Anterior–posterior asymmetry in iris mechanics measured by indentation

Abstract

Indentation and histological analysis of the porcine iris were done to assess the relative stiffness of the anterior (stroma) and posterior (dilator and sphincter) layers. The dimensions of the constituent structures were documented histologically by staining with a monoclonal anti-human α-smooth muscle actin antibody to determine the location of the stroma, sphincter, and dilator. Intact porcine irides (4–8 h post-mortem) were bisected into two equal C-shaped halves to indent both surfaces. Indentation experiments were performed using a 1 mm cylindrical indenter tip. The load–displacement curve for each experiment was used to estimate effective instantaneous and equilibrium moduli for the anterior and posterior surfaces of the tissue. A total of 18 irides (9 pairs) with 3–5 indentations per iris surface was performed. The average thickness of the samples was 550 μm; the indentation depth was limited to 60–100 μm depending on the thickness of the sample at each point. Posterior surface indentation gave larger forces than anterior, with the resulting instantaneous modulus of 6.0 ± 0.6 kPa versus 4.0 ± 0.5 kPa (mean ± 95% CI, n = 45, p < 0.001) and equilibrium modulus of 4.4 ± 0.9 versus 2.3 ± 0.3 (p = 0.007). The stress–relaxation analysis revealed that the anterior surface had a shorter relaxation time (121.31 ± 6.84 s) than the posterior surface (210.61 ± 9.41 s, p = 0.03), perhaps due to the permeability of the stroma. Recognizing that our effective modulus calculations in this study did not account for heterogeneity, viscoelasticity, or poroelasticity, we conclude that the posterior components of the iris – dilator, pigment epithelium, and sphincter – are on average stiffer than the stroma and anterior border layer.

1. Introduction

The interest in understanding iris mechanics arises from the influence of abnormal iris morphology on specific ocular disorders. For example, in angle closure (Epstein et al., 1997), the iris bows anteriorly, and the abnormal shape and position of the iris impede aqueous humor outflow, resulting in increased intraocular pressure. In contrast, pigment dispersion syndrome (Ritch, 2004; Niyadurupola and Broadway, 2008) is associated with posterior bowing of the iris and liberation of pigment due to iris–lens and/or iris–zonule contact. A relatively new ophthalmic disorder, intraoperative floppy iris syndrome (IFIS) also involves iris mechanics. First reported in 2005 by Chang and Campbell (Chang and Campbell, 2005), IFIS has been observed in patients who are taking or have taken Tamsulosin (Flomax) for the urinary complications of benign prostatic hypertrophy. Tamsulosin, a selective α_1A-adrenergic receptor antagonist, has the side effect of inhibiting the iris dilator, and its effects can be irreversible, suggesting that the muscle atrophies from the drug (Takmaz and Can, 2007a,b). This theory has recently been verified structurally by visible alterations of the iris in patients taking Tamsulosin, such that the dilator muscle region measured as half of the distance between the scleral spur and the pupillary margin began to significantly thin with long-term use of the drug (Prata et al., 2009).

The contour of the iris is determined by two factors: external stresses arising from the flow of the aqueous humor, and internal stresses due to the passive and active components defining the structures of the iris. The external stresses can be understood in terms of fluid mechanics of the aqueous humor (Huang and Barocas, 2004; Heys and Barocas, 2002; Silver and Quigley, 2004). The internal stresses are more complex because the iris is a composite tissue (Fig. 1), in which each individual component may have different mechanical properties that can affect the iris’s natural contour.

Fig. 1.

(a) Histological cross-section of the pupillary and mid-peripheral portions of the porcine iris. A monoclonal anti-human α-smooth muscle actin stain was used to differentiate the muscular tissues, including the sphincter (S) and dilator (D). Also identifiable is the stroma (ST), a loosely arranged collagen network, and a layer of pigment epithelium (PE) cells. Magnification of the mid-periphery regions illustrates that sphincter muscle is a larger than the dilator muscle in the porcine model.

The anterior border layer (ABL), principally composed of melanocytes and fibroblasts, is the predominant factor for the visible external color of the iris. Directly beneath the ABL is the stroma, composed of a loosely organized collagenous network of fibroblasts and melanocytes. The stroma is an open mesh containing mucopolysaccharide and fluid with minimal additional tissue (Hogan et al., 1971). The sphincter muscle, lying within the stromal framework, is a circumferentially aligned smooth muscle located at the pupillary margin. The sphincter in the pig is slightly larger (Fig. 1) than in the human. The posterior surface of the iris is composed of the dilator muscle and a layer of modified pigment epithelial cells. The radially aligned dilator extends along the posterior stromal layer from the base of the iris root to the midportion of the sphincter muscle similarly found in the porcine model (Hogan et al., 1971).

Toward our goal of understanding the mechanisms determining the iris contour in vivo, it is imperative that we explore how the different constituents contribute to the overall mechanical behavior of the passive iris. We have recently shown (Whitcomb et al., 2009) that the global stiffness of the ex vivo porcine iris increases with pharmaceutical stimulation, but to our knowledge, no work has been done to determine the relative contributions of the compromising segments of the iris. Our motivation to use indentation to further mechanically characterize the iris was due to its ability to identify individual constituents within composites or heterogenous samples (Ebenstein and Pruitt, 2006) and its depth-sensitivity at very small length scales (Oyen and Cook, 2009). As a preliminary study, we used indentation on the anterior and posterior surfaces of the ex vivo porcine iris to examine mechanical differences. Since indentation is far more sensitive to the tissue properties near the indenter than to those far from the indenter (Costa and Yin, 1999), the asymmetry between the anterior and posterior surfaces was revealed.

2. Materials and methods

2.1. Isolation and preparation of tissue

Eyes were harvested from animals sacrificed by Experimental Surgical Services and the Visible Heart Laboratory^® at the University of Minnesota. The porcine model was used for convenience, rapid post-mortem availability, size, and similarity of anatomical structure to the human eye (Olsen et al., 2002). Enucleated eyes were isolated and refrigerated in a modified Krebs–Ringer (KRB) bicarbonate buffer and the irides were indented within 4–8 h post-mortem. The KRB buffer was composed of the following millimolar composition: 118 NaCl, 25 NaHCO₃, 4.7 KCL, 1.2 KH₂PO₄, 1.2 MgSO₄, 1.25 CaCl₂, and 10.0 glucose and oxygenated with 95% O₂ and 5% CO₂ to maintain a pH of 7.5. Removal and dissection of the iris from the eye globe was done in a manner similar to the dissection technique for penetrating keratoplasty (Vajpayee et al., 1994). First, a small incision was made through the sclera, roughly 1 mm below the edge of the limbus, which was used to cut around the globe perimeter. The anterior portion of the eye, containing the cornea, iris, and lens, was discarded, and the iris was peeled away from the sclera. The irides were isolated and bisected into two equal C-shaped halves, which were glued to a flat aluminum surface with an cyanoacrylate adhesive (Nexaband^®). This procedure was done to enable indentation of the two surfaces – anterior and posterior – using the same iris. After the specimens had been mounted, a thin layer of KRB solution was placed on the exposed surface to limit tissue degradation. The KRB solution was previously observed (Whitcomb et al., 2009) not to activate the smooth muscle cells.

2.2. Histology

The dimensions of the porcine iris structures are not as well documented as those of the human iris (Hogan et al., 1971; Freddo, 1996); therefore, histological images were taken to characterize the thickness of the stroma, dilator, and sphincter to guide our indentation experiments. Monoclonal anti-human α-smooth muscle actin antibody was used to stain the muscular tissues of the porcine iris: the dilator pupillae and sphincter iridis. As seen in Fig. 1, the musculature was stained a darker red compared to the collagenous matter of the stroma. Dimensions of the stroma, sphincter, and dilator were found by taking multiple measurements throughout the cross-sectional length to find an average thickness of the constituent components. Histological cross-sections were examined using a Leica DMR microscope and imaging analysis software Bioquantnova prime (BIOQUANT, Nashville, TN).

2.3. Indentation

A Nanoindenter XP (MTS; Eden Prairie, MN) with a force resolution of 0.02 mN, displacement resolution of 30 nm, and a stainless steel flat-end cylindrical tip of 1 mm diameter was used. Indentation data were collected from paired points on the anterior and posterior surface, as indicated in Fig. 2. The indentations for both surfaces were done near the mid-region of the tissue to minimize edge effects of glueing. Two loading protocols were used for the modulus calculations: a single and double loading step. Each step consisted of loading to 1.5 mN followed by a stress relaxation for 400 s. The loading rate was initially varied, and the optimal loading rate for reproducibility was found to be at 2.0 mN/s. The indentation depth was between 60 and 100 μm to ensure that the upper (i.e., the indented) portion of the tissue had a much larger effect on the indentation results than the lower (i.e., glued surface) portion.

Fig. 2.

Bisected ex vivo iris sections with the anterior surface face up on the left and the posterior surface face up on the right. Irides were indented in the mid-periphery, as indicated by the X.

A larger deformation of 60–100 μm was used to ensure that the response captured was due to the tissue and not due to the thin layer of buffer on the tissue or due to movement of air across the indenter. This depth also enabled a clear distinction between the reaction of the anterior and posterior surfaces, which was more apparent at larger depths. For linear elastic materials of finite thickness a indented by $a$ flat cylindrical tip of radius $b$, the modulus $E_i$ is given by Simha et al. (2007)

$$\frac{dP}{dh}=\frac{2b{E}_i\kappa }{1-{\nu }^2}\Rightarrow E_i=\frac{1-{\nu }^2}{2b\kappa }\frac{dP}{dh}$$ (1)

where $P$ is the indentation force, $v$ is Poisson’s ratio, and $\kappa =\kappa (b/a,\nu )$ is a correction factor for finite sample thickness. The $\kappa$ values are known for a given $v$ (Hayes et al., 1972); we assumed $v$ to be 0.5 for the iris based on our previous experiments (Heys and Barocas, 1999). Table 1 gives the values of the parameters used in Equation (1). Since the iris exhibited a viscoelastic response, Equation (1) cannot define a single modulus. Rather, we calculated the instantaneous modulus $\left(E_0\right)$ from the linear slope of the load–depth curve between 0.5 and 1.0 mN on the first loading ramp and an equilibrium modulus ($E_{\mathrm{\infty }}$) from the load–depth values at the end of the 400 s hold. Using these values for $E_{\mathrm{\infty }}$, a linear fit was found enforcing a zero intercept. Fig. 3 gives a visual illustration of the calculation for the two effective moduli. Both moduli were calculated for indentation of the anterior and the posterior surfaces.

Table 1.

Parameters referenced in Equation (1) for the porcine iris.

Parameter	Definition	Value
P	Indenter force	0–1 mN
h	Indenter depth	0–100 μm
a	Sample thickness	200–750 μm
ν	Poisson’s ratio	0.5 (Heys and Barocas, 1999)
b	Indenter radius	0.5 mm
κ	Finite-thickness correction	2.2–6.2 (Hayes et al., 1972)

Fig. 3.

A representation of the calculation used to determine the effective instantaneous ($E_0$) and equilibrium ($E_{\mathrm{\infty }}$) moduli for the iris. The slope of the load–depth curve between 0.5 and 1.0 mN on the first ramp was used to calculate $E_0$. The load–depth values after the two 400 s holds was extrapolated to zero and fit linearly, assuming the tissues long-term behavior was linear, to determine $E_{\mathrm{\infty }}$.

Since the stroma is composed of a network of loosely organized collagen and interstitial fluid it does not behave strictly elastically; therefore, viscoelastic analysis was also done via a two-relaxation-time linear solid model.

$$R\left(t\right)=E_1{e}^{\left(-t/{\tau }_1\right)}+E_2{e}^{\left(-t/{\tau }_2\right)}+E_3$$ (2)

This relaxation function is similar to the Weichert model (Machiraju et al., 2006), and assumes the material to behave as two spring-dashpot elements in parallel with a spring, resulting in a sum of exponentials. The time constant ${\tau }_i(i=1,2)$ represents the dynamic relaxation time of the $i$th dashpot, and $E_i(i=1,2,3)$ represents the modulus of the $i$th spring. The load $P\left(t\right)$ for any displacement history $h\left(t\right)$ can be obtained using the following integral representation:

$$P(t)={\int }_0^{t}R(t-\xi )\frac{\partial }{\partial \xi }h(\xi )d\xi$$ (3)

Chiravarambath (2006) developed a discrete method that deduces the relaxation function (Equation (2)) from a non-ideal step displacement, such as arises from the slow time response of the Nanoindenter XP. In discrete form, Equation (3) can be written as

$$P\left(t_n\right)=\sum _{i=0}^{n-1}R\left(t_n-t_i\right)\left(h_{i+1}-h_i\right)$$ (4)

The displacement at any time $t$ can be expressed as a sum of $n$ discrete small displacement steps. The iris was allowed to relax under each displacement step according to $R\left(t\right)$. Thus, the load at any time $t_n$ due to $n$ displacement steps was computed from the summation in Equation (4). The parameters for the relaxation function in Equation (2) were calculated by minimizing the error from the experimental load values $P_e\left(t_n\right)$ and the discrete load representation in Equation (4).

$$er=\sum _{n=1}^N{\left[P_e\left(t_n\right)-\sum _{n=0}^{n-1}R\left(t_n-t_i\right)\left(h_{i+1}-h_i\right)\right]}^2$$ (5)

2.4. Statistical analysis

The differences between the mechanical responses of the anterior and posterior iris surfaces were quantified using Student t-tests for paired data with equal variance by pairing each indentation point as an independent measure on the same iris. Indentation responses were not reliable for tissue more than 10 h post-mortem, so only tissue less than 10 h post-mortem was used in the analysis. Statistical analysis was performed using Matlab and Origin software (OriginLab, Northampton, MA).

3. Results

3.1. Histology

A total of 25 histological cross-sections per iris with a thickness of 20 μm each were used to determine the dimensions of the iris components. The 25 slides per iris were chosen randomly, but were located in the mid-section of the iris to ensure that accurate dimensions of the tissues were captured. The average thickness of the dilator (D) was 26.0 ± 1.2 μm (mean ± 95% CI, n = 100 cross-sectional locations), sphincter (S) was 133.7 ± 2.0 μm, central stroma plus sphincter (CSeS) was 430.6 ± 20.9 μm, and peripheral stroma (PS) was 735.8 ± 19.7 μm, shown graphically in Fig. 4. The average length of the sphincter from the pupillary margin was 2445.1 ± 133.6 μm. It is important to note that during the fixation process shrinkage did occur, especially in the stroma, and the measured dimensions of the iris thickness may differ in vivo.

Fig. 4.

(a) Graphical cartoon representation of the thickness of the constituent components of the tissue; peripheral stroma (PS), dilator (D), central stroma plus sphincter (CS–S), and the sphincter (S). (b) The values represent the mean thickness of the components marked by histological staining from 2 pairs of eyes. A total of 100 (25 per eye) cross-sections of thickness 20 μm were used. The error bars represent a 95% CI.

3.2. Indentation

A total of 18 irides (9 pairs) with 3–5 indentations per iris surface were used for analysis. The thickness ranged between 200 and 600 μm at each individual indentation point, averaging 550 μm at the mid-periphery where the tissue was indented. The thickness of the tissue was measured as a depth difference between the depth of the aluminum surface and the depth of the iris surface at each indentation. Fig. 5a shows typical force–displacement curves for the anterior verses posterior indentation. Both curves show an initial toe region of low force, followed by higher forces with steeper slopes as the indenter depth increases. The toe regions were not significantly different between the anterior and posterior indentations, but the curves separated as the indentation depth increased. Fig. 5b shows the modulus measurements; the instantaneous effective modulus $\left(E_0\right)$ for the porcine iris was 1.5 times higher on the posterior surface compared to the anterior surface (p < 0.001). The posterior effective instantaneous modulus was 6.0 ± 0.6 kPa (mean ± 95% CI, n = 45) in contrast to the anterior value of 4.0 ± 0.5 kPa. Asymmetry was also observed in the equilibrium effective modulus, $E_{\mathrm{\infty }}$. The equilibrium effective modulus was 1.9 times as high on the posterior compared to the anterior surface (4.4 ± 0.9 kPa versus 2.3 ± 0.3 kPa, p = 0.007). There was no significant difference between the right and left iris samples on either the posterior (p = 0.40) or anterior (p = 0.56) surface indentations. The intrasample variation for the anterior indentations averaged 1.22 (measured by SD), and the posterior intersample variation was 1.53, suggesting little variability occurred between indentation locations.

Fig. 5.

(a) Typical load–displacement curve at a constant load rate of 2.0 mN/s for the anterior (square) and posterior (circle) surfaces. The slope of the linear region between 0.5 and 1.0 mN was used to calculate $E_0$. (b) Effective moduli for the anterior and posterior surfaces, the error bars represent 95% CI (n = 45 for $E_0$ and n = 28 for $E_{\mathrm{\infty }}$).

Fig. 6 shows a typical stress–relaxation curve for the two surfaces, including a typical fit minimizing the error in Equation (5) (Fig. 6c). In Equation (2), τ₁ captured the initial rapid relaxation, whereas τ₂ described the slower time response. For the anterior surface τ₁ was 6.12 ± 0.62 s, and for the posterior it was 10.32 ± 0.91 s (n = 21, p < 0.05). The larger time constant τ₂ also showed a statistically significant difference between the surfaces, 121.31 ± 6.84 s for the anterior and 210.61 ± 9.41 s for the posterior (p < 0.03). There was no significant difference in the relaxation (p < 0.03) moduli $E_1$ and $E_3$ represented in Equation (2) between the surfaces. There was, however, a notable difference in $E_2$, the modulus corresponding to the larger time constant τ₂. On the anterior $E_2$ was 0.35 ± 0.2 and 0.62 ± 0.10 (p = 0.06) on the posterior surface. A quantification of the mechanical asymmetry found between the anterior and posterior surfaces of the iris is shown in Table 2.

Fig. 6.

(a) Typical stress–relaxation experiment on a sample held at a constant displacement of 80 μm for 400 s for the anterior (A) and posterior (P) surfaces. (b) Cross-section of a histological image illustrating the different structures composing the posterior and anterior surfaces. (c) A typical fit of the first step in the relaxation hold for the anterior and posterior surfaces.

Table 2.

Effective material properties of the iris found during the indentation experiments.

Property	Anterior	Posterior	p-value
E0* (kPa)	4.0 ± 0.5	6.0 ± 0.6	p < 0.001
E∞* (kPa)	2.3 ± 0.3	4.4 ± 0.9	p = 0.007
E 1	0.74 ± 0.06	0.81 ± 0.08	p = 0.54
E 2	0.35 ± 0.02	0.62 ± 0.10	p = 0.06
E 3	0.32 ± 0.03	0.38 ± 0.04	p = 0.39
τ1* (s)	6.12 ± 0.62	10.32 ± 0.91	p = 0.05
τ2* (s)	121.31 ± 6.84	210.61 ± 9.41	p = 0.03

The reported values represent the mean ± 95% CI. A * indicates a significant difference between the surfaces defined by p < 0.05.

4. Discussion

It must first be noted that the indentation test is not measuring a true modulus but rather an effective modulus. The iris in both human and pig is heterogeneous, viscoelastic, anisotropic (Heys and Barocas, 1999), and possibly poroelastic (discussed below), so combining the entire indentation mechanical response into a linear modulus is a simplification. The effective moduli $E_0$ and $E_{\mathrm{\infty }}$ reported here must be seen as measures of the tissue response as a whole and not as specific moduli with formal meaning. The porcine iris is not identical to the human, notably in its larger sphincter, but the similarity in size and structure suggests that conclusions drawn from the porcine experiment translate to the human. Nevertheless, we stress that our conclusions must be interpreted recognizing that interspecies differences may be significant.

Those warnings notwithstanding, a significant difference was found between the anterior and posterior effective moduli. Little variance was found between right and left irides nor the locations of the indentations since all indentations were done in the mid-section of the tissue. Since the indentation experiment was most sensitive to material properties of the tissue near the indenter (Costa and Yin, 1999), it must be concluded that the posterior components of the iris – the dilator, the pigment epithelium, and perhaps the sphincter – are stiffer than the anterior border layer and stroma. Because of the interaction among the different components and the complex anatomy, interpretation of the data beyond the qualitative conclusion above is difficult. The results showed no difference between the anterior and posterior indentation during the early stages of indentation (<30 μm in Fig. 5a). One would expect that if the dilator, which is very near the posterior surface of the iris, is stiffer than the iris stroma, the early stages of indentation would show the most pronounced differences. A number of possible reasons exist for our unexpected result. First, at the early stages of indentation, the tissue beneath the indenter is primarily compressed, but as the indentation continues, the tissue surrounding the indenter is placed in tension. If the dilator is significantly stiffer in extension than in compression, as one would expect for smooth muscle (Walraevens et al., 2008), the stiffer response would not be seen initially. Another explanation would be that the highly compressible stroma deforms easily enough under the dilator that the stiffness of the dilator is masked in the early stages of the experiment. Finally, it is possible that at very short times, effective incompressibility of the biphasic stromal tissue (due to lack of time for water to escape) creates the appearance of greater stiffness as in the well-studied case of articular cartilage (Mow et al., 1980; Mak et al., 1987).

It is impossible to tell from our experiment whether the time-dependent response is due to viscoelasticity (i.e., relief of stress via Brownian motion of the entangled constituent molecules) or poroelasticity (i.e., relief of stress via exudation of pressurized interstitial fluid, resisted by the drag on the fluid as it permeates out of the tissue). The latter effect is certainly possible, especially in the highly hydrated iris stroma, and could be important in light of the recent report (Quigley et al., 2009) that iris cross-sectional area changes during dilation. Quigley et al. reported that the change in cross-section was different for narrow-angle versus normal patients. If the cross-sectional area change differences correspond to volume change differences, then the permeability of the stroma could be important. Since the non-indented surface of the sample was glued to an impenetrable plate, one would expect that if stromal permeability should be a significant factor, the relaxation time would be shorter when the tissue was indented on the anterior surface, providing a shorter path for pressurized interstitial fluid to escape. The results from the experimental analysis were consistent with this hypothesis, showing shorter relaxation times on the anterior surface.

We found previously (Whitcomb et al., 2009) that the modulus of the iris in uniaxial stretch was 3.0 kPa for circumferential stretch and 4.0–5.0 kPa for a radial stretch. Since the stretch experiments were performed in the plane of the iris, any differences between layers of the iris would have been averaged out. The average values for $E_{\mathrm{\infty }}$ were 2.0 kPa (anterior indentation) and 4.0 kPa (posterior indentation) are comparable to those results. In comparing stretch to indentation data, however, it is important to recognize that fibrous tissues are usually much stiffer in tension than in compression. The indentation test both stretches the tissue along the indented surface and compresses the tissue beneath the indenter, so a purely elastic analysis such as was done here could underestimate the stretch response and overestimate the compression response.

Although these results do not permit quantitative conclusions regarding the properties of specific components of the iris, they do raise important questions about the role of heterogeneous iris mechanical properties in certain conditions. For example, we have recently (Whitcomb et al., 2010) put forward the idea that the posterior positioning of the dilator muscle within the iris contributes to the anterior bowing observed during dilation, and we have shown theoretically that when the location of the contractile mechanism within the iris is changed, the iris contour changes. It is important to consider whether the heterogeneity of the iris affects the contour in narrow-angle glaucoma (Epstein et al., 1997) or pigment dispersion syndrome (Farrar and Shields, 1993) patients.

One must also consider the role of tissue heterogeneity in intraoperative floppy iris syndrome (IFIS, Chang and Campbell, 2005). The bending of a single uniform structure is very different from the bending of a structure with multiple layers of different stiffness, and it may be that the perceived floppiness is due not to a weakening of the entire iris but to weakening or thinning of the dilator muscle specifically. The work of Prata et al. (2009) suggests that there is an overall thinning of the iris in the region of the dilator but does not identify whether the dilator muscle is thinning, the stroma is thinning, or it is some combination of the two. However, more recently Santaella et al. (2010) found on average the dilator muscle, not the stromal thickness, was 23.2% thinner in Tamsulosin treated patients versus the control, suggesting that the dilator could a significant factor in the iris’s flexibility.

This study also suggests two important avenues for future work. The current study showed a significant difference between anterior and posterior indentation, but with only force–displacement data, it is difficult to assign properties to specific components of the iris. It would be possible to construct a finite element model based on the histology and extract properties for the sphincter, dilator, and stroma that provide force–displacement curves similar to those obtained experimentally. With only the force–displacement data, however, interpretation is somewhat difficult. A better option, although experimentally more challenging, would be to use an image correlation method, e.g. Al-Qaisi and Akkin (2008), to extract the entire displacement field, which could then be used in conjunction with an inverse finite element model, e.g. Raghupathy and Barocas (2010), to determine the properties of different regions – sphincter, dilator, pigment epithelium, and stroma. Such an experiment would be difficult to perform and analyze, but we hope that continuing advances in both imaging and analysis will make it possible and will lead to greater insight into the mechanical behavior of the iris in healthy and diseased eyes.