Joint Segmentation of Multi-Class Hyper-Reflective Foci in Retinal Optical Coherence Tomography Images

Yao, Chenpu; Wang, Meng; Zhu, Weifang; Huang, Haifan; Shi, Fei; Chen, Zhongyue; Wang, Lianyu; Wang, Tingting; Zhou, Yi; Peng, Yuanyuan; Zhu, Liangjiu; Chen, Haoyu; Chen, Xinjian

doi:10.1109/tbme.2021.3115552

Cited by 16 publications

(3 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They obtained an F1-measure score of 70.7% on 33 SD-OCT volumes from DR patients. Yao et al [31] developed a global information fusion and dual decoder collaboration-based network (GD-Net). It could aggregate global semantic information effectively and learn the semantic correlation between multiclass HRF that contained hard EXs and microglia.…”

Section: Hrf Segmentationmentioning

confidence: 99%

Automatic exudate and aneurysm segmentation in OCT images using UNET++ and hyperreflective-foci feature based bagged tree ensemble

Tanthanathewin,

Wongrattanapipat,

Khaing

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

Diabetic retinopathy’s signs, such as exudates (EXs) and aneurysms (ANs), initially develop from under the retinal surface detectable from optical coherence tomography (OCT) images. Detecting these signs helps ophthalmologists diagnose DR sooner. Detecting and segmenting exudates (EXs) and aneurysms (ANs) in medical images is challenging due to their small size, similarity to other hyperreflective regions, noise presence, and low background contrast. Furthermore, the scarcity of public OCT images featuring these abnormalities has limited the number of studies related to the automatic segmentation of EXs and ANs, and the reported performance of such studies has not been satisfactory. This work proposes an efficient algorithm that can automatically segment these anomalies by improving key steps in the process. The potential area where these hyper-reflective EXs and ANs occur was scoped by our method using a deep-learning U-Net++ program. From this area, the candidates for EX-AN were segmented using the adaptive thresholding method. Nine features based on appearances, locations, and shadow markers were extracted from these candidates. They were trained and tested using bagged tree ensemble classifiers to obtain only EX-AN blobs. The proposed method was tested on a collection of a public dataset comprising 80 images with hand-drawn ground truths. The experimental results showed that our method could segment EX-AN blobs with average recall, precision, and F1-measure as 87.9%, 86.1%, and 87.0%, respectively. Its F1-measure drastically outperformed two comparative methods, binary thresholding and watershed (BT-WS) and adaptive thresholding with shadow tracking (AT-ST), by 78.0% and 82.1%, respectively.

show abstract

Section: Hrf Segmentationmentioning

confidence: 99%

Automatic exudate and aneurysm segmentation in OCT images using UNET++ and hyperreflective-foci feature based bagged tree ensemble

Tanthanathewin,

Wongrattanapipat,

Khaing

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…The segmentation of hyper-reflective foci (HRF) in retinal OCT images, which is essential for the diagnosis of retinal disorders, was the subject of the research in [10]. They suggested the GD-Net, which combined dual decoder cooperation with global information fusion.…”

Section: Review Of Existing Modelsmentioning

confidence: 99%

An In-Depth Statistical Review of Retinal Image Processing Models from a Clinical Perspective

Jagadale Sachin Mohan

2023

IJRITCC

View full text Add to dashboard Cite

The burgeoning field of retinal image processing is critical in facilitating early diagnosis and treatment of retinal diseases, which are amongst the leading causes of vision impairment globally. Despite rapid advancements, existing machine learning models for retinal image processing are characterized by significant limitations, including disparities in pre-processing, segmentation, and classification methodologies, as well as inconsistencies in post-processing operations. These limitations hinder the realization of accurate, reliable, and clinically relevant outcomes. This paper provides an in-depth statistical review of extant machine learning models used in retinal image processing, meticulously comparing them based on their internal operating characteristics and performance levels. By adopting a robust analytical approach, our review delineates the strengths and weaknesses of current models, offering comprehensive insights that are instrumental in guiding future research and development in this domain. Furthermore, this review underscores the potential clinical impacts of these models, highlighting their pivotal role in enhancing diagnostic accuracy, prognostic assessments, and therapeutic interventions for retinal disorders. In conclusion, our work not only bridges the existing knowledge gap in the literature but also paves the way for the evolution of more sophisticated and clinically-aligned retinal image processing models, ultimately contributing to improved patient outcomes and advancements in ophthalmic care.

show abstract

“…Macular edema can cause visual impairment or even visual loss, and is secondary to retinal diseases 3 . Optical coherence tomography (OCT) has the characteristics of high resolution, non-invasive, and non-contact, and has become an indispensable imaging method in the diagnosis and treatment of macular diseases 4 , 5 . OCT can identify three types of liquid lesions, namely, intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED) 6 .…”

Section: Introductionmentioning

confidence: 99%

A deep network embedded with rough fuzzy discretization for OCT fundus image segmentation

Chen

Zeng

Lin

2023

Sci Rep

View full text Add to dashboard Cite

The noise and redundant information are the main reasons for the performance bottleneck of medical image segmentation algorithms based on the deep learning. To this end, we propose a deep network embedded with rough fuzzy discretization (RFDDN) for OCT fundus image segmentation. Firstly, we establish the information decision table of OCT fundus image segmentation, and regard each category of segmentation region as a fuzzy set. Then, we use the fuzzy c-means clustering to get the membership degrees of pixels to each segmentation region. According to membership functions and the equivalence relation generated by the brightness attribute, we design the individual fitness function based on the rough fuzzy set, and use a genetic algorithm to search for the best breakpoints to discretize the features of OCT fundus images. Finally, we take the feature discretization based on the rough fuzzy set as the pre-module of the deep neural network, and introduce the deep supervised attention mechanism to obtain the important multi-scale information. We compare RFDDN with U-Net, ReLayNet, CE-Net, MultiResUNet, and ISCLNet on the two groups of 3D retinal OCT data. RFDDN is superior to the other five methods on all evaluation indicators. The results obtained by ISCLNet are the second only inferior to those obtained by RFDDN. DSC, sensitivity, and specificity of RFDDN are evenly 3.3%, 2.6%, and 7.1% higher than those of ISCLNet, respectively. HD95 and ASD of RFDDN are evenly 6.6% and 19.7% lower than those of ISCLNet, respectively. The experimental results show that our method can effectively eliminate the noise and redundant information in Oct fundus images, and greatly improve the accuracy of OCT fundus image segmentation while taking into account the interpretability and computational efficiency.

show abstract

Joint Segmentation of Multi-Class Hyper-Reflective Foci in Retinal Optical Coherence Tomography Images

Cited by 16 publications

References 48 publications

Automatic exudate and aneurysm segmentation in OCT images using UNET++ and hyperreflective-foci feature based bagged tree ensemble

Automatic exudate and aneurysm segmentation in OCT images using UNET++ and hyperreflective-foci feature based bagged tree ensemble

An In-Depth Statistical Review of Retinal Image Processing Models from a Clinical Perspective

A deep network embedded with rough fuzzy discretization for OCT fundus image segmentation

Contact Info

Product

Resources

About