Improving accuracy and efficiency of mutual information for multi-modal retinal image registration using adaptive probability density estimation

Abstract

Mutual information (MI) is a popular similarity measure for performing image registration between different modalities. MI makes a statistical comparison between two images by computing the entropy from the probability distribution of the data. Therefore, to obtain an accurate registration it is important to have an accurate estimation of the true underlying probability distribution. Within the statistics literature, many methods have been proposed for finding the 'optimal' probability density, with the aim of improving the estimation by means of optimal histogram bin size selection. This provokes the common question of how many bins should actually be used when constructing a histogram. There is no definitive answer to this. This question itself has received little attention in the MI literature, and yet this issue is critical to the effectiveness of the algorithm. The purpose of this paper is to highlight this fundamental element of the MI algorithm. We present a comprehensive study that introduces methods from statistics literature and incorporates these for image registration. We demonstrate this work for registration of multi-modal retinal images: colour fundus photographs and scanning laser ophthalmoscope images. The registration of these modalities offers significant enhancement to early glaucoma detection, however traditional registration techniques fail to perform sufficiently well. We find that adaptive probability density estimation heavily impacts on registration accuracy and runtime, improving over traditional binning techniques.

Retinal image registration under the assumption of a spherical eye

“…For retinal images, usually methods based on mutual information (Pluim et al, 2003;Legg et al, 2013) have been proposed. Instead of employing all image pixels, certain feature-based methods rely on carefully selected, localized features.…”

Photo-Optical In-Situ Measurement of Drop Size Distributions: Applications in Research and Industry

“…N = 29,861 in e06 (Sturges, 1926). Since Sturges' Rule is known to lead to an over-smoothed histogram, especially for large samples, and only considers normal, not skewed distributions (Legg et al, 2013), a higher number was chosen. This also results in a more representative Cumulative Distribution Function (CDF) and leads to preservation of the shape of both density distributions in Figure 9 although they span different intervals.…”

KDE based outlier detection on distributed data streams in multimedia network

“…The best kernel width selection methods include rules of thumb, over smoothing, least squares cross-validation, biased cross-validation, direct plug-in methods, and the smoothed bootstrap. And the widely used rule for approximating the kernel width is the Scott's rule [22,24,32], where h is expressed in (4). However, these strategies depend on the complete sample, which is impracticable in data stream scenario.…”