SYSTEM IDENTIFICATION APPROACH to OCULAR TACTILE TONOMETRY

Frequent intraocular pressure (IOP) measurements are necessary for the diagnosis and management of glaucoma. Most current tonometers utilize some form of corneal deformation in order to estimate the IOP. Trans-corneal tonometers require sterile parts, or are complex and expensive (air-puff tonometer). Therefore, their use at home is not practical. Trans-scleral and trans-palpebral tonometry on the other hand, could be utilized at home but is less sensitive and prone to measurement errors. One such method is based on tactile palpation of the eye (palpation tonometry). During tactile palpation, the IOP increases significantly. The goal of tactile tonometry is to estimate the initial (starting) IOP, based on contact forces and displacements obtained during the palpation. This paper describes the development of two phenomenological models relating the contact forces, displacement, and intraocular pressure during instrumented mechanical palpation of porcine eyes. The first model predicts the changes of intraocular pressure (IOP) and corresponding forces as a function of the applied displacements and initial IOP. The model was build in two parts. First, a pressure vs force and displacement was developed and subsequently a force vs displacement and pressure was added. It is then used to derive a second model that predicts the initial IOP from the applied forces and displacements. Leastsquares techniques have been used to extract parameter values of the two models. The first model (14 parameters) produced a maximum error of 1.65 mmHg, while the second model (17 parameters) produced a maximum error of 0.85 mmHg over the test range of 10 to 35 mmHg. The second model leads to an explicit solution for the initial IOP as a function of the applied forces and displacements. When calibrated, this model can be used to interpret palpation data and estimate the starting IOP. Most commercial tonometers are certified to be accurate within 2-3 mmHg against Goldmann applanation tonometry. They also suffer from larger errors for low (IOP 11 mmHg) and high (IOP 21 mmHg). Both models presented here demonstrate sufficient accuracy for the development of a tactile tonometer. Through simulations, it has been shown that accurate prediction of IOP from tactile data requires data from multiple displacements, i.e. an active palpation system is needed in order to realize a tactile tonometer. The ability to predict changes of IOP during palpation tonometery can also be useful in the development of mechanical eye stress tests, replacing the drug-induced Priscol provocative test.