Poroviscoelastic Modeling of Perfused Bovine Liver Tisue

INTRODUCTION Blunt and penetrating liver injuries make up one‐third of all abdominal injuries, and 5% of all trauma admissions are due to injuries to this organ [1]. Additionally, the majority of blunt liver injuries stem from motor vehicle crashes (Moore 2004). Validated computational models of liver are useful for understanding injury to the liver from mechanical loading. It has been shown previously that the mechanical behavior of perfused ex vivo liver closely approximates the in vivo behavior, while unperfused ex vivo liver response differs significantly from the in vivo case [2]. Therefore, to achieve a realistic approximation of in vivo liver mechanics for injury prediction, it is beneficial to adopt a constitutive model for liver that incorporates both fluid and solid phases of the tissue, to reflect this perfusion effect. The goal of this study is to characterize the mechanical response of perfused bovine liver tissue to unconfined compression using poroviscoelastic (PVE) modeling. This modeling approach quantifies both the solid and fluid phases of the tissue [3], considering the solid phase as viscoelastic [4]. The key advantage of PVE liver models over traditional viscoelastic models is the ability to examine pore fluid pressure, since fluid pressure has been shown to correlate well to injury severity [5]. In previous work, PVE modeling was shown to capture the mechanical behavior of unperfused samples of porcine liver tissue in unconfined compression [6]. The current study extends this work to examine the effectiveness of PVE modeling to capture perfused liver mechanical response. MATERIALS AND METHODS Unconfined Compression Experiments Bovine livers were acquired at a local abattoir and stored in Dulbecco’s Modified Eagle Medium (DMEM) until tested (within 36 hours post mortem). Samples approximately 27mm thick and 40mm in diameter with an intact portal vein of 4cm were collected manually using a surgical blade. Compression testing was conducted using the Bose Electroforce Test Instrument and WinTest Software. The perfusate reservoir was placed 66cm above the testing surface in order to create physiologically accurate interstitial fluid pressures (0‐6mmHg) in the liver sample. The sample was then perfused with room‐temperature DMEM for 10 minutes until thermal equilibrium was established. Continuous flow of perfusate exiting the sample created an approximately frictionless boundary between sample and platens. A total of four samples were pre‐loaded to ‐10g at a rate of 0.001s in order to achieve uniform surface contact and then thickness was measured. After pre‐load, a ramp displacement to 5% strain was applied at a rate of 0.001s. Reaction forces were monitored with a 1000g submersible load cell, and samples were monitored for 900 seconds of relaxation following ramp loading.