The properties of the cornea based on hyperspectral imaging

Michael, Kaleena and Siti Salwa Binti Md Noor, S and Tschannerl, Julius and Ren, Jinchang and Marshall, Stephen (2017) The properties of the cornea based on hyperspectral imaging. Investigative Ophthalmology and Visual Science, 58 (8). 2439. ISSN 0146-0404

[thumbnail of Kaleena-etal-IOVS-2017-The-properties-of-the-cornea-based-on-hyperspectral]
Preview
Text. Filename: Kaleena_etal_IOVS_2017_The_properties_of_the_cornea_based_on_hyperspectral.pdf
Accepted Author Manuscript

Download (239kB)| Preview

Abstract

Purpose : Hyperspectral Imaging (HSI) is a hybrid modality that combines imaging and spectroscopy. Here we investigate the ability of a hyperspectral device in extracting data from the layers in the porcine corneal tissue through the wavelength spectrum, in foreseeing its potential in clinical diagnostics by simplifying methods of examination by clinicians in detecting corneal injuries. Methods : Hyperspectral imaging using 400 to 1000nm visible wavelength camera was used to scan five porcine eyes, containing a mix of eyes with intact and unintact epithelial layer. Images were saved and analysed in three dimensional rows, columns and depth slices at 1200 to 1300 x 804 x 604 resolution. The Matlab image processing toolbox was utilised to process the images for inspection in grayscale, HSV format and reflectance spectrum. All laboratory works were performed in accordance with the general risk assessment of University of Strathclyde. Results : The obtained hyperspectral images were able to demonstrate distinct differences through the wavelength. Images at longer wavelength reveal distinct shapes in regular arrangements, and could be descriptive of the physical properties of the corneal tissue in particular layers. When comparing reflectance spectrum obtained from both eyes with intact and unintact epithelium, we were able to demonstrate distinct separation in reflectance values from 578 to 818nm wavelenghth. Conclusions : Our analysis was able to demonstrate a gap in the reflectance spectrum between the intact and unintact epithelium of a porcine’s cornea, illustrating its potential value in the assessment of corneal tissue integrity. Further image processing with grayscale slices reveal distinct tissue properties at varying wavelengths strongly suggests a novel role for hyperspectral image technology in the diagnostics of corneal tissues, alongside traditional methods such as microscopy. These findings support our proposition for the role of hyperspectral imaging in aiding the development of innovative, mobile devices.