Abstract
Introduction
In Total Hip Arthroplasty (THA), proper bone preparation technique is fundamental to preventing intraoperative fracture. Anecdotally, surgeons suggest they can avoid fracture by listening for changes in the pitch of a mallet strike during broaching. Consequently, it is not surprising that researchers have explored vibroacoustic methods to prevent [1] and identify bone fractures [2, 3]. For instance, a shift in frequency of the acoustic signals during impaction has been correlated with initial stability [4, 5]. In-spite of these research-based successes, we are unaware of an intraoperative application for THA. We submit that idiosyncratic variability during impaction [6] may overwhelm analytical techniques developed in a controlled laboratory environment. The purpose of this test, therefore, was to evaluate the effect of several strike parameters on the vibro-acoustic response during impaction. Specifically, we hypothesized that the angle, location, and force of impaction would produce ‘false-positives’ in frequency regions that have been used to identify fracture [7].
Methods
A Sawbones femur (SKU1121, Medium) was prepared and broached using standard surgical technique for the Summit hip system (DePuy Synthes) progressing from size 0 to 4. The size 4 broach was firmly seated and impacted ten times (n=10) for each of the prescribed conditions (Table 1) while securely holding the femur by hand. Vibroacoustic data from an accelerometer attached distally on the femur and a directional microphone located within 1 metre (Figure 1) were acquired at a sampling rate of 40kHz and postprocessed using LabView. Spectrograms were generated for qualitative comparisons, while fast fourier transform (FFT) with normalised amplitudes for each strike facilitated quantitative analysis of the area under the FFT curve (AU-FFT). Strike conditions were monitored to ensure the groups were consistent and distinct (Table 1).
Results
There were statistically significant differences in strike conditions for angle (30°vs 60°), location (centre vs medial and lateral) and force (medium vs low and high) (Figure 2). Data describing the strike conditions revealed consistent and distinct groups (data not shown).
Discussion and Conclusion
We have demonstrated that variability in striking does influence the vibroacoustic signal during impaction; however, contrary to our hypothesis, this variability does not overwhelm the ability to distinguish between fractured and intact impaction signals. Consequently, the AU-FFT comparator could be a robust and useful metric. Future work could evaluate this technique under more diverse conditions with multiple samples of varying anatomies, densities, and degrees of fracture. The above methods and paradigms could further be investigated to discern when a broach is properly seated and thereby avoid the risk of fracture altogether.
