Curvatures with uncertainties derived in conformal space to characterize tendon microstructure

There is a growing demand for non-invasive methods to diagnose tendon injuries and monitor the healing processes of their repair. To date there is limited knowledge on their structure and function and the interlink between these. One of the potential targets in this investigation is the extracellular matrix (ECM) that captures its structural changes. Recently we reported on the assessment tendon damage on a macroscopic level from high field MR signals. In this paper, we present a methodology that enables structural description on a microscopic level. We derived curvature values from the conformal monogenic signal, which however can become unreliable in the presence of noise. To account for this we use non parametric noise properties and a 1D feature based uncertainty measure in an iterative framework using Hidden Markov Measure Field (HMMF). The proposed method reveals that curvature values derived from normal tendon tissue microscopy images are higher and more homogenous than curvature values derived from the damaged tendon images