Automated segmentation of the ciliary muscle in OCT images using fully convolutional networks

Cabeza-Gil, Iulen; Ruggeri, Marco; Chang, Yu-Cherng; Calvo, Begoña; Manns, Fabrice

doi:10.1364/boe.455661

Cited by 9 publications

(8 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional image-segmentation algorithms, such as edge detection-based segmentation, threshold-based segmentation, region-based, and specific filter-based algorithms, cannot readily segment the sub-ciliary muscle boundary and the ciliary muscle tip area because of the weak contrast and blurred boundaries, whereas manual segmentation is time-consuming, laborious, and shows poor stability. 25 Cabeza et al 26 proposed a U-Net with EfficientNetb2 as the backbone to segment the ciliary muscle in OCT images with a CM-IoU of 90.23%, which was slightly higher than ours (88.7%). However, only 1039 relatively clear images were selected TherapeuTic advances in chronic disease from 16,640 images for development of the full convolutional network (FCN).…”

Section: Discussioncontrasting

confidence: 55%

CMS-NET: deep learning algorithm to segment and quantify the ciliary muscle in swept-source optical coherence tomography images

Chen

et al. 2023

Therapeutic Advances in Chronic Disease

View full text Add to dashboard Cite

Background: The ciliary muscle plays a role in changing the shape of the crystalline lens to maintain the clear retinal image during near work. Studying the dynamic changes of the ciliary muscle during accommodation is necessary for understanding the mechanism of presbyopia. Optical coherence tomography (OCT) has been frequently used to image the ciliary muscle and its changes during accommodation in vivo. However, the segmentation process is cumbersome and time-consuming due to the large image data sets and the impact of low imaging quality. Objectives: This study aimed to establish a fully automatic method for segmenting and quantifying the ciliary muscle on the basis of optical coherence tomography (OCT) images. Design: A perspective cross-sectional study. Methods: In this study, 3500 signed images were used to develop a deep learning system. A novel deep learning algorithm was created from the widely used U-net and a full-resolution residual network to realize automatic segmentation and quantification of the ciliary muscle. Finally, the algorithm-predicted results and manual annotation were compared. Results: For segmentation performed by the system, the total mean pixel value difference (PVD) was 1.12, and the Dice coefficient, intersection over union (IoU), and sensitivity values were 93.8%, 88.7%, and 93.9%, respectively. The performance of the system was comparable with that of experienced specialists. The system could also successfully segment ciliary muscle images and quantify ciliary muscle thickness changes during accommodation. Conclusion: We developed an automatic segmentation framework for the ciliary muscle that can be used to analyze the morphological parameters of the ciliary muscle and its dynamic changes during accommodation.

show abstract

Section: Discussioncontrasting

confidence: 55%

CMS-NET: deep learning algorithm to segment and quantify the ciliary muscle in swept-source optical coherence tomography images

Chen

et al. 2023

Therapeutic Advances in Chronic Disease

View full text Add to dashboard Cite

show abstract

“…The mean difference between the PA derived from the segmentation of Ciloctunet and the two examiners was 5.35 µm ( Ciloctunet –SW) and −3.80 µm ( Ciloctunet –TS), respectively, which is smaller than the mean difference of −9.60 µm between the two examiners (SW–TS) and considerably lower than those reported by [ 60 ] for the comparable parameter CMTMAX, derived from segmentation of two examiners (relaxed ciliary muscle: 20 µm, accommodated ciliary muscle: 25 µm). Cabeza-Gil et al report a mean difference of 1.2 µm with a standard deviation of about 23.72 µm between CMTMAX derived from CNN-based segmentations and those performed by a human expert [ 53 ], therefore slightly better than the difference between CNN and human examiners found in this study.…”

Section: Discussioncontrasting

confidence: 51%

“…By leveraging existing datasets from previous studies for training, validation, and testing, Ciloctunet not only proved the feasibility of the automated segmentation of the ciliary muscle in AS-OCT images like a similar approach published recently [ 53 ], but moreover demonstrated to be on par with experienced examiners. Thereby, Ciloctunet enables the analysis of high numbers of images of large study cohorts by avoiding a time-consuming manual segmentation with possible examiner biases.…”

Section: Discussionmentioning

confidence: 92%

“…Furthermore, the Infogain loss has been shown to achieve a better performance than the cross-entropy loss [ 37 ]. Using other loss functions like the Dice coefficient, which works similarly to the IoU metric and allows for dealing with the class imbalances [ 51 ], a focal loss, as suggested by [ 52 ], or a combination of both [ 53 ], could further improve the performance and will be evaluated in the future. An alternative approach could use network architectures tailored to edge or contour detection like the holistically nested edge detection (HED) network [ 54 , 55 ].…”

Section: Discussionmentioning

confidence: 99%

“…However, in this study, the Freiburg U-Net led to better results than HED Future work might evaluate different network architectures, like the DeconvNet, SegNet, DeepLabv3+, Criss-Cross Network (CCNet), or Context Encoding Network (EncNet), for further improving the accuracy of the segmentation. Cabeza-Gil et al recently published a comparison of several CNN architectures (U-Net and LinkNet, both with different backbone structures, like MobileNetv2, Vgg19, and EfficientNetb4) for the segmentation of the ciliary muscle in OCT images [ 53 ] and found the U-Net to have the highest performance compared to the others [ 53 ].…”

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Performance of the Deep Neural Network Ciloctunet, Integrated with Open-Source Software for Ciliary Muscle Segmentation in Anterior Segment OCT Images, Is on Par with Experienced Examiners

Straßer

Wagner

2022

Diagnostics

View full text Add to dashboard Cite

Anterior segment optical coherence tomography (AS-OCT), being non-invasive and well-tolerated, is the method of choice for an in vivo investigation of ciliary muscle morphology and function. The analysis requires the segmentation of the ciliary muscle, which is, when performed manually, both time-consuming and prone to examiner bias. Here, we present a convolutional neural network trained for the automatic segmentation of the ciliary muscle in AS-OCT images. Ciloctunet is based on the Freiburg U-net and was trained and validated using 1244 manually segmented OCT images from two previous studies. An accuracy of 97.5% for the validation dataset was achieved. Ciloctunet’s performance was evaluated by replicating the findings of a third study with 180 images as the test data. The replication demonstrated that Ciloctunet performed on par with two experienced examiners. The intersection-over-union index (0.84) of the ciliary muscle thickness profiles between Ciloctunet and an experienced examiner was the same as between the two examiners. The mean absolute error between the ciliary muscle thickness profiles of Ciloctunet and the two examiners (35.16 µm and 45.86 µm) was comparable to the one between the examiners (34.99 µm). A statistically significant effect of the segmentation type on the derived biometric parameters was found for the ciliary muscle area but not for the selective thickness reading (“perpendicular axis”). Both the inter-rater and the intra-rater reliability of Ciloctunet were good to excellent. Ciloctunet avoids time-consuming manual segmentation, thus enabling the analysis of large numbers of images of ample study cohorts while avoiding possible examiner biases. Ciloctunet is available as open-source.

show abstract

Efficient artificial intelligence approaches for medical image processing in healthcare: comprehensive review, taxonomy, and analysis

Alnaggar,

Jagadale,

Saif

et al. 2024

Artif Intell Rev

View full text Add to dashboard Cite

In healthcare, medical practitioners employ various imaging techniques such as CT, X-ray, PET, and MRI to diagnose patients, emphasizing the crucial need for early disease detection to enhance survival rates. Medical Image Analysis (MIA) has undergone a transformative shift with the integration of Artificial Intelligence (AI) techniques such as Machine Learning (ML) and Deep Learning (DL), promising advanced diagnostics and improved healthcare outcomes. Despite these advancements, a comprehensive understanding of the efficiency metrics, computational complexities, interpretability, and scalability of AI based approaches in MIA is essential for practical feasibility in real-world healthcare environments. Existing studies exploring AI applications in MIA lack a consolidated review covering the major MIA stages and specifically focused on evaluating the efficiency of AI based approaches. The absence of a structured framework limits decision-making for researchers, practitioners, and policymakers in selecting and implementing optimal AI approaches in healthcare. Furthermore, the lack of standardized evaluation metrics complicates methodology comparison, hindering the development of efficient approaches. This article addresses these challenges through a comprehensive review, taxonomy, and analysis of existing AI-based MIA approaches in healthcare. The taxonomy covers major image processing stages, classifying AI approaches for each stage based on method and further analyzing them based on image origin, objective, method, dataset, and evaluation metrics to reveal their strengths and weaknesses. Additionally, comparative analysis conducted to evaluate the efficiency of AI based MIA approaches over five publically available datasets: ISIC 2018, CVC-Clinic, 2018 DSB, DRIVE, and EM in terms of accuracy, precision, Recall, F-measure, mIoU, and specificity. The popular public datasets and evaluation metrics are briefly described and analyzed. The resulting taxonomy provides a structured framework for understanding the AI landscape in healthcare, facilitating evidence-based decision-making and guiding future research efforts toward the development of efficient and scalable AI approaches to meet current healthcare needs.

show abstract

Automated segmentation of the ciliary muscle in OCT images using fully convolutional networks

Cited by 9 publications

References 42 publications

CMS-NET: deep learning algorithm to segment and quantify the ciliary muscle in swept-source optical coherence tomography images

CMS-NET: deep learning algorithm to segment and quantify the ciliary muscle in swept-source optical coherence tomography images

Performance of the Deep Neural Network Ciloctunet, Integrated with Open-Source Software for Ciliary Muscle Segmentation in Anterior Segment OCT Images, Is on Par with Experienced Examiners

Efficient artificial intelligence approaches for medical image processing in healthcare: comprehensive review, taxonomy, and analysis

Contact Info

Product

Resources

About