Analysis Wave Speed of Optic Nerve and Optic Nerve Head for Assessing Normal Tension Glaucoma by using Vibro-Elastography Technique

Abstract

Background:

This research aims to evaluate wave speed and viscoelasticity of ocular tissues including the optic nerve and optic nerve head of human eyes between normal tension glaucoma patients and healthy controls by using vibro-elastography techniques.

Methods:

Participants included 12 patients and 12 controls. Wave speed was measured at the optic nerve and optic nerve head in each subject and viscoelasticity was estimated by using Voigt model. Wave speed and viscoelasticity of the optic nerve and optic nerve head were compared between patients and controls by linear mixed models via a restricted maximum likelihood method. The correlation between intraocular pressure and wave speed, elasticity, and viscosity of patients was performed using the Pearson correlation coefficient.

Findings:

Significant differences in wave speed (p < 0.0005), elasticity (p = 0.0001) and viscosity p < 0.0001) between patients and controls at the optic nerve head. There was a moderate negative correlation (r = −0.50, p < 0.05) between wave speed and elasticity and intraocular pressure at the optic nerve of patients but no correlation (p > 0.05) between wave speed, elasticity, and viscosity and intraocular pressure at the optic nerve head of patients. No significant difference and correlation among wave speed, elasticity, and viscosity vs. intraocular pressure of the control group at the optic nerve and optic nerve head.

Interpretation:

Ultrasound vibro-elastography is useful for noninvasive measurement of viscoelasticity of ocular structures. The glaucoma patient is associated with biomechanical property changes in the optic nerve and optic nerve head, suggesting another way to assess glaucoma focusing on the optic nerve and optic nerve head.

1. Introduction

Glaucoma is the second leading cause of blindness in adults aged 50 and older and the fourth leading cause of moderate to severe vision impairment. It is the most common cause of irreversible blindness and the second most frequent cause of irreversible moderate to severe vision impairment in adults aged 50 years and over (Adelson et al., 2020). Glaucoma is a complex eye disease characterized by elevated intraocular pressure (IOP), which leads to progressive vision loss if untreated (Jamie et al., 2024). However, around 50% of glaucoma cases exhibit normal IOP values at diagnosis, and many cases continue to experience progressive vision loss despite IOP-lowering therapy (AGIS Investigator et al., 2000). Conversely, the majority of normal tension glaucoma (NTG) patients with elevated IOP do not develop glaucoma, indicating that irregular ocular biomechanical properties may play a role in the pathogenesis of glaucoma (Trivli et al., 2019). To gain a better understanding of the pathogenesis of glaucoma, and identify potential targets for novel treatment strategies, it is crucial to comprehend the differences in biomechanical properties between individuals with glaucoma and healthy individuals.

Multiple previous studies of glaucoma focus on analyzing the size, mechanical properties, and stiffness of the cornea (Elsheikh et al., 2007; Han et al., 2015). These studies included testing to measure stiffness of the cornea by using direct and indirect methods such as tensile strength, stress–strain relationship, stress-relaxation properties, and elastography (Aoki S et al., 2023). However, these studies reported that the structure of the cornea is complex and affects wave dispersion when applying the stiffness measurement method by using shear wave elastography (SWE). In diabetic patients, glucose metabolism and electron transport overloaded lead to mitochondrial transport overload and the formation of large amounts of reactive oxygen species (ROS). ROS stimulates increased phosphorylation of p38 mitogen-activated protein kinase (p38 MAPK), activating the pro-apoptotic p38 MAPK pathway, thereby inducing apoptosis of neurons and regeneration of damaged neurons (Shao et al., 2020; Zhang S. et al., 2020). These pathological changes aggravate stiffness of the optic nerve (ON), so assessing ON stiffness is also a method to assess the pathological condition in diabetic patients (Yue et al., 2024). In addition, studies on different types of diseases, such as pseudotumour cerebri, pre-eclampsia, behçet’s disease, show that these diseases also affect the stiffness of the optic nerve (ON) and optic nerve head (ONH) (Asal et al., 2020, et al., 2019; Elsaid et al., 2022; Inal et al., 2017; Ma et al., 2020; Yilmaz et al., 2022; Yue et al., 2024). With the advantage of the ultrasound vibro-elastography (UVE) technique, we are capable of measuring wave speed (WS) in the ON and ONH of the human eye. Therefore, we applied this technique to analyze the mechanical properties in the ON and ONH of glaucoma patients.

Ocular tissue is often vulnerable to physical damage, so eye research requires noninvasive and safe methods to avoid damage to tissue from mechanical and thermal factors (Guazzaroni et al., 2020). The vibro-acoustic technique is an innovative technology that measures biomechanical properties of soft tissue (Bui et al., 2024; Zhou et al., 2020). A lot of studies have attempted to use acoustic wave vibration, a relatively noninvasive stimulation, to infer the biomechanical properties of ocular structures (Asal et al., 2019; Guazzaroni et al., 2020; Li R et al., 2023; Qian et al., 2021). Recent advances in noninvasive and clinically applicable methods to quantitatively assess the mechanical behavior of ocular structures, such as clinical shear wave (SW) elastography, could significantly help to this end (Asal et al., 2020, et al., 2019; Ramier et al., 2020).

SWE imaging is an innovative medical technology that utilizes ultrasound imaging to evaluate the mechanical properties of tissues through the measurement of SW (Guazzaroni et al., 2020; Zhang J et al., 2022). The technique involves the application of a transient ultrasound pulse to the tissue, inducing a small deformation, and subsequently tracking the pulse to determine the SW velocity (Bui et al., 2024; Li R et al., 2023). The elasticity of the tissue can then be derived from the measured SW velocity and has been demonstrated to be positively correlated with its stiffness (Asal et al., 2019).

However, most patient studies used the UVE with low excitation frequencies from 10 to 500Hz. Shao et al (Shao Y et al., 2022) studied breast cancer with an excitation frequency from 21 to 126Hz. Zeng et al (Zeng Q et al., 2021) studied the liver with external frequencies from 40 to 65Hz. Sara Mahdavi et al (Mahdavi SS et al., 2011) studied the prostate gland using excitation frequencies from 2 to 10Hz. Alternately, a study on phantoms (Hashemi H.S. et al., 2023) and one on ex vivo models (Li R et al., 2023) used the higher frequencies up to several kHz.

Here, we report on UVE measurements of WS for analyzing tissue viscoelasticity of both the ON and ONH in NTG and normal human eyes. Based on results from our previous studies with ex vivo (Zhang X et al., 2016, et al., 2011, et al., 2019) and in vivo (Zhang X et al., 2020) models, we selected three different frequencies, 100Hz, 150Hz, and 200Hz, for measuring wave propagation inside the tissue. At higher frequencies the penetration ability is rapidly reduced, and it is difficult to observe wave propagation in the optic nerve. This measurement was performed using an ultrasound system with a shaker to safely generate low-energy elastic wave excitation vibrations on the human eye at three selected frequencies (i.e., 100Hz, 150Hz, and 200Hz). With each excitation frequency, the measurements were performed three continuous times by a trained physician.

2. Methods

2.1. Participants

In accordance with an institutional review board-approved protocol (Mayo Clinic, Rochester, MN, USA), this study evaluated both eyes of 12 NTG cases (from December 2021 to May 2024) and 12 controls. The NTGs consisted of 6 males (50%) and 6 females (50%) and 12 controls consisting of 2 males (16.7%) and 10 females (83.3%). We enrolled mild to moderate NTG patients and healthy individual as controls. Some of the NTG patients were already on medication and washed out their eye drops after clinical examination but prior to the research study. However, the majority were newly diagnosed as NTG and treated with either eye drops or selective laser trabeculoplasty after enrolling into the study. All participants agreed to provide written consent and knew the purpose and methodology of this study. Details of NTGs and controls are shown in Table 1.

Table 1.

Details of all subjects

NTG	Age (years)	Control	Age (years)
1	76.0	1	47.5
2	75.0	2	43.6
3	70.5	3	42.9
4	71.1	4	47.9
5	70.1	5	61.9
6	54.3	6	66.1
7	59.0	7	62.5
8	74.0	8	72.0
9	67.0	9	67.8
10	64.0	10	67.3
11	72.0	11	72.1
12	74.0	12	63.6

A total of 24 subjects were recruited (consisting of 12 patients with NTG and 12 individuals as controls). In terms of age, significant differences (p = 0.0175) were found between the NTG and control groups in terms of mean deviation values (NTG group, 68.9 ± 6.43years vs. control group, 59.6 ± 10.51years).

2.2. Intraocular pressure measurement, ultrasound examinations, and two-dimensional speed map processing

The IOP was measured in both eyes of subjects by using a pneumatonometer (Model 30 Classic, Reichert Inc., Buffalo, NY) (Fig. 1a). Subjects were then placed in the supine position and IOP was remeasured after 5minutes. Calibration of the tonometer was verified according to the manufacturer’s instruction and the tip was cleaned before each set of measurements. Topical proparacaine 0.5% was instilled before each IOP measurement. The right eye was always measured first and around 15–20 minutes before the ultrasound examination (Kazemi et al., 2021).

Fig. 1.

(a) IOP measurement device, (b) Schematic of ultrasound experimental system, (c) an example of two analysis locations ON and ONH presents on B-mode and 2D speed map image.

Fig. 1b (Bui et al., 2024) describes the experimental model, which includes a 0.1-second harmonic vibration signal generated using a function generator (Model 33120A, Agilent, Santa Clara, CA). The excitation signal is amplified by an audio amplifier (Model D150A, Crown Audio Inc., Elkhart, IN) and then applied to a handheld shaker (Model: FG-142, Labworks Inc., Costa Mesa, CA 92626) on the eyelid using a small indenter with a 3mm diameter (Kazemi et al., 2021). The indenter of the shaker was solely utilized for generating wave propagation with a fixed 0.1second duration of excitation for all frequencies (Bui et al., 2023; Kazemi et al., 2021; Sit et al., 2017). An ultrasound system (Vantage 1, Verasonics, Kirkland, WA) with an L11–5v ultrasound probe (ultrasound transducer with a central frequency of 6.5MHz and 128 linear array elements) is used for detecting WS propagation at the ON and ONH of eye. The ultrasound probe contacts with the eyelid after the application of coupling gel (Aquasonic, Inc., Clinton Township, MI, USA) on the probe surface (Bui et al., 2024, et al., 2023; Kazemi et al., 2021).

The subjects were positioned in a supine position with both eyes closed and directed in an upward gaze. Measurements were made three times at each excitation frequency in both eyes. It took about twenty minutes for a subject. No pain or discomfort was reported from participants while using this device in the preliminary studies (Bui et al., 2023; Kazemi et al., 2021; Sit et al., 2017).

The reconstruction B-mode and 2D speed map with data obtained from ultrasound system were shown in some our previous studies (Bui et al., 2024, et al., 2023; Zhang et al., 2020a, et al., 2020b) and supplement 1. The 2D speed map presents the WS of the whole eye and area where redder color means higher WS, while bluer color means lower WS. Fig. 1c shows an example of a B-mode image, 2D speed map of two analysis locations (ON and ONH). The 2D speed map describes that WS at the ONH area is higher than WS at the ON.

2.3. Measurement of WS and estimated elasticity, viscosity at the optic nerve, optic nerve head of human eye

First step, we checked wave propagation at the ON and ONH to ensure high accuracy of calculating WS. Details of this step are shown in supplement 3. Second step, we chose the sizes and shapes of regions of interest (ROI) for calculating WS. Other studies show that there is no consensus on the optimal type, size or plane of the ROI for the SWE measurements (Asal et al., 2019; Lahtinen et al., 2022). Typically, a small 1–5mm circular ROI is placed over the stiffest part of the analysis area on the axial plane (Lahtinen et al., 2022). Thus far, no recommendations exist for ROI types in eye assessment with SWE. The sequence of analysis steps follows: First, we selected 5 ROIs (circle with diameter of 1mm) on B-mode image, then these ROIs was automatically mapped to a 2D speed map (B-mode and 2D speed map are same size). Second, WS was calculated based on mean and standard deviation (SD) intensity of all pixels inside ROIs of the 2D speed map. The WS was chosen with the standard SD/ mean ≤ 20% (Bui et al., 2024, et al., 2023).

In this study, we used the Voigt model (i.e., eq 1) with WS at multiple frequencies for calculating elasticity μ₁, and viscosity μ₂ (Bui et al., 2023; Zhang et al., 2020b; Zhou et al., 2018). Details were shown in supplements 2 and 3.

$$c=\sqrt{\frac{2\left({\mu }_1^2+{\omega }^2{\mu }_2^2\right)}{\rho \left({\mu }_1+\sqrt{{\mu }_1^2+{\omega }^2{\mu }_2^2}\right)}}$$ (1)

where c is local WS, ω is the angular frequency, μ₁ is shear elasticity, μ₂ is viscosity, and p is the mass density of the tissue in 1000kg/ m³. According to our experience with tissue biochemical application, WS is typically measured at three frequencies (100Hz, 150Hz, and 200Hz).

2.4. Statistical analysis

Mean values of WS, elasticity, viscosity, and IOP were compared between NTGs and controls using a linear mixed models via restricted maximum likelihood (REML) method for both the ON and ONH with the associated p-value adjusted using the Holm method per analysis. We also estimated the Pearson correlation coefficient between WS, elasticity, and viscosity vs. IOP separately for both NTGs and controls. Two quantitative variables are correlated if the p-value is < 0.05 (Miot et al., 2018). Statistical analysis was performed using R programing language (Bui et al., 2023).

2.5. Repeatability Analysis

Repeatability of WS measurement and reliability were evaluated with intra-class correlation coefficients (ICCs) of the first and second measurement of WS at the same location of the five ROIs. Repeatability was considered poor for an ICC less than 0.50, moderate for 0.50 ≤ ICC ≤ 0.59, good for 0.60 ≤ ICC ≤ 0.74, and excellent for 0.75 ≤ ICC ≤ 1.00 (Bui et al., 2023; Kazemi et al., 2021).

3. Results

There were no significant differences (p = 0.4898) found across either group in terms of IOP (NTG group, 17.8 ± 3.55mmHg vs. control group, 16.85 ± 2.82mmHg). The ICCs were evaluated in all 48 eyes at three different frequencies while calculating WS with MATLAB v.2023 software (Bui et al., 2024, et al., 2023). ICCs were calculated with a value always higher than 0.85 of first and second analysis.

3.1. Wave speed, elasticity, and viscosity analysis between NTG and control

The WS, elasticity, and viscosity were calculated for all 24 subjects that were used for statistical analysis. Fig. 2 shows the comparison results of WS between NTGs vs. controls of both the ON and ONH at 100Hz, 150Hz and 200Hz. There were significant differences of WS in NTGs vs. controls at the ONH (p < 0.005) and no significant difference of WS between NTG vs. control groups at the ON of all three frequencies. Details of WS values are shown in Table 1 of supplement 4.

Fig. 2.

Comparison WS of optic nerve and optic nerve head at three frequencies between NTG and control.

Fig. 3 shows the estimation and comparison results of elasticity (Fig. 3a) and viscosity (Fig. 3b) using the Voigt model on the ON and ONH for both NTG and control groups. There were significant differences of elasticity (NTG: 10.11 ± 4.37kPa vs. control: 5.39 ± 1.89kPa, p < 0.0001) and viscosity (5.94 ± 2.19Pa.s vs. 3.39 ± 1.61Pa.s, p < 0.0001) between NTGs and controls at the ONH. No significant difference between elasticity and viscosity between NTG vs. control groups at the ON. Details of elasticity and viscosity values for these analyses are shown in Table 2 of supplement 4.

Fig. 3.

Comparison Elasticity, Viscosity of optic nerve and optic nerve head between NTG and Control.

3.2. Correlation between IOP and WS, Elasticity, and Viscosity

We found that there was a moderate negative correlation (r < −0.5) between WS and elasticity vs. IOP at the ON of the NTG group. However, we found no significant correlation among IOP with other parameters at the ONH in the NTG group and both the ON and ONH in control groups. In addition, we observed the R² values in Table 2 and the distribution of WS values (@200Hz) in Fig. 4. The R² values are relatively small (< 0.35) which suggests WS and elasticity at the ON in the regression model are not effectively explaining the variation in IOP. Details of results are described in Table 2 and Fig. 4.

Table 2.

Correlation between IOP and WS, elasticity, viscosity of optic nerve and optic nerve head as well as with age for both NTG vs. control.

Location	Parameters	NTG			Control
		r	p-value	Linear regression	r	p-value
Optic Nerve	WS-100Hz	−0.52	0.009	3.50–0.0845*IOP; R2 = 0.27	0.30	0.159
	WS-150Hz	−0.47	0.019	3.47 −0.0836*IOP; R2 = 0.23	0.38	0.063
	WS-200Hz	−0.50	0.012	3.61 −0.081*IOP; R2 = 0.25	0.24	0.268
	Elasticity	−0.59	0.002	13.2 −0.503*IOP; R2 = 0.35	0.30	0.412
	Viscosity	0.11	0.600	x	0.11	0.621
Optic Nerve Head	WS-100Hz	−0.38	0.068	x	0.26	0.213
	WS-150Hz	−0.30	0.147	x	0.32	0.125
	WS-200Hz	−0.36	0.087	x	0.38	0.068
	Elasticity	0.15	0.540	x	0.30	0.157
	Viscosity	−0.19	0.365	x	0.29	0.171

Fig. 4.

Correlation (Pearson method) between WS vs. IOP at optic nerve and optic nerve head of NTG: WS200_ON/WS200_ONH as wave speed of optic nerve/ optic nerve head at 200Hz.

4. Discussion

Typically, studies of glaucoma have primarily focused on analyzing changes in mechanical properties of the cornea that lack of tissue characteristics at different locations of the eye (Kazemi et al., 2021; Medeiros et al., 2013; Ramier et al., 2020; Tanter et al., 2009; Weng et al., 2021). However, our study takes a more comprehensive approach by examining two deeper locations of eye such as the ON and ONH utilizing the UVE technique. Then, we conducted similar tests on both NTG cases and healthy individuals for evaluating the biomechanical properties between these groups. The results show that there were significant differences of WS, elasticity, and viscosity at the ONH between NTG and normal eyes. This finding indicates glaucoma predominantly affects the ONH rather than the ON. Therefore, we hypothesize that this difference is attributable to the biomechanical properties of the ONH, which render the ONH more susceptible compared to the ON.

Our study has limitations that are mentioned. Firstly, the small number of subjects was a challenge in recruiting age-matched participants with both normal and NTG cases. Consequently, NTG cases included in our study were relatively advanced in age, ranging from 54.3 to 76.0years and normal controls ranging from 42.9 to 72.1years. The small sample size of this study along with the significant difference in the age of each group also affected the accuracy of comparison results. Also important note is that our sample size and the homogeneity among samples may not be large enough to accurately detect correlation. Specifically, we have a NTG patient with IOP (around 9.5mmHg) lower than the rest of NTG group that may affect correlation results in Fig. 4 and Table 2. Therefore, in the future, we need to increase the number and diversity of NTG patients. Based on the WS provided in Table 1 of the supplementary material and G*Power software (Faul et al., 2007), we estimated the number of subject of each group around 25. These subjects can provide better results of significant differences in WS between NTGs and controls at the ON. Secondly, our study only focused on the NTG group without patients with high tension glaucoma (HTG). Therefore, we do not have a comparison of different mechanical properties of the ON and ONH between NTG and HTG (Chuangsuwanich et al., 2023). Thirdly, we measured WS at the ON and ONH with three frequencies (i.e., 100Hz, 150Hz, and 200Hz) which are low frequencies to ensure good penetration when examining deep areas of the eye. The WS values are combined with the Voigt model and Levenberg-Marquardt non-linear least-squares algorithm to estimate the viscoelasticity of the ON and ONH, however the Voigt model may be too simplified for this problem. Therefore, the elasticity viscosity values are relatively low compared to the stress and train tests in other studies (Park et al., 2024). In the future, we will work with different viscoelastic models (i.e., Maxwell, standard linear solid, spring pot and fractional Voigt model) (Zhou et al., 2018) or develop complex models for estimating the viscoelasticity of the ON and ONH.

One potential concern regarding ultrasonic wave elastography (UWE) is the possibility of interference caused by measurements taken through the closed eyelid and upper tissues, which could impact wave velocity measurement. However, due to the direct visualization of wave propagation in different tissues and layers using an ultrasonic transducer, stacked tissues are unlikely to have a direct effect on wave propagation. Fig. 2(a, c, e) of supplement 3 show the wave propagation and wave displacement of the ONH in time domain by selecting the pixels at an area of the ONH in B-mode images. Nevertheless, our current implementation of UWE is limited by the measurement of wave propagation in a single direction, while eye tissues exhibit significant anisotropy (Guazzaroni et al., 2020; Weng et al., 2021). UVE offers advantages over conventional ultrasound elastography by using harmonic vibrations rather than ultrasound radiation force (URF) for wave propagation. UVE induces tissue movement up to 20μm, compared to 1–2μm with URF, which requires high-intensity ultrasound and poses risks of tissue damage making it unsuitable for ocular studies. UVE’s low-frequency mechanical waves penetrate deep tissue safely with the ultrasound probe serving only to detect shear wave propagation. This makes UVE effective for examining deep ocular tissues unlike optical coherence tomography (OCT), which is limited to corneal studies despite its higher resolution (Han et al., 2015). UVE achieves a spatial resolution of ~0.3mm with wave velocity measured around 5mm and a temporal resolution of 0.5ms at 2000 frames per second, ensuring safe and accurate assessments (Zhou et al., 2020).

In another study, (Yue et al., 2024) used the Mindray Resona 7 ultrasonic diagnostic apparatus with an L14–5 WU linear array probe, frequency of 5–14 MHz, to assess the stiffness of the ON in 30 healthy participants and 95 patients with three different types of diabetes. However, to our knowledge, the Mindray Resona 7 ultrasound system uses URF technology to generate excitation vibrations and measure the shear wave speed in ROI. This technique requires strict control of the URF intensity and the medical index (MI) and thermal index (TI) parameters, especially when applied to vulnerable tissues such as the eye (TI < 1.0, MI < 0.23) (Guazzaroni et al., 2020). Our technique uses external vibrations, so there is no need for strict control of URF. Moreover, the frequency and amplitude of oscillation are relatively low, thus there is no damage to the tissues, which is also the advantage of the vibro-elastography technique compared to other methods. On the other hand, the Verasonic-256 system can collect raw data as radio-frequency signals where the components in the signal can be analyzed in detail to restore speed maps on ROI based on different algorithms (Kijanka P et al., 2024; Z. Dong et al., 2024). This is also the main difference between the Verasonic system and other ultrasound systems used in clinical applications.

In addition, (Qian et al., 2021) used Verasonic and external vibration to evaluate the stiffness of the ON and posterior sclera in 8 ex vivo porcine eyes. Different from our system, Qian used a single element-transducer mounted to a 3-axis translation motorized linear stage for scanning 100 A-scans of porcine eye. To collect the 100 A-scan signals during the experiment, we estimated the collection time of their experiment to be 20 seconds (step size 42μm, scanning distance of 8.4mm and time per each measurement of 100ms) which is a relatively long time when implemented in patient’ studies. Normally, our system has a collection time less than 1second, which made it comfortable for the patient.

To mitigate potential sources of variation, all measurements and visual analyses were conducted by a skilled investigator who maintained consistent positioning of the shaker and super probe for each eye (Bui et al., 2023; Kazemi et al., 2021). Despite these precautions, it is conceivable that the differences in elasticity between the two groups may have been limited to specific layers or regions of the ON and ONH that were not captured by our measurements.

However, our study examined both the ON and ONH which gives a better overview of the effects of glaucoma on human eyes. This is different from many previous studies that only focused on one part of the eye such as the cornea (Kazemi et al., 2021), optic nerve(Yue et al., 2024), optic nerve head (Qian et al., 2021), or sclera(Ma et al., 2020). Moreover, the calculation of WS based on the 2D-speed map technique can calculate the WS at every point in the eye, which also makes a difference compared to our previous study when applying the WS calculation technique on the surface.

5. Conclusions

Our study shows that WS, elasticity, and viscosity of the ONH were significantly different between NTG and normal eyes but no significant difference of the ON between NTG and normal eyes. Another peculiarity found was that the stiffness of the ONH is highest in both NTG and normal eyes. Furthermore, we still found significant negative correlations between WS, elasticity and IOP for patients with NTG. In the future, the UVE technique can be applied to measure and compare cornea and scleral elasticity between glaucoma and papilledema with a higher number of cases.

• Measure wave speed and analyze viscoelasticity of optic nerve and optic nerve head.

• Comparison between patients and healthy controls.

• Correlation between intraocular pressure and wave speed.

• Quantitative method to assess glaucoma.