Hyperspectral characterization of re‐epithelialization in an in vitro wound model

Abstract

In vitro wound models are useful for research on wound re-epithelialization. Hyperspectral imaging represents a non-destructive alternative to histology analysis for detection of reepithelialization. This study aims to characterize the main optical behavior of a wound model in order to enable development of detection algorithms. K-Means clustering and agglomerative analysis were used to group spatial regions based on the spectral behavior, and an inverse photon transport model was used to explain differences in optical properties. Six samples of the wound model were prepared from human tissue and followed over 22 days. Re-epithelialization occurred at a mean rate of 0.24 mm 2 /day after day 8 to 10. Suppression of wound spectral features was the main feature characterizing re-epithelialized and intact tissue. Modeling the photon transport through a diffuse layer placed on top of wound tissue properties reproduced the spectral behavior. The missing top layer represented by wounds is thus optically detectable using hyperspectral imaging.

Exploiting scale-invariance: a top layer targeted inverse model for hyperspectral images of wounds

“…Histologies were not available for the wound model samples in this study, and repetition of the experiment is necessary for proper attribution of the reflectance changes to corresponding changes in epidermal layer composition. However, the epidermal layer presence indicated by the inverse model results is in agreement with statistical characterization of these data [38]. .…”

“…Photon transport modeling of in vitro wounds can be challenging. The spectral characteristics of the dermal part of the wound models are less defined than in normally circulated tissue due to the lack of blood absorption, and the spectra are significantly influenced by the spectral properties of the growth medium [38]. However, in wounds, both regions with and without the upper tissue layer are present.…”

“…The high number of available wound spectra in a hyperspectral image makes selection of one unique spectrum somewhat prohibitive. However, the same availability of multiple spectra with a high spectral resolution makes it possible to use dimensionality reduction methods to remove redundant information and represent the spectral information in a low-dimensional space [38]. Principal component analysis (PCA) is such a method, which can decompose a dataset in terms of orthonormal components (loadings) that can linearly transform each observation into new variance-maximizing coordinates (scores) [39].…”

“…Principal component analysis (PCA) is such a method, which can decompose a dataset in terms of orthonormal components (loadings) that can linearly transform each observation into new variance-maximizing coordinates (scores) [39]. This technique has been used as a pre-processing technique [40][41][42] and for investigation of spectra in a low-dimensional space [6,19,38]. The method is used in the current paper to reduce a discrete wound spectrum choice to a continuous choice of PCA scores that can be back-transformed to a wound-like spectrum using the inverse transform.…”

Comparative analysis of denoising techniques in burn depth discrimination from burn hyperspectral images

“…Not the same attention was paid to this subject in medical hyperspectral imaging. Only the use of maximum noise fraction (MNF) transform [13,14] and principal component analysis [15,16] as noise reduction techniques in skin hyperspectral images have been reported so far without any evaluation of their effects on the performance of medical hyperspectral data analysis.…”