System for intraoperative evaluation of soft-tissue-generated forces during total hip arthroplasty by measurement of the pressure distribution in artificial joints

Abstract

A system for evaluating the soft-tissue-generated forces at the hip joint was developed. The system enabled measurement of contact pressure distribution at hip joint surfaces, as well as evaluation of the artificial hip joint condition during total hip arthroplasty (THA). First, a pressure sensor module that forms part of the artificial joint was constructed. Eight small pressure sensors were installed in the spherical head component of the ball-and-socket joint. Next, software for recording and visualizing the detected pressures at 1-millisecond intervals was developed. The pressure distribution was displayed in real time via 3D computer graphics on a monitor. The system enabled intuitive recognition of the direction of soft-tissue-generated forces and pressure distribution in three dimensions. Accuracy tests were conducted using a high-accuracy 6-degree-of-freedom positioning device and digital force gauge. The error between the applied loads and measured forces was 3.42 ± 3.26 N (mean ± standard deviation) for each coordinate in 10 trials involving load application from 10 different directions. Next, a clinical evaluation was conducted during THA. The relative positions of the cup and stem component were measured using a surgical navigation system simultaneously with the pressure measurement. The system allowed real-time acquisition of information regarding the artificial hip joint, as well as comparison of the differences in the hip condition when several types of neck were used. Further improvements to the calibration method should enable more accurate measurements. We believe this system will be a useful tool for selecting an appropriate implant that fits a patient's hip joint or for estimating the risk of complications following surgery.

• Total hip arthroplasty
• pressure distribution
• in vivo measurement
• real-time measurement

Introduction

In total hip arthroplasty (THA), selection of the optimum size and type of implant to fit the anatomical characteristics of the individual patient is important in order to avoid postoperative complications such as dislocation or loosening [1–3]. In recent studies [4–6], 3D skeletal models created from CT data for individual patients were used to select the proper implant with optimum fit, and such systems have been applied clinically. However, a cadaveric study by Bartz et al. [7] has indicated that the causes of postoperative dislocation are not restricted to impingement between hard tissues. Specifically, the study indicated that the primary mechanisms of dislocation following THA can be mainly classified into three types: impingement of the femoral neck on the cup liner; impingement of the femur on the pelvis; and spontaneous dislocation due to excessive external force versus the muscular force. Using geometric structural data on the hard tissues derived from CT data, the surgical planning system can estimate the proper implant needed to avoid dislocation arising via the first two types of mechanism. However, for the third type of mechanism, the causes of dislocation depend not only on the hard tissue alignment, including the anatomical structure of the bones or the position of the implant, but also on the soft-tissue balance around the hip joint. In other words, the forces induced from the surrounding soft tissues, such as muscles, tendons or ligaments, or the condition of the articular cartilage may contribute to the risk of dislocation. Therefore, we considered that the ability to measure such forces intraoperatively before fixing the real implant would provide additional important information for determining the best implant for the patient.

The purpose of the present study was to develop a system that can intraoperatively measure the hip joint contact forces and pressure distribution at the sliding surface and evaluate the balance of the forces induced by the surrounding soft tissues. By using this system in conjunction with the preoperative CT-based planning system, surgeons would be able to select the optimum size and type of implant for each patient, which should reduce the risk of postoperative complications.

Methods

Pressure sensor module

Eight subminiature pressure sensors incorporating foil strain gauges were used in our pressure measurement system. The diameter of the diaphragm of each sensor was 6 mm and its thickness was 2 mm. Since the measurement system was developed for use during surgery under anesthesia, gravity loading and acceleration during movement were not considered for these measurements. Hence, in order to quantify the pressure distribution with maximum sensitivity under relatively low pressure, we used a 3-MPa capacity sensor for each point. Furthermore, to measure the pressures over the entire spherical surface, 8 pressure sensors were distributed over the surface at almost equal intervals. A customized femoral head component was designed using a 3D CAD system in order to embed the 8 pressure sensors, including their wiring, under the spherical surface (Figure 1). Although the diameter of the head can easily be changed to fit individual patients, a head of diameter 32 mm was created for the initial trial reported here.

Figure 1. CAD data of the customized femoral head component. The head compartment was designed using a 3D CAD system to allow 8 pressure sensors, including their wiring, to be embedded under the spherical surface. [Color version available online.]

The physical model was manufactured from plastic based on CAD data using a laminate molding technique. The 8 sensors were embedded in the customized head and their pressure-sensing diaphragms were covered with spherical plastic parts to reduce friction against the cup component during movement (Figure 2). All the wiring was connected through a single point on the rim of the head component, to avoid disturbing the movement during measurement, and then waterproofed with silicone and fine poly tubing. The other sensor parts were also waterproofed with silicone, and the whole assembly could be sterilized. Surgical glues were used for adhesion of the individual parts of the sensor assembly. The sensors themselves and other non-sterilizable devices, such as the personal computer and A/D converter, were connected by an easy handling connector to enable quick connection just prior to the measurement during surgery.

Figure 2. The physical model of the femoral head component manufactured from the CAD data in plastic using a laminate molding technique. All sensors and wiring are collected within the head component, which has a diameter of 32 mm. [Color version available online.]

Measuring software

Next, software was developed to control the measurement system and to display and record the measured data (Figure 3). The pressure values and 3D directions were recorded at a frequency of 1000 Hz. The directions and values of the forces could be intuitively assumed by the surgeons with the help of 3D computer graphics. The distribution of forces loaded by the cup component was also displayed on the computer monitor using color mapping on the spherical surface according to the force values (Figure 4). The loads at points between those with embedded sensors were linearly interpolated using the pressure data from the points with sensors. In addition, a digital video camera recorded a motion picture sequence in synchrony with the pressure data measurements.

Figure 3. The displays of the measuring software. (a) Graphs indicating the transitions of the pressure values of each sensor. (b) Pressure direction (top) and distribution (bottom) over the spherical surface. (c) Real-time video captured by the digital video camera. (d) The window that controls the sensor system and records the data. [Color version available online.]

Figure 4. Examples of the estimated pressure distributions. The pressure distributions when the sensor receives loading from the cup component in three directions, designated a, b and c, are depicted. [Color version available online.]

Accuracy evaluation test

To assess the accuracy of the sensor system, we conducted experiments to measure the pressures under known loading conditions. A sensor was fixed by a high-accuracy 6-degree-of-freedom positioning device at a certain angle, followed by load application using a digital force gauge controlled by the positioning stand. The cup component was attached to the tip of the force gauge. Load was applied from several directions while the center of the sensor assembly and center of the cup component were accurately aligned to the same position. A total of 10 trials, each of 30 seconds duration, were conducted for 10 different directions.

Clinical experiments

The sensor system was evaluated in a clinical trial. To measure the soft-tissue balance under anesthesia, the pressure sensors were used following anesthesia and just before insertion of the real femoral head. The pressure at the hip joint surface was measured for 17 hip positions. As the hip joint angle at each hip position was measured quantitatively using a surgical navigation system, the surgeon could evaluate the relationship between the pressure at the hip joint and the hip joint angle in real time.

Results

Pressure sensor module

Figure 1 shows the CAD data created when the customized femoral head component was designed. Eight holes were created in the spherical surface at almost equal intervals to enable the sensors to be embedded, as well as providing a passage to channel the wiring of each sensor to just below the spherical surface. Figure 2 shows the manufactured physical model, including the 8 pressure sensors and waterproofed wiring. The diaphragm of each sensor was covered with plastic parts to reduce friction against the inner surface of the cup component.

Measuring software

Figure 3 shows the display of the measurement software. Two-dimensional graphs indicating the transitions of the pressure values of each sensor (panel a) and 3D computer graphics depicting the direction of the resultant force on each sensor and the pressure distribution over the spherical surface (panel b) can be observed simultaneously. Panel c in Figure 3 shows the video captured in real time by the digital video camera connected to the computer, while panel d shows the window that controls the sensor system and records the data. Examples of the estimated pressure distribution are shown in Figure 4. Estimated pressure distributions when the sensor receives loading from the cup component in three directions, designated a, b and c, are depicted. Figure 5 shows the system configuration for the intraoperative pressure measurement experiments using our system. The whole system was basically divided into two parts: the sensor part and the control part. The sensor part was sterilized prior to surgery and connected to the control part just before the experiment via the easy handling connector. All the subminiature pressure sensors operated correctly during the surgical procedure and the pressures at 17 hip positions were recorded.

Figure 5. System configuration for the clinical evaluation experiment. The sensor part was sterilized prior to surgery and connected to the control part just before the experiment. [Color version available online.]

Accuracy evaluation test

Figure 6 shows the results of one trial of the accuracy test. The graph shows the measured and applied forces during the 30-second trial. The force was applied from the oblique direction indicated by the red vector in the left portion of the figure. The forces were divided into three components, designated F1, F2 and F3, as indicated by the blue, pink and yellow arrows, respectively. When the results of all 10 trials for 10 different load directions were calculated together, the error between the applied loads and measured forces was 3.42 ± 3.26 N (mean ± standard deviation).

Figure 6. Results of one trial of the accuracy test. The graph at right shows the measured and applied forces during a 30-second trial. The applied force direction is shown at left. [Color version available online.]

Clinical experiments

The results of the clinical trial for measured pressure for one hip position are shown in Figure 7. A skeletal structure model constructed from the patient's CT data was simultaneously visualized on the computer monitor. The red arrow indicates the direction and intensity of the pressure. A surgical navigation system recorded the hip joint angle in synchrony with the pressure value measurement.

Figure 7. Display of measured pressure at a given moment. (a) Computer display of the software. The red arrow in the left window indicates the direction and intensity of the pressure. The pressure distributions were mapped using color. (b) and (c): 3D views from the posterior and looking diagonally backward. [Color version available online.]

Discussion

A novel system that can analyze the pressure distribution on the femoral head component in real time during THA was developed. Using this system, surgeons can evaluate the soft-tissue balance around the hip joint and select the best-fitted implant for each individual patient from several alternatives chosen in accordance with the preoperative planning system.

Regarding pressure sensors for hip prostheses, several types of device have been developed. Some of these use a few strain gauges and only measure the contact force direction loaded by the cup component [8–10]; they cannot measure the distribution of pressure over the surface of the head component. A flexible flat array of capacitive pressure sensors was previously developed and applied to measure the pressure distribution over the hip joint [11]. However, the measurement frequency of that system was 50 Hz, while that of our system is 1000 Hz, and the capacitive sensory array has not been applied clinically due to its problematic thickness. Some researchers have developed wireless sensor assemblies consisting of several force or temperature sensors and a telemetry system and embedded them in the stem component [12–14]. Although these systems could acquire data, even for postoperative daily activity, and provide useful information for understanding human hip dynamics, they cannot be applied to all patients undergoing surgery due to their cost and certain ethical issues. In addition, since these telemetry systems require relatively large devices for modulators and/or the power supply, they cannot be assembled in the head component. In our sensor system, all the sensors and wiring are assembled in the head component, such that it can manufactured at relatively low cost and used multiple times with sterilization after each use.

We believe that our pressure sensor system has two main advantages. The first advantage is the real-time measurement and display of the 3D pressure distribution. The sensor can be used for successive measurements in real time, and the recorded pressure distribution is directly displayed on a computer monitor in the form of 2D graphs for quantitative analysis and 3D computer graphics of the head and stem shape with color maps to indicate the pressure distribution for an intuitive understanding of the current hip situation. The surgeon can check the pressure during movement and also confirm the dynamic situation, such as the moment immediately before dislocation. Thus, real-time measurement is especially important since the nature of joints is dynamic, not static.

The second advantage is the ease of customization for each patient. Since all the sensors and wiring are assembled in the head component, we can test several types of neck during surgery when a stem with an interchangeable neck is used. Furthermore, the head is designed using a 3D CAD system so it is easy to change the diameter of the head or taper angle at the connection with the stem component. Hence, in the foreseeable future, it will be possible to manufacture several types of customized head with varying diameters and taper angles.

We integrated our pressure distribution analysis system with the surgical navigation system and measured hip joint angles and pressures simultaneously. In the next step, we will integrate this measurement system with a simulation system that we have been developing to estimate the postoperative daily motion of patients following THA [15],[16]. This simulation system uses the patient-specific skeletal model and the patient's motion data acquired by an optical motion capture system, and can visually represent the movement of the patient's bones and estimate the impingement between the hard tissues around the hip joint. By integrating the intraoperative pressure measurement system with the simulation system, we will be able to estimate the soft-tissue balance of postoperative activities during surgery and select the best implant type with consideration of postoperative complications such as dislocation or loosening.

We believe that our system has possibilities for revealing the dynamic mechanics of an installed artificial hip joint. Further development of the calibration method should enable more accurate measurements, thus making it a useful tool for selecting the appropriate implant to be fitted in a patient or for estimating the risk of complications following surgery. The system will also be useful for evaluating joint dislocation mechanisms by comparing the measured soft-tissue-generated forces with those estimated by the simulation system.