IMPROVEMENT OF ACCURACY OF IMAGELESS NAVIGATION USING GENERALIZED ESTIMATION MODEL OF SOFT TISSUE THICKNESS IN THR

Abstract

Typical navigation system to insert hip implants in the accurate position consists of a 3D position measurement device and a computer. These navigation systems are classified into two categories according to the method of identifying the anterior pelvic plane that works as the reference of the orientation of the acetabulum cup. The preparation process for imageless navigation system is very easy because it uses three anatomical bony markers to define the anterior pelvic plane. When these anatomical bony markers are hard to locate, especially at the pubic symphysis due to the thick soft tissue, the accurate direction of the cup cannot be secured. The aim of this study is to estimate the soft tissue thickness without using the patient’s specific data such as the A-mode ultrasound image or C-arm image.

In our previous study, it was pointed out that the thickness of the hypodermic fat obtained through an ultrasound image could be estimated using the patient’s BMI and the displacement created by a specific force. Considering the probe shape, the soft-tissue thickness estimation formula is expressed as follows:

k: constant for the shape of the probe end

Only two kinds of the probe end shapes (flat-ended probe and spherical-ended probe) were considered, and the change in the k value corresponding to the radius was calculated using the FE model of the soft tissue for each subject. The finite-element model was constructed as axisymmetric.

The simulation result of the initially assumed variables and the measured result were compared, and the optimization method was used to minimize the error: The RMS difference between the result of the experiment and that of the analysis was taken as the objective function. With the FE analysis for the two kinds of probe shapes with one subject, we determined the shape variable (k).

From the formula composed by a model with data from 28 people, the average error was 3 mm equivalent to the angle error of less than 1°. Therefore, the use of the method suggested in this study will help to improve the acetabulum cup navigation in THA, when we use only the surface points on the soft tissue. In addition, it seems that the soft-tissue thickness estimation formula suggested in this study may be generally used.