This study presents a method for quantifying the anterior–centripetal movement of the ciliary muscle during accommodation using dynamic OCT imaging based on Procrustes analysis. The method was demonstrated in a preliminary study on 4 participants. The dynamic analysis allows us to quantify the displacement of the body of the ciliary muscle and its relationship with the change in LT during accommodation. The analysis provides parameters such as the amplitude, direction, time constant, and velocity of the ciliary muscle movement, which are relevant in studies of the biomechanics and neural control of accommodation.
5,6
A larger sample size is required to gain information regarding the response of the ciliary muscle during accommodation and its relation to changes in lens shape. However, our preliminary study helps to demonstrate the knowledge that can be gained using this method and its potential applications in the study of accommodation dynamics. For instance, our preliminary results show that the method allows us to quantify differences in the amplitude and orientation of the ciliary muscle movement and quantify their impact on the amplitude and dynamics of the changes in lens shape. These differences could influence the forces applied to the lens. Further studies utilizing this method on a larger sample size could therefore help determine if there are age-related changes in these parameters that could be a factor in presbyopia.
Previous OCT studies typically quantified the changes in ciliary muscle with accommodation by using thickness measurements at discrete positions, or the thickness profile.
19–27 Quantifying the ciliary muscle thickness can offer valuable insight into the overall change in shape of the muscle or changes at discrete positions, such as the forward and inward movement of the apex.
19,20,26,29 However, thickness measurements do not capture the overall net movement of the body of the ciliary muscle during accommodation quantitatively, which determines the direction of the force of accommodation
28 that releases the resting zonular tension and pulls the choroid forward (the anterior–centripetal movement of the ciliary muscle). This movement of the entire muscle is clearly visible in dynamic UBM images,
10,13,15 and we have previously shown that it can also be visualized using dynamic OCT imaging.
21 Our study demonstrates the feasibility of quantifying this movement from dynamic OCT recordings by measuring the displacement of the apex or centroid of the ciliary muscle contour using Procrustes analysis. The centroid displacement represents a measurement of displacement of the more posterior longitudinal fibers of the ciliary muscle that cannot be captured using thickness measurements. The apex displacement produced using our method is different from the apex displacement predicted from thickness measurements. The Procrustes analysis calculates the displacement of contour points from the deformation of the contour, whereas methods based on thickness define the apex as the point of maximum thickness measured in the transverse direction.
36 In our study, we find that the magnitude of the apex displacement produced using the Procrustes analysis (≤0.13 mm/D) is much more pronounced than the changes in apex thickness (≤0.05 mm/D).
19 We, therefore, expect that the ciliary muscle displacement will provide greater sensitivity in quantifying changes with accommodation or age than thickness measurements.
Overall, the results of our preliminary study are consistent with findings from prior studies using OCT and with the Helmholtz theory of accommodation. As in prior OCT studies, we find an increase in thickness in the anterior region of the ciliary muscle close to the apex and a thinning in the posterior region at 3 mm from the scleral spur.
19,20,26,29 Our change in thickness profile (
Fig. 2B) is also comparable with the findings of Wagner at al.
19 (their Figure 7). Our thickness changes at maximum thickness (0.01 up to 0.05 mm/D) are within the range of previously published studies, which found values of 0.01 up to 0.03 mm/D.
19,20,26,29 In addition, as in our prior study,
21 we find that there is a lag between the response of the muscle and the lens, which is more pronounced in older adults. No obvious lag between the movement of the lens and ciliary processes was observed in gonioscopic video recordings acquired in nonhuman primates during accommodation induced by stimulation of the Edinger-Westphal nucleus stimulated using implanted electrodes. However, this difference could be due to differences in methodology (stimulated response in iridectomized monkeys vs. natural response in human participants). Further studies on a larger population sample are required to confirm the presence of the lag and its age dependence.
In the same age range, the centroid displacements found by Stachs et al.
17 using UBM (0.04 to 0.26 mm) are lower than our results (0.20 to 0.40 mm), even though their study relied on pharmacological stimulation, which is expected to produce more pronounced changes. The reason for these differences is most likely that Stachs et al.
17 truncated the ciliary muscle more anterior, at approximately 3 mm from the scleral spur, that they truncated the ciliary muscle at the same fixed distance from the scleral spur in the relaxed and accommodated states, and that they included the ciliary processes in their contour. In a separate analysis, we found that the magnitude of the centroid displacement decreases by approximately 25% if the OCT images are truncated at 3 mm. The effect of truncation is discussed further elsewhere in this article.
The proposed OCT-based method offers several biometric parameters that could be used to better understand the age-related changes in the ciliary muscle, such as, for example, the displacement of the ciliary muscle apex and centroid. We believe that the most suitable metric will depend on the research question being addressed. It seems the centroid movement might be driven more by the longitudinal fibers, whereas the ciliary muscle apex might be driven more by radial and circular fibers. Apex displacement is more directly related to the interaction with the lens, because the displacement of the apex more directly determines the tension on the zonules. We expect that the combination of multiple metrics will help to better understand the muscle response as a whole. For instance, comparing the centroid displacement with the apex displacement and studying age-related changes could help understand if the ciliary muscle displacement remains constant with age, but apex movement changes, or the age-related changes, are the same.
The proposed method was developed using the iris as a reference plane and Procrustes analysis to relate the shape of the ciliary muscle across different time points during the accommodative process, providing a new approach to measure the anterior–centripetal movement of the ciliary muscle during accommodation in dynamic OCT imaging.
For this preliminary study, we somewhat arbitrarily chose a fixed area of 1.45 mm2 for the Procrustes analysis, because we found that it captured most of the ciliary muscle in the study participants (>85% of the area and >4.0 mm from the scleral spur), except for the distal end. The position of the centroid may be shifted if instead we would have included the entire ciliary muscle all the way to the posterior insertion zonule. In addition, the calculated centroid displacement will be affected by interindividual variations in the shape of the muscle or the size of the truncated area. For instance, with a fixed truncated area, we expect that we will find a smaller overall displacement with a more pronounced inward component for a shorter muscle relative to a longer muscle. However, in a separate analysis, we found that these effects can be largely compensated for by expressing the displacement in percent of the initial length. Further analyses on a larger sample will help to quantify the effect of interindividual variations and determine the optimal approach to quantify displacement (e.g., relative vs. absolute, size of the area).
Acquiring dynamic images of the ciliary muscle during accommodation has also the advantage of ensuring that the same region of the ciliary muscle is imaged at different accommodative states during the same imaging session, whereas imaging of the ciliary muscle at static accommodative states may introduce misalignment errors owing to the need of acquiring images in different sessions.
21 In our method, the greatest source of error is the variability of the segmentation at the ciliary muscle's apex owing to the lower OCT image contrast in this region.
30,32 Our previous studies show that the variability of ciliary muscle thickness measurement in the region of the apex is on the order of ±45 µm. This is a known limitation that could be addressed by increasing the signal to noise ratio of the images, for instance by using a high incident power.
31 The maximum power of the research-grade commercial spectral domain OCT system used in the present study was 4 mW, well below the exposure limit at 1300 nm. In the two participants (1 and 4) where the images of the ciliary muscle apex had high quality, the variability of the measurements was minimal.