Algorithms of Experimental Medical Data Analysis

The paper is devoted to the development of a computational technique for assessing the performance of tissue regeneration in an experiment using mesh nickel-titanium implants with shape memory. Observational data obtained from electron microscopy and classical histological examination are processed and analyzed by the use of proprietary algorithms and their modifications. This can significantly facilitate the procedure of data analysis and increase the accuracy of the estimates by 5–10%. As a computational method for examining the dynamics of the studied process and determining the internal geometrical characteristics of empirical images of objects concerned, the suggested technique includes algorithms of shearlet transform, wavelet transform and construction of elastic maps for efficient visualization of spatial data. The significant part of the recommended method is the computing means of visual data preprocessing for increasing the brightness and contrast of the examined images based on the Retinex technology. This part has a significant impact on the quality of applying the tools of the computer-based evaluation presented in this work.

I. Introduction

Currently, there is extensive use of implants of mesh in different fields of plastic surgeries [1]. Among the modern hernioplasty tasks solved by utilizing implants is the restoration of deformities of the frontal wall of the abdomen, i.e., treating the defects of hernia by surgery. As part of this task, various materials and methods of histologic tissues analysis are used in experiments [2]. The wavelet and shearlet transform represents the most popular solutions as a computational method for analyzing visual data [3], [4]. So, the work of [5] describes the sophisticated processing methods of empirical medical data. The transformation of shearlet is employed as the basis for a multidimensional examination of spatial data and medical images to solve problems of histology.