It is well known that glaucoma is one of the primary leading causes of blindness [1
]. Without proper treatment, glaucoma can degenerate visual ability, and eventually result in an irreversible vision loss [2
]. Although glaucoma is a multifactorial disease, increased intraocular pressure (IOP) is one of the major risk factors for its progression. Therefore, in many cases glaucoma can be treated by reducing the IOP [3
]. The IOP is the fluid pressure of the eye, that is, the magnitude of the force generated by the aqueous humor on the internal surface area of the anterior eye [4
]. Increasing the IOP can cause mechanical stress and ischemic effects on the retinal nerve fiber layer, thus, accurate IOP measurements are needed for efficient diagnosis to perceive the risk of developing glaucoma in advance as well as for assessment of progress after treatment.
Although there are several methods for measuring the IOP, Goldmann applanation tonometry (GAT) is known well as a classical gold standard method. The GAT determines the IOP from the force required to flatten a pre-defined area of the central cornea [9
]. This assessment, however, has some limitations, requiring patients to maintain open the upper eyelid during measurement, necessity for anesthesia and fluorescein staining, direct contact with the cornea, immobility due to the slit-lamp set up, and the influence of the central corneal thickness (CCT) measurable with an additional instrument such as an ultrasound pachymeter [11
To solve these problems, new methods for IOP measurement are being developed, especially, a rebound tonometer which uses an induction/impact principle [15
]. It is based on launching a magnetized probe with a plastic end-tip towards the cornea and monitoring the voltage induced in a solenoid coil as it returns [11
]. This instrument is portable and easy to use, and does not require topical anesthesia [11
]. However, it causes patients discomfort and may inflict cornea injury since it is accomplished while they open their eyes, and the probe hits cornea directly. In addition, the CCT has influence on the measurement of the rebound tonometer as well as GAT, resulting in need for any assessment to always be accompanied by a pachymeter measurement [19
Unlike contact tonometers, a non-contact tonometer (NCT) is free from mechanical contact with the eye, providing a non-invasive test and minimal risk of infection [13
]. The NCT, introduced by Grolmann in 1972 [25
], generates a puff of air, whose force increases linearly over time and collimates a beam of light. The cornea is flattened by the air puff and reflects a light beam onto a sensor. The time required to produce the peak reflection is measured and converted to the IOP [13
]. Although there are several advantages such that the NCT does not require the use of anesthesia and can be operated by paramedical personnel because of its simple operating process, the air puff still startles patients, and the measured results are more affected by the CCT compared to the GAT. In addition, it has decreasing reliability at higher pressure ranges [22
In the same vein, to measure the IOP using a non-contact method, this paper presents a novel mechanism to assess the IOP using acoustic radiation force. Acoustic radiation force has been widely used in elasticity imaging, visually providing elastic information of a tissue which is hard to identify using a conventional brightness mode (B-mode) image. This radiation force, produced by a burst of ultrasound waves, generates a localized tissue displacement that can be detected by using an ultrasound correlation-based method. The tissue displacement response is dependent on the magnitude of the applied force and inversely proportional to the tissue stiffness [28
]. That is, when a certain force is applied, measuring the tissue displacement can provide information about its elasticity. If the ultrasound transmission intensity is within a safe range, it does not have any harmful effects on the human body, does not require anesthesia or direct contact with the cornea, and is not limited by the subject’s position or angle, thereby minimizing patient discomfort. A single transducer can generate an acoustic radiation force to produce a tissue response and track the resulting displacement at the same time.
There is another imaging technique that measures tissue displacement known as the optical coherence elastography (OCE) technique. In this scheme, a laser pulse, needle probe, or acoustic radiation force are used to produce an external stimulation to generate tissue displacement, and the resulting changes are measured by the optical coherence technique (OCT) [36
]. However, OCE eventually requires two different energy sources for tissue displacement generation and measurement. Thus, a complicated probe and a system in which the external simulation source is integrated with the OCT are required [36
]. In addition, to the best of our knowledge, the OCE has been studied for elastic imaging that presents the mechanical properties of corneal tissue, but it has not been applied to tonometry.
The purpose of this research is to demonstrate the correlation between the IOP and displacement generated by an acoustic radiation force, and verify the feasibility through experiments. The analysis of the correlation mechanism was performed by calculating a formula that converts the displacement into an IOP value. For displacement assessment, gelatin-based phantoms having different elasticity were fabricated by controlling the gelatin concentration. A pushing beam for generation of acoustic radiation force and tracking beams for detection of displacement were produced by a specially designed ultrasound beam sequence using a customized transducer having a 5 MHz center frequency. The radio frequency data was obtained and processed off-line to calculate the axial displacement along the beam axis. To demonstrate the performance of the proposed method, we compared the resulting displacement with the IOP obtained by a commercial rebound tonometry apparatus.
4. Discussion and Conclusions
In this study, we demonstrated that an acoustic radiation force capable of generating tissue displacement can be used for IOP measurements. According to the commonly used formulas to calculate the true IOP, it can be seen that the final corneal displacement is derived by calculating the difference between the deformation due to the external force and the deformation due to the internal force of the eye. In an applanation tonometer such as Goldmann tonometer, the final degree of deformation is predetermined to a specific value, and then the IOP is calculated using the force required to deform it to that value. From the same concept, it can be inferred that the final degree of deformation reflects the IOP when a constant force is applied. In this experiment, we tried to verify whether the acoustic radiation force can induce deformation by replacing the applanation force (=external force), and it was confirmed that the displacement value changes according to the IOP.
For this experimental demonstration, we fabricated a 5 MHz ultrasound transducer, and produced eye phantoms with various gelatin concentrations to observe changes in intraocular pressure step by step. The data obtained using the sequence composed of pushing and tracking beams was converted into the axial displacement information over time for each phantom through offline processing. If the stiffness control of fabricated phantoms was conducted properly, the phantoms should show the incrementally changed IOP values as the gelatin concentration increases. As a result of checking the IOPs of the fabricated phantoms with a commercial tonometer as shown in Figure 7
a, various IOP values were obtained as we intended, and thus it can be said that the customized phantoms are suitable for use in IOP measurement experiments. As depicted in Figure 7
b, it was confirmed that axial displacement of the phantom surface was caused by the acoustic radiation force of the applied pushing beam. Also, after the pushing beam application was completed, the peak deformation was found in the first lobe, and its value gradually decreased over time.
The applied pushing beam is a 500 cycle 5 MHz sine wave, which is ten times larger in amplitude and has a relatively longer time duration compared to the tracking beam. Therefore, the echo caused by the pushing beam was not negligible, and this echo signal makes it difficult for the tracking beam to be transmitted immediately after the pushing beam is terminated. Since both the pushing and the tracking beams of the sequence were composed of 5 MHz sine waves, there was a limit to removing echo signals from the pushing beam by the frequency filtering scheme. Thus, the first tracking beam was transmitted 0.3 ms later after the pushing beam was sent, in this case, the influence of the echo signal of the pushing beam can be minimized. However, if the displacement information can be obtained immediately after the pushing beam, the accurate value closer to the actual displacement can be obtained.
As a result of measuring the IOP values of the six types of gelatin phantoms using the rebound tonometer, it was confirmed that the IOP values were different depending on the concentration. Thus, these gelatin phantoms can be used in experiments to measure IOP values using the acoustic radiation force. The final displacement induced by the acoustic radiation force was dependent on the internal pressure of the object as expected from the formulas, and the R-squared value of the exponential regression curve for those two values was 0.9315, which shows a relatively large goodness to fit.
In this study, it was not possible to directly convert the eye displacement measured in the experiment into IOP due to some limitations. For one thing, the fabricated phantom did not have a thin layer like the actual corneal layer, and we were unable to measure the Young’s modulus and acoustic radiation pressure in our laboratory environment. If the above parameters are ready, it will be possible to get the IOP right away and to compare other conventional procedures. Although our results sufficiently show that the displacement induced by the acoustic radiation force is related to the IOP, further studies including ex-vivo and in-vivo tests will be undertaken to increase the reliability of the proposed technique and investigate the effects of other anatomical confounders.
The main purpose of this study is to validate the method of measuring IOP using displacements generated and tracked by the acoustic radiation force. Therefore, in order to deliver sufficient ultrasound energy, the experiments were performed using a gel between the eye phantom and the ultrasound transducer. However, the use of a gel on a human cornea can cause great discomfort to the patient. Thus, when this technique is applied to a human, the ultrasound transducer should be placed on the eyelid using a gel, and the influence of the eyelid should be removed for precise displacement measurement. In general, the ultrasound acoustic radiation force has been widely used for measuring the elasticity of the organs where the targets are covered by the skin such as breast, thyroid or abdomen [28
]. Thus, the acoustic radiation force can pass through the eyelid and be transmitted to the cornea surface of the human eyes. In addition, since CCT generally affects the measurement results of Goldmann and rebound tonometers, a corneal thickness measurement using an ultrasound based pachymeter should be performed before and after IOP measurement to improve accuracy. However, the proposed acoustic radiation force method has the advantage of being able to measure the corneal thickness at the same time as the IOP measurement.
The reason for using the gelatin phantoms in this study is to ensure that the axial displacement, which can be simultaneously generated and measured by the acoustic radiation force, has correlation with the IOP values or not. In order to achieve this goal, experimental targets with gradually changing IOP values were required, the gelatin phantom capable of changing property was the proper target. Moreover, the gelatin-based phantom is a commonly used target for validation experiments for measuring elasticity or mimicking the structure of corneal tissues [52
]. In this regard, we conducted experiments with the gelatin phantoms for the feasibility validation of our proposed technique. We will try to find a way to gradually change the IOP value of the in-vitro target as the further work. In this experiment, the increase amount in the concentration of the gelatin phantom was determined by considering the range of human IOP. If the error range of the rebound tonometer used in the experiment and the concentration uniformity error of the hand-made gelatin phantom in the laboratory can be reduced, the increase amount in concentration of the gelatin phantom can be more reduced resulting in the increased IOP resolution.
As we mentioned the above, OCE techniques also apply acoustic radiation force to transfer it to the interior of corneal tissues, such as crystalline lens and retinas for elastic imaging [36
]. Our proposed technique also fires acoustic radiation force to the eyes, but it could have lower force than OCE since it does not need to transfer the force to the inside of eyes and only delivers the force on the surface. In addition, though the acoustic force is also used in the tracking beams for measuring the resulting deformation, it is supposed to cause no problems since it has a short duration and an amplitude equal to one tenth of the pushing beams. Even so, a precise safety test using ex-vivo and in-vivo should be conducted in the further study.
In this study, the feasibility of the acoustic radiation force-based IOP measurement was verified experimentally. Using the dedicated transducer, customized beam sequence, and gelatin phantoms, we have confirmed that the acoustic radiation force can cause axial deformations on the phantom surface, and this amount of displacement is also related to the IOP. In other words, as the IOP increases, the resistance to the deformation caused by the acoustic radiation force increases and thus the amount of ocular displacement decreases. Conversely, when the IOP decreases, the amount of ocular displacement by the acoustic radiation force increases. Although this study shows that the displacement induced by the acoustic radiation force is related to the IOP, in order to increase the reliability of the proposed method, additional experimental studies such as in-vitro, in-vivo, and sensitivity tests are needed. In addition, since safety issues are very important in the proposed technique, safety tests focused on their effects on corneal endothelial or optic nerve cells are indispensable.