Computing in Biomedicine

Facts  
Duration: 1 semester (6 weeks)
Period: Fall (1) Semester
Credits: 3 ECTS
Contact Hours: 36
Self-study: 80
Hours: 108

Main Objectives

The course is focused on developing skills for solving information issues using biomedical high performance computing and synthesis of interdisciplinary knowledge and skills. The course includes basic information about parallel computing algorithms and their application to solve typical issues of data processing in the fields of biology and medicine. The examples and typical tasks are taken from the field of modelling in neurophysiology and medical visualization. The course has a practical orientation and is focused on developing skills of using modern computing systems to meet the challenges of processing large volumes of biomedical data.

Learning Outcomes

As a result of the course, a student must:

  • know: the principles of biomedical information processing using high-performance computing;
  • be able to: apply the necessary areas of the discipline to prepare data and the realization of biomedical information processing algorithms using high-performance computing;
  • master: interdisciplinary knowledge synthesis techniques, skills of information issues preparation in biomedicine in order to solve those with the use of high-performance computing systems.

Professor

Konstantin Brazovskii

Apply

Read more

Course annotation

Computing in Biomedicine (Teaching practice) (3 European Credits)

Taught by: Assoc. Prof. Konstantin Brazovskii

The course is focused on developing skills for solving information issues using biomedical high performance computing and synthesis of interdisciplinary knowledge and skills. The course includes basic information about parallel computing algorithms and their application to solve typical issues of data processing in the fields of biology and medicine. The examples and typical tasks are taken from the field of modelling in neurophysiology and medical visualization. The course has a practical orientation and is focused on developing skills of using modern computing systems to meet the challenges of processing large volumes of biomedical data.

The module covers the following topics:

  • Introduction.
  • Neurophysiology information computing issues. Practical approaches to data preparation.
  • Typical algorithms for solving issues in computational information neurophysiology. Application of parallel algorithms and software systems.
  • Presentation of the results of neural networks numerical simulation.
  • Information processing issues of large and very large images. Preparation of images for parallel processing.
  • Typical image processing algorithms, application of parallel algorithms and software systems.
  • Presentation of solving issues of image processing.
  • Statistical processing of biomedical data of very large volume. Preparing data for supercomputer systems.
  • Typical algorithms of biomedical data statistical processing.
  • Presentation of the results of statistical processing.

Learning objectives

As a result of the course, a student must:

  • know: the principles of biomedical information processing using high-performance computing;
  • be able to: apply the necessary areas of the discipline to prepare data and the realization of biomedical information processing algorithms using high-performance computing;
  • master: interdisciplinary knowledge synthesis techniques, skills of information issues preparation in biomedicine in order to solve those with the use of high-performance computing systems.

Content of the module

The course is focused on developing skills for solving information issues using biomedical The course is focused on developing skills for solving information issues using biomedical high-performance computing and synthesis of interdisciplinary knowledge and skills. The course includes the following topics:

  • Introducing biomedical data processing features in supercomputer systems. Methods of transmission and storage of large and very large volumes of data, requirements for the initial data parallel processing.
  • Typical computational neurophysiology issues. Modelling of individual neurons and neural networks, algorithms for high-performance computing systems. Suboptimal ways to store descriptions of neural structures, the most suitable data structure, NeuroML description language.
  • Study of parallel algorithms for solving issues of this information type. Selecting one of the algorithms for self-study and implementation on supercomputing cluster using NEURON software.
  • Protection of an individual task, discussion of the results.
  • Medical visualization areas that create the largest image size. Three-dimensional images and the methods for their preparation for parallel processing.
  • Processing algorithms of large and very large images. Choice and independent conducting of one of the algorithms on the supercomputer cluster.
  • Protection of an individual task, discussion of the results.
  • Sources of very large volume statistical data, large-scale epidemiological and clinical studies. Modelling sample, data preparation.
  • Algorithms for large and extra-large size statistical data processing. Choice and independent conducting of one of the algorithms on the supercomputer cluster.
  • Prerequisite knowledge, skills and competencies of students are: knowledge programming language, mathematics, algebra, logic, principles of biomedical data processing.

Overview of tasks and lectures

The length of this course in the 1st semester is 6 weeks (a lecture and practical classes of 36 hours in total).

Topic of lecture:

1. Computing in biomedicine. Introduction.

Topics of the practical classes:

The practical class 1

  • Basic problems of computational neurophysiology. Practical approaches to data preparation.
  • Typical algorithms for solving computational tasks in neurophysiology. Application of parallel algorithms and software systems.
  • Presentation of the results of neural networks numerical simulations.

The practical class 2

  • Large and very large image processing problems. Preparation of the images for parallel processing.
  • Typical image processing algorithms, application of the parallel algorithms and software systems.
  • Presentation of the results of solving image processing problems.

The practical class 3

  • Statistical processing of very large biomedical data. Preparing data for the supercomputer systems.
  • Typical algorithms of biomedical data statistical processing.
  • Presentation of the results of  biomedical data statistical processing.

The practical classes involve 3 seminars for students (3 hours each), 3 practical work (19 hours in total), and 3 presentations (2 hours each) with a task of solving typical biomedical data processing issues using a high-performance computing cluster.

Position within the programme

This is a unique course of the programme, which relates to the application of high performance computing in solving biomedical data processing issues. The knowledge and skills are important for the study and application of modern research methods and mathematical modelling in biomedicine. This module provides a framework for the courses High-performance Computing in Biomedicine and Data Analysis in Biomedicine.

Teaching format

Structure

The course is scheduled for the first semester. The course involves 108 hours, including 36 hours in the classroom. The sections of the course are given 6 weeks (a lecture and practical classes of 36 hours). Lecture is given in a multimedia auditorium, equipped with technical means for video conferencing, as well as presentation and interactive equipment. Practical classes are given with the use of remote access technology to high-performance computing cluster resources.

Grading

The form of final assessment is an credit  test. The credit  test consists of an overall grade for the performance quality of practical tasks included in the mandatory minimum, testing of the practical skills and knowledge of the program and an oral exam on theory.