CDNet : Contrastive Disentangled Network for Fine-Grained Image Categorization of Ocular B-Scan Ultrasound
Precise and rapid categorization of images in the B-scan ultrasound modality is vital for diagnosing ocular diseases. Nevertheless, distinguishing various diseases in ultrasound still challenges experienced ophthalmologists. Thus a novel contrastive disentangled network (CDNet) is developed in this work, aiming to tackle the fine-grained image categorization (FGIC) challenges of ocular abnormalities in ultrasound images, including intraocular tumor (IOT), retinal detachment (RD), posterior scleral staphyloma (PSS), and vitreous hemorrhage (VH). Three essential components of CDNet are the weakly-supervised lesion localization module (WSLL), contrastive multi-zoom (CMZ) strategy, and hyperspherical contrastive disentangled loss (HCD-Loss), respectively. These components facilitate feature disentanglement for fine-grained recognition in both the input and output aspects. The proposed CDNet is validated on our ZJU Ocular Ultrasound Dataset (ZJUOUSD), consisting of 5213 samples. Furthermore, the generalization ability of CDNet is validated on two public and widely-used chest X-ray FGIC benchmarks. Quantitative and qualitative results demonstrate the efficacy of our proposed CDNet, which achieves state-of-the-art performance in the FGIC task.
Medienart: |
E-Artikel |
---|
Erscheinungsjahr: |
2023 |
---|---|
Erschienen: |
2023 |
Enthalten in: |
Zur Gesamtaufnahme - volume:27 |
---|---|
Enthalten in: |
IEEE journal of biomedical and health informatics - 27(2023), 7 vom: 01. Juli, Seite 3525-3536 |
Sprache: |
Englisch |
---|
Beteiligte Personen: |
Dan, Ruilong [VerfasserIn] |
---|
Links: |
---|
Themen: |
---|
Anmerkungen: |
Date Completed 03.07.2023 Date Revised 16.11.2023 published: Print-Electronic Citation Status MEDLINE |
---|
doi: |
10.1109/JBHI.2023.3271696 |
---|
funding: |
|
---|---|
Förderinstitution / Projekttitel: |
|
PPN (Katalog-ID): |
NLM356299775 |
---|
LEADER | 01000naa a22002652 4500 | ||
---|---|---|---|
001 | NLM356299775 | ||
003 | DE-627 | ||
005 | 20231226070152.0 | ||
007 | cr uuu---uuuuu | ||
008 | 231226s2023 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.1109/JBHI.2023.3271696 |2 doi | |
028 | 5 | 2 | |a pubmed24n1187.xml |
035 | |a (DE-627)NLM356299775 | ||
035 | |a (NLM)37126620 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a eng | ||
100 | 1 | |a Dan, Ruilong |e verfasserin |4 aut | |
245 | 1 | 0 | |a CDNet |b Contrastive Disentangled Network for Fine-Grained Image Categorization of Ocular B-Scan Ultrasound |
264 | 1 | |c 2023 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a ƒaComputermedien |b c |2 rdamedia | ||
338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
500 | |a Date Completed 03.07.2023 | ||
500 | |a Date Revised 16.11.2023 | ||
500 | |a published: Print-Electronic | ||
500 | |a Citation Status MEDLINE | ||
520 | |a Precise and rapid categorization of images in the B-scan ultrasound modality is vital for diagnosing ocular diseases. Nevertheless, distinguishing various diseases in ultrasound still challenges experienced ophthalmologists. Thus a novel contrastive disentangled network (CDNet) is developed in this work, aiming to tackle the fine-grained image categorization (FGIC) challenges of ocular abnormalities in ultrasound images, including intraocular tumor (IOT), retinal detachment (RD), posterior scleral staphyloma (PSS), and vitreous hemorrhage (VH). Three essential components of CDNet are the weakly-supervised lesion localization module (WSLL), contrastive multi-zoom (CMZ) strategy, and hyperspherical contrastive disentangled loss (HCD-Loss), respectively. These components facilitate feature disentanglement for fine-grained recognition in both the input and output aspects. The proposed CDNet is validated on our ZJU Ocular Ultrasound Dataset (ZJUOUSD), consisting of 5213 samples. Furthermore, the generalization ability of CDNet is validated on two public and widely-used chest X-ray FGIC benchmarks. Quantitative and qualitative results demonstrate the efficacy of our proposed CDNet, which achieves state-of-the-art performance in the FGIC task | ||
650 | 4 | |a Journal Article | |
700 | 1 | |a Li, Yunxiang |e verfasserin |4 aut | |
700 | 1 | |a Wang, Yijie |e verfasserin |4 aut | |
700 | 1 | |a Chen, Xiaodiao |e verfasserin |4 aut | |
700 | 1 | |a Jia, Gangyong |e verfasserin |4 aut | |
700 | 1 | |a Wang, Shuai |e verfasserin |4 aut | |
700 | 1 | |a Ge, Ruiquan |e verfasserin |4 aut | |
700 | 1 | |a Qian, Guiping |e verfasserin |4 aut | |
700 | 1 | |a Jin, Qun |e verfasserin |4 aut | |
700 | 1 | |a Ye, Juan |e verfasserin |4 aut | |
700 | 1 | |a Wang, Yaqi |e verfasserin |4 aut | |
773 | 0 | 8 | |i Enthalten in |t IEEE journal of biomedical and health informatics |d 2013 |g 27(2023), 7 vom: 01. Juli, Seite 3525-3536 |w (DE-627)NLM217081614 |x 2168-2208 |7 nnns |
773 | 1 | 8 | |g volume:27 |g year:2023 |g number:7 |g day:01 |g month:07 |g pages:3525-3536 |
856 | 4 | 0 | |u http://dx.doi.org/10.1109/JBHI.2023.3271696 |3 Volltext |
912 | |a GBV_USEFLAG_A | ||
912 | |a GBV_NLM | ||
951 | |a AR | ||
952 | |d 27 |j 2023 |e 7 |b 01 |c 07 |h 3525-3536 |