Filtering of Corneal Images using Hybrid Wavelet Transform in the Cases of Keratoconus

Abstract

One of the most prevalent, bilateral, asymmetric, and progressive corneal diseases, keratoconus can have a slight to severe impact on vision. Early on, the condition is frequently misdiagnosed as irregular astigmatism, delaying diagnosis. Although we have cutting-edge diagnostic techniques, the results are insufficient to fully assess the corneal health at different areas, making it challenging to plan additional treatment programmes. Here, image pre-processing techniques using a Hybrid Wavelet Transform of Discrete Wavelet Transform (DWT) and Stationary Wavelet Transform (SWT), followed by soft and/or hard thresholding and Inverse Wavelet Transform, are proposed in order to achieve early and accurate diagnosis and assess the health of the cornea. The qualitative and quantitative metrics are reached by taking into account the several Electronic Corneal Topography picture modes, which would be useful to an ophthalmologist in moving on with therapy. This approach has been proven to have greater promise than the ones currently in use, particularly in relation to corneal diseases like keratoconus. Additionally, this approach aids in more accurate keratoconus stage determination.