Estimation of mechanical parameters in cancers by empirical orthogonal function analysis of poroelastography data

Md Tauhidul Islam, Raffaella Righetti

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


Ultrasound poroelastography is a non-invasive imaging modality that has been shown to be capable of estimating mechanical parameters such as Young's modulus (YM), Poisson's ratio (PR) and vascular permeability (VP) in cancers. However, experimental poroelastographic data are inherently noisy because of the requirement of relatively long temporal data acquisitions often in hand-held mode conditions. In this paper, we propose a new method, which allows accurate estimation of YM and PR from denoised steady state axial and lateral strains by empirical orthogonal function (EOF) analysis of poroelastographic data. The method also allows estimation of VP from the time constant (TC) of the first expansion coefficient (EC) of the temporal axial strain, which has larger dynamic range and lower noise in comparison to the actual temporal axial strain curve. We validated our technique through finite element (FE) and ultrasound simulations and tested the in vivo feasibility in experimental data obtained from a cancer animal model. The percent relative errors (PRE) in the estimation of YM, PR and VP using the EOF analysis as applied to ultrasound simulation data were 3.27%, 3.10%, 14.22%, respectively (at SNR of 20 dB). Based on the high level of accuracy by EOF analysis, the proposed technique may become a useful signal processing technique for applications focusing on the estimation of the mechanical behavior of cancers.

Original languageEnglish (US)
Article number103343
JournalComputers in Biology and Medicine
StatePublished - Aug 2019


  • Cancer imaging
  • Elastography
  • Poisson's ratio
  • Poroelastography
  • Vascular permeability
  • Young's modulus

ASJC Scopus subject areas

  • Health Informatics
  • Computer Science Applications


Dive into the research topics of 'Estimation of mechanical parameters in cancers by empirical orthogonal function analysis of poroelastography data'. Together they form a unique fingerprint.

Cite this