The Ministry of Education of the Republic of Belarus

The Belarusian State University

The English Language Department for Sciences

Simulation of microbiological objects fluorescent images

Maksim Shytsik

a Master’s Course Student

of the System Analysis

and Computer Modelling Department

of the Radiophysics and

Computer Technology Faculty

Supervisor:

PhD N. N. Yatskou

Minsk, 2015

CONTENTS

ABSTRACT 3

АННОТАЦИЯ 3

SIMULATION OF MICROBIOLOGICAL OBJECTS FLUORESCENT IMAGES 5

CONCLUSION 13

BIBLIOGRAPHY 14

GLOSSARY 15

Abstract

Key words: confocal microscopy, modelling, automatic analysis, cells, microbiological objects, cancer.

The success of digital technologies in image acquisition has promoted the development of automatic cytometry – cells and their substructures properties analysis. The efficiency and robustness of automatic analysis algorithms may be improved by modelling synthetic images, which allows defining basic features of objects and the measurement system. This paper proposes a simulation algorithm and its practical implementation to create fluorescent images of microbiological objects. The comparison of generated and experimental cancer tumors images confirms their similarity, which allows using the developed method to study and debug algorithms.

Аннотация

Ключевые слова: конфокальная микроскопия, моделирование, автоматический анализ, клетки, микробиологические объекты, рак.

Успехи применения цифровой техники при получении изображений способствовали развитию автоматической цитометрии – анализа свойств клеток и их подструктур. Повысить эффективность и устойчивость алгоритмов автоматического анализа может моделирование синтетических изображений, позволяющее определить основные свойства объектов и измерительной системы. В данной работе предложен алгоритм моделирования и его реализация для создания люминесцентных изображений микробиологических объектов. Результаты сравнения полученных изображений раковых опухолей с экспериментальными подтверждают их схожесть, что позволяет использовать предложенный метод при исследовании и отладке алгоритмов.

Introduction

The success of digital technologies in image acquisition has promoted the development of automatic cytometry – cells and their substructures properties analysis. The efficiency and robustness of automatic analysis algorithms may be improved by modelling synthetic images, which allows defining basic features of objects and the measurement system [1]. Varying simulation parameters allows one to study robustness of automatic analysis algorithms to different influences which appear in the process of image acquisition and to define the most effecting factors during experiments [2].

It is a complex process to simulate images with parameters similar to real features. Nevertheless, basic features of objects and the measurement system can be studied when some real objects characteristics are neglected. Furthermore, cell simulation is not possible without simplifications [2].

This paper proposes a simulation algorithm and its practical implementation to create fluorescent images of microbiological objects. It has allowed producing a list of fluorescent images of cancer tumors. The statistical analysis was carried out to check the model significance. The comparison of generated and experimental images confirms their similarity, which allows using the developed method to study and debug algorithms.

The obtained images enable to reveal qualitative morphological system properties. They can be used to measure certain tissue areas characteristics. A wide possible simulation parameters list provides generating diverse sets of images.