The Contribution of Geometrical Properties to Pubic Rami Fractures in Lateral Impact

Journal of Young Investigators
Undergraduate, Peer-Reviewed Science Journal

Volume Four

RESEARCH ARTICLE

RECENT ISSUES | ARCHIVES | RESOURCES | JYI NEWS | ABOUT JYI

Issue 1, June 2001

Engineering & Applied Sciences
The Contribution of Geometrical Properties to Pubic Rami Fractures in Lateral Impact

Adam J. Engler
Department of Bioengineering, University of Pennsylvania
Advisor: Alan W. Eberhardt, Ph.D.
Department of Biomedical Engineering, University of Alabama at Birmingham

Abstract

Pelvic fracture has mainly been studied in terms of motor vehicle crashes (MVC) or through component studies of the mechanical properties of bone. Consequently, there are many knowledge gaps in the dynamic behavior of various regions of the pelvis. One such region is the pubic ramus bone, in which fractures are common in both dynamic lateral impact tests and automobile crash data (Guillemot 1997). This suggests that pubic rami, a group of four bones in the front of the pelvis (inferior and superior bones for both the left and right sides), have insufficient resistance against lateral stress and fracture. The present study showed that the influence of the geometrical properties of area ratios and area moments of inertia on pubic rami play a significant role during impact. This study confirmed that fractures commonly occur at the smallest total area and centroid-external surface distance where this stress maximums and minimum area moments are found. Area ratios were significantly different for post-impact fractured and non-fractured superior rami (p = 0.02), but not in the inferior rami (p = 0.97). Significant differences were observed as a function of position along the superior rami (p = .009), due to local increases in cortical bone area. Notably, the area moments were significantly lower for the unfractured superior rami than the fractured. The previously described trend of increased area ratio for the unfractured superior rami, this suggests that highly dense, yet compact cross sections are more resistant to fracture. This location coincided with the minimum modeled rami cross section, supporting the current results.

NOMENCLATURE

AR
PL
I_xx^'
I_yy^' Area ratio (a ratio of the cortical bone area divided by the trabecular bone area)
Percent length from the pubic symphysis joint
Area moment of Inertia about the inferior-superior axis
Area Moment of Inertia about the anterior-posterior axis

Introduction

Pelvic fracture, cracking or breaking of a portion of the pelvis, is an extremely common injury in side-impact collisions of motor vehicles. As the National Accident Sampling System (NASS) showed, side impacts represented 22% of the collisions from 1980 to 1992. Of all pelvic fractures, pubic rami fractures have been shown to be common in motor vehicle collisions (Guillemot 1997) and in lateral impact tests (Guillemot 1997, Cesari and Ramet 1982, and Arbelaez 1999). This suggests that pubic rami (outlined in Fig. 1 radiograph), a group of four bones in the front of the pelvis (inferior and superior bones for both the left and right sides), have insufficient resistance against lateral loading characteristic of side impacts. To better understand pubic rami fractures and their severity, it is necessary to describe the properties that contribute to the likelihood of pubic rami fracture. Geometrical properties of bone cross sections have been shown to correlate with bone strength in long bones (Ruff and Hayes 1984) and in rami strength in lateral impact tests (Cesari and Ramet 1982).

Figure One

Previous research at the Musculoskeletal Mechanics Laboratory at the University of Alabama at Birmingham (UAB) has involved cadaveric dynamic impact testing of isolated human cadavers (Molz et al.. 1997 and Arbelaez 1999). Molz et al. conducted impact tests on pelves supported in a containment fixture through which a compressive pre-load is applied through the lower spine, and lateral support was applied at the iliac wing. This support condition resulted in loading rates of 4 - 18 kN/msec. Arbelaez used an additional lateral support at the greater trochanter, which resulted in loading rates that did not exceed 3 kN/msec. In all, 25 cadavers were impacted, yielding 11 pubic rami fractures and five acetabular fractures. Three-dimensional finite element (FE) models of a human pelvis (Plummer et al. 1996) has suggested temporal failure relations under these different impact events. The present study was undertaken to quantify geometrical properties of pubic rami from the impacted pelves, as a function of position along the bone. The positional variations of these properties were examined for correlation with fracture site, and compared with peak stress predictions from the FE models.

Materials & Methods

Anatomically oriented cross sections of bone (Fig. 2) were removed from each post-impact pubic ramus of seven rami-fractured and six non-fractured cadavers (impacts preformed by Arbelaez 1999) using a diamond circular saw and jig to insure precision cutting and uniform thickness. Digital images of each section were taken using a Nikon digital camera using a 120 mm Medical Nikkor lens. Adobe PhotoShop 3.0.3 with IP Plug-ins (Adobe Systems Inc., San Jose, CA) was used to outline cortical and trabecular regions of bone. These traces were then used to calculate area ratios of the cross sections (in which a ratio was generated from the cortical bone area divided by the trabecular bone area), on a pixel by pixel basis.

Figure Two

The images were then imported into I-DEAS software (Structural Dynamics Research Corporation, Milford, OH) for calculation of area moments of inertia about a centroidal x-y coordinate system. Principal moments of inertia were found as the eigenvalues of the matrix of area moments.

Mean area ratios and mean moments of inertia of each rami section were plotted as a function of the position along the rami (see Fig. 4 for relative position information) for fractured and non-fractures rami. Statistical analyses (Statview, Abacus Concepts, Inc., Berkeley, CA) were performed to determine if there were significant differences between the regressions. While area ratio and area moment trends show differences in the overall strength and structure of the pubic rami, they do not point to why pubic rami constantly fractured at certain locations under similar loading conditions. Focusing on the fracture site of the rami, minimum area ratios and area moments were correlated with stress peaks, using a dynamic FE model from Plummer et al., in order to see if the weakest cross section correlated to the site of fracture and to evaluate the best indicator of fracture. The dynamic model simulated an impacted on the right acetabulum with a fixed left iliac wing (top left fixture point). It did not include mechanical and material properties for the sacroiliac joint and sacrum (back of the pelvis), but these are not as important since this study focuses on the front of the pelvis.

Results

Figures 3 (a) and (b) show the mean + standard deviations of the area ratios plotted as a function of position along the rami for superior and inferior cases of fractured and non-fractured cortical bone. Non-fractured rami are shown in red, while fractured cases are in blue.

Figures Three (a) and (b)

Analysis of covariance showed that the area ratios were significantly different comparing fractured and non-fractured superior rami (p = 0.02) but not in the inferior rami (p = 0.97). Significant differences were observed as a function of position along the superior rami (p = .009), due to increases in cortical bone area. Pooled t-tests indicated no significant differences between females and males with respect to area ratios for either the inferior or superior rami. With regard to age, a significant increase in area ratio was observed comparing superior rami for under-65 years old to over-65 years old (p = 0.005), however, this was not the case for the inferior rami (p = 0.13).

The area moments of inertia of each cross section showed unexpected trends. I_xx^' and I_yy^' calculations were performed comparing fractured versus nonfractured, gender and age, as well as positional dependence along the superior and inferior rami. Notably, the area moments were significantly lower for the unfractured superior rami than the fractured. Noting the previously described trend of increased area ratio for the unfractured superior rami, this suggests that highly dense, yet compact cross sections are more resistant to fracture. Other important differences were observed comparing I_xx^' and I_yy^' for both inferior and superior rami for the over-65 age group and the under-65 group. In this case, comparisons yielded significant differences in both principal area moments for both inferior and superior rami, with greater values observed for the older group. Taken in conjunction with the area ratios, this suggests substantial increases in cross section with age. No significant differences in area moments were observed with differences in gender.

Of the 11 total experimental ramus fractures (from 7 fractured pelves), five were examined for fracture site-specific area ratio and moments (all full support cases, Arbelaez 1999), while the other six were excluded due to inconclusive radiographs caused from inaccurate location of the micro and macro fractures in the bone. The mean fracture location was 36.4 + 29.1% of the total pubic rami length. Observation of these fracture sites revealed a strong correlation with minimum area moments, but a weak correlation with minimum area ratio. Of the 5 fractures examined, the cortical area ratio was 34.66 ± 3.66%, the I_xx^' value at the fracture site was 24,700 ± 13,800 mm4 and the I_yy^' value was 10,200 + 4,850 mm4. The minimum I_xx^' and I_yy^' occurred, respectively, at 40% and 43% of the total pubic rami length (Fig. 4).

Figure Four

It was also hypothesized that the maximum stress location, as predicted by dynamic FE modeling, would correspond with the fracture sites. For the FE model of Plummer et al., which modeled wing and greater trochanter support, peak stress in the cortex occurred at 45% of the total pubic rami length on the superior ramus (Fig. 5). This location coincided with the minimum modeled rami cross section, supporting the current results.

Figure Five

Discussion

As described in the literature, pubic rami fractures are a very dangerous and common pelvic injury resulting from the abnormal loading of the pelvis in lateral motor vehicle collisions (MVCs). Young and Resnick (1990) indicate that pelvic fractures were the most costly injuries in automobile crashes, resulting in 18-24% sustaining debilitating injury. The current study, which included fracture specific analysis, found correlations between fracture site and geometrical parameters, which further coincided with stress maxima in FE models. These indicators all point to the smallest cross sectional area as the most vulnerable point during lateral impact, which confirms Cesari and Ramet's finding of correlation between the value of area moments and impact tolerance. Furthermore, it appears that increased cortical area ratios in the superior rami reduce the fracture potential in side impact.

The present study is limited by its relatively small size. While specific numbers may change with the inclusion of more pelves, the authors believe the current trends are reasonable. The calculations of I_xx^' and I_yy^' could be further refined through the inclusion of trabecular bone. The significant trends associated with age, but not with gender, were surprising and warrant further study.

Conclusions

The present study measured geometrical properties, including area ratios and moments of inertia, of pubic rami cross-sections, and correlated the measures with the occurrence of fracture in experimental side impacts. The results suggest that the area moment of cortical bone in the rami best indicates fracture resistance, while the ratio of cortical to trabecular bone is a good secondary predictor. Current research in car door padding technology has been greatly focused on the side-impact force and stress distribution on the pubic rami as a result of this study. New padding designs have been focused on shielding the impact seen by a pubic rami and limiting the stress wave, thus improving pubic rami fractures survivability.

Acknowledgements

This study was supported by the University of Alabama at Birmingham Injury Control Research Center (CDC Grant #R49/CCR403641) through the National Center for Injury Prevention and Control at CDC. The authors would like to thank Jason Plummer, of the UAB Department of Biomedical Engineering, for his assistance with the dynamic models used in the study and of Martha Wilkins, of the UAB Orthopedics Laboratory, for her help with all digital imagining.

References

Arbelaez, R.A. (1999) "Fracture Tolerance and Fracture Control Strategies for the Pelvis in Automotive Side Impact:" Master's Thesis, University of Alabama at Birmingham Press: Birmingham.

Cesari, D. and Ramet, M. (1982) SAE 821159.

Guillemot, H et al (1997) SAE 973322.

Molz, FJ; George, PD; Bidez, MW; King, AI; Alonso, JE (1997) Simulated Automotive Side Impact on the Isolated Human Pelvis: Phase 1: Development of a Containment Device. Phase 2: Analysis of Pubic Symphysis Motion and Overall Pelvic Compression. STAPP Car Crash Conference SAE #973321.

Plummer, JW; Bidez, MW; Alonso, JE (1996) Parametric Finite Element Studies of the Human Pelvis: The Influence of Load Magnitude and Duration on Pelvic Tolerance during Side Impact. STAPP Car Crash Conference. SAE #962411.

Ruff CB and Hayes WC. (1984) Bone-Mineral Content in the Lower Limb. Relationship to Cross-Sectional Geometry. J Bone & Joint Surgery - American Volume. 66(7):1024-31

Young JW and Resnik CS. (1990) Fracture of the Pelvis: Current Concepts of Classification. Am. J of Roentgenology. 155(6):1169-75

SEARCH | SITE MAP | RECENT WEB SITE ADDITIONS PRIVACY POLICY | CONTACT US

JYI is supported by: The National Science Foundation, The Burroughs Wellcome Fund, Glaxo Wellcome Inc., Science Magazine, Science's Next Wave, Swarthmore College, Duke University, Georgetown University, and many others.