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

Quantification of the Anterior-Centripetal Movement of the Ciliary Muscle During Accommodation Using Dynamic OCT Imaging

Purpose

Although the lens undoubtedly plays a major role in presbyopia, altered lens function could be in part secondary to age-related changes of the ciliary muscle. Ciliary muscle changes with accommodation have been quantified using optical coherence tomography, but so far these studies have been limited to quantifying changes in ciliary muscle thickness, mostly at static accommodative states. Quantifying ciliary muscle thickness changes does not effectively capture the dynamic anterior-centripetal movement of the ciliary muscle during accommodation. To address this issue, we present a method to quantify the movement of the ciliary muscle during accommodation using trans-scleral optical coherence tomography images obtained dynamically.

Methods

An image processing framework including distortion correction, geometric transformation, and Procrustes analysis, was used to quantify the anterior-centripetal movement of the ciliary muscle apex and centroid during accommodation. The method was applied in a preliminary study to quantify ciliary muscle displacement and its relation to lens thickness change with accommodation on two young adults and two prepresbyopes.

Results

The magnitude and the direction relative to the pupil plane of the apex/centroid displacement in response to a two diopters (2D) stimulus were 0.16/0.20 mm at 11.3°/30.5° and 0.26/0.34 mm at 6.6°/33.2° for the young adults and 0.20/0.20 mm at 29.7°/40.6° and 0.24/0.40 mm at 33.0°/31.7° for the prepresbyopes, respectively.

Conclusions

This study demonstrates the feasibility of quantifying dynamic anterior-centripetal movement of the ciliary muscle during accommodation using optical coherence tomography. The method better captures the functional response of the muscle than the quantification of thickness changes.

Translational Relevance

We provide a method that holds potential to better understand the age-related changes of the ciliary muscle on presbyopia.

The effect of aging on the ciliary muscle and its potential relationship with presbyopia: a literature review

Background

The ciliary muscle is known to play a part in presbyopia, but the mechanism has not received a comprehensive review, which this study aims to achieve. We examined relevant articles published from 1975 through 2022 that explored various properties of the muscle and related tissues in humans and rhesus monkeys. These properties include geometry, elasticity, rigidity, and composition, and were studied using a range of imaging technologies, computer models, and surgical methods. We identified a notable age-related displacement of the ciliary muscle apex that is characterized by anterior and medial shifts, and hypothesized to be primarily attributed to the accrual of connective tissue and tension exerted by the thickening lens. Other factors could also contribute to the movement, particularly the "inward bowing" of the sclera. Another noteworthy observation is that while the ciliary muscle experiences increasing constraint with advancing age due to adjacent anatomical structures, its contractile capacity remains unaltered, alongside the sustained constancy in both the concentration of muscarinic receptors and their binding affinity. Overall, more studies on human ciliary muscle are needed, as it ages differently from that of monkeys' ciliary muscle. These studies should explore other perspectives, including those regarding changes in the physical properties of the tissue and its relationship with other connected tissues.

Methodology

This literature review utilized a systematic methodology to identify and analyze pertinent studies of the presbyopia and ciliary muscles. The approach encompassed a thorough examination of available literature across different academic databases, such as PubMed, Embase, and Cochrane Library.

Results

Many studies have identified age-related thickening in the ciliary muscle and its potential causes, including the heightened deposition of connective tissues and traction exerted by the thickening lens. Notably, these inquiries culminated in the formulation of a geometric theory positing that the morphology of the ciliary muscle and its spatial relationship with adjacent structures exert pivotal influence over the tension exerted on zonular fibers.

Conclusion

The decline in the accommodative response of the muscle is prevalent in advanced age, with reduced mobility likely attributable to the increased stiffness of the Bruch's Membrane-Choroid Complex (BMCC), where the tendons of the ciliary muscle insert, as well as the thickening and stiffening of the lens. Importantly, the ciliary muscle forces do not change with age.

Quantification of scleral changes during dynamic accommodation

The mechanics of accommodation is a complex process that involves multiple intraocular ocular structures. Recent studies suggest that there is deformation of the sclera during accommodation that may also play a role in accommodation, influencing ciliary muscle contraction and contributing to the accommodative response. However, the type and magnitude of the deformations measured varies significantly across studies. We present high-resolution synchronous OCT measurements of the anterior sclera contour and thickness and lens thickness acquired in real-time during accommodative responses to 4D step stimuli. The lens thickness was used as an assessment of objective accommodation. No changes in nasal and temporal anterior scleral contour and scleral thickness were found during accommodation within the precision of our measurements. Our results demonstrate that there are no significant scleral deformations during accommodation.

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

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.

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

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.

Wavelet attention network for the segmentation of layer structures on OCT images

Automatic segmentation of layered tissue is critical for optical coherence tomography (OCT) image analysis. The development of deep learning techniques provides various solutions to this problem, while most existing methods suffer from topological errors such as outlier prediction and label disconnection. The channel attention mechanism is a powerful technique to address these problems due to its simplicity and robustness. However, it relies on global average pooling (GAP), which only calculates the lowest frequency component and leaves other potentially useful information unexplored. In this study, we use the discrete wavelet transform (DWT) to extract multi-spectral information and propose the wavelet attention network (WATNet) for tissue layer segmentation. The DWT-based attention mechanism enables multi-spectral analysis with no complex frequency-selection process and can be easily embedded to existing frameworks. Furthermore, the various wavelet bases make the WATNet adaptable to different tasks. Experiments on a self-collected esophageal dataset and two public retinal OCT dataset demonstrated that the WATNet achieved better performance compared to several widely used deep networks, confirming the advantages of the proposed method.