for clinical application. The dual-tree complex wavelet transform (DTCWT) can extract rich information from different orientations of image boundaries and extract details that improve OCT fluid semantic segmentation results in difficult conditions. This paper presents a comparative study of using DTCWT subbands in the segmentation of fluids. To the best of our knowledge, no previous studies have focused on the various combinations of wavelet transforms and the role of each subband in OCT cyst segmentation. In this paper, we propose a semantic segmentation composite architecture based on a novel U-net and information from DTCWT subbands. We compare different combination schemes, to take advantage of hidden information in the subbands, and demonstrate the performance of the methods under original and noise-added conditions. Dice score, Jaccard index, and qualitative results are used to assess the performance of the subbands. The combination of subbands yielded high Dice and Jaccard values, outperforming the other methods, especially in the presence of a high level of noise.

KW - deep learning

KW - dual-tree complex wavelet transform (DTCWT)

KW - fluid accumulation

KW - optical coherence tomography (OCT)

KW - segmentation

KW - subband

UR - http://www.scopus.com/inward/record.url?scp=85146796437&partnerID=8YFLogxK

U2 - 10.3390/photonics10010011

DO - 10.3390/photonics10010011

M3 - Journal article

AN - SCOPUS:85146796437

SN - 2304-6732

VL - 10

JO - Photonics

JF - Photonics

IS - 1

M1 - 11

ER -

Dual-Tree Complex Wavelet Input Transform for Cyst Segmentation in OCT Images Based on a Deep Learning Framework

Abstract

Adgang til dokumentet

Andre filer og links

Fingeraftryk

Citationsformater