• S A Shah, N Bachrach, S J Spear, D S Letbetter, R A Stone, R Dhir, J W Prichard, H G Brown, and W A LaFramboise.
• TissueInformatics, Pittsburgh, PA, USA.
• BioTechniques. 2003 Feb 1; 34 (2): 408-13.

AbstractA correlative bright-field and hyperspectral analysis of full-thickness, cutaneous wounds in a porcine model was undertaken to investigate the efficacy of hyperspectral imaging as an alternate method for wound identification. Analysis of a randomly selected specimen yielded distinct spectral signatures for cutaneous regions of interest including the epidermis, injured dermis, and normal dermis. The scanning of the entire specimen group using these hyperspectral signatures revealed an exclusionary, pseudo-color pattern whereby a central wound region was consistently defined by a unique spectral signature. An algorithm was derived as an objective tool for the comparison of the wound regions defined by the hyperspectral classification versus the pathologists' manual tracings. The dimensions of the wound identified in the hyperspectral assay did not differ significantly from the wound region identified by the pathologists using standard bright-field microscopy. These data indicate that hyperspectral analysis may provide a high-throughput alternative for wound estimation that approximates standard bright-field imaging and pathologist evaluation.

29 keywords...

hide…