Prediction of tissue rupture from percolation of local strain heterogeneities for diagnostics

TL;DR

This paper explores how merging weak regions in tissues can predict tissue rupture before it happens, using medical imaging data.

Contribution

The study demonstrates that percolated local strain heterogeneities can predict tissue failure locations before rupture occurs.

Findings

Tissue rupture is preceded by merging of locally weak regions in mechanical strains.

Strain percolation analysis successfully predicted an aortic aneurysm in a Marfan syndrome patient from MRI data.

Mechanical heterogeneities detected via imaging correlate with future tissue failure loci.

Abstract

A plethora of medical conditions, ranging from torn ligaments to aneurysmic blood vessels, are caused by failure of mechanically stressed biological tissues until rupture. Clearly prediction of the potential loci of tissue failure prior to rupture is highly desirable for prophylactic measures, preferentially in sufficiently early stages of the disease.

Mechanical heterogeneities are identified from local mechanical strains obtained from image sequences recorded during uniaxial tensile testing of reconstituted collagen (both, in experiments and finite element model (FEM) calculations) and horse aorta explants, respectively, as well as of the pulsating aorta using magnetic resonance imaging (MRI).

Within this work we present a comprehensive study on the biomechanical concept that percolated local mechanical strain heterogeneities can serve as valid indicators to predict the loci of tissue rupture already from straining behavior within the elastic regime. While we first experimentally validate the predictive capabilities of our strain percolation analysis for reconstituted rat tail collagen fibers and horse aorta explants, we unveil the structural origins of mechanical heterogeneities on the network level using FEM calculations based on digitized confocal laser scanning microscopy (CLSM) measurements. To demonstrate the diagnostic capabilities, we successfully predict potential occurrence and location of an aortic aneurysm in a patient with documented Marfan syndrome from MRI video sequences recorded of the pulsating aorta six years prior to surgery.

Detection of local mechanical heterogeneities and their percolation behavior bears predictive capabilities for tissue failure before it actually has occurred and thus promises large potential for diagnostics and therapy.

Numerous medical conditions are related to failure of biological tissues due to mechanical stresses, resulting in tearing and rupture of tissue. We explore whether it is possible to predict tissue rupture before failure occurs. We find that tissue rupture is preceeded by a merging phenomenon of locally mechanically weak regions. Our approach can be applied to data obtained during medical imaging and so could potentially be used clinically in the future.

Schutte et al. investigate whether percolated local mechanical strain heterogeneities can predict tissue rupture. They successfully predict an aortic aneurysm in a patient with documented Marfan syndrome from magnetic resonance imaging (MRI) data, showing clinical utility of their method.

Figures