Experience

Here is the description of my past working experience:

April 2022 - November 2023: Postdoctoral researcher in computational biology, University of Montpellier, Montpellier, France

  • Objectives:
    • better understand how a single cell (fertilized egg) can develop into a complex biological organism
    • simulate embryo morphology, mainly shape, based on available embryo data and historical description of cells development
  • Results:
    • represented embryo cells shape as parameters of a spherical weighted Voronoi diagram developed earlier
    • ported from MATLAB to Python an optimization module, which extracts diagram parameters from embryo images, to execute the computation freely on the Jean Zay supercomputer
    • added peanut-shape cells processing and improved initialization of optimization parameters to construct the spherical weighted Voronoi diagram more accurately
    • described the evolution of Voronoi diagram parameters during the development process and for different embryos as multidimensional trajectories in space, processed them with space reduction techniques, and compared results
    • created a neural network using Transtab software to predict cells parameters and division probability for the next step based on the recent development history of the cell and its neighbors, the results analysis reveals the necessity of further investigation and system improvements
    • compared the shape of cells with iterative closest point and field level set approaches and their modifications, the results indicated that cells similarity notion has to be determined in the context of the considered problem
  • skills:
    • Computational biology, Morphology, Voronoi diagrams, Image processing, Machine learning, Graph databases
  • technologies:
    • Python, PyCharm, NumPy, SciPy, Scikit-image, Matplotlib, Transtab, Neo4j, Cypher

May 2017 - September 2021: PhD student, Research engineer, INRIA, University of Strasbourg, Strasbourg, France

  • Objectives:
    • estimate liver’s boundary conditions to improve the simulation accuracy in an augmented reality system created for the surgery
    • simulate a realistic liver behavior as well as an influence of the surrounding organs and tissues in real time
    • compute statistical atlas for liver attachments
  • Results:
    • constructed a hyperelastic finite element method with Saint Venant-Kirchhoff material for liver deformations and a cubic spring system for boundary conditions
    • developed a component describing springs with polynomial stress strain relation, which was integrated into SOFA framework
    • employed an approach based on a reduced-order unscented Kalman filter as the best choice for real-time performance/accuracy trade off
    • participated in a software module implementation that allows estimating various model parameters using nonlinear Kalman filters
    • carried out an optimization based on a preconditioner to improve the performance of the estimation process (computation time was 26% less)
    • created an approach based on deformable registration of liver shapes (large deformation diffeomorphic metric mapping) to obtain statistics for ligament positions
    • performed experiments based on synthetic and real data to show the improvement in accuracy for simulations with estimated boundary conditions (average error was 66% less for synthetic data and 23% less for real data)
  • skills:
    • Augmented reality, Numerical simulation, Finite element method, Mass-spring system, Kalman filters, Diffeomorphic metric mapping
  • technologies:
    • OpenCV, PCL, C++, Python, SOFA, FEBio, Deformetrica, QtCreator

May 2014 - November 2016: Software and research engineer, Scientific-Production Affiliate Energiya, Factory Energiya, Saint Petersburg, Russia

  • Objectives:
    • create modules to find various signals in the radio waves spectrum, maintain a database to store the elements detected in records, collect statistics about characteristics of signals with an analyzing tool
  • Results:
    • employed a software module that analyzes noise-like signal appearance and collects their basic information
    • ported from MATLAB to C++ a module that searches for frequency-hopping spread spectrum signals and extracts their basic parameters
    • developed a graphical user interface to show the spectrum frequency/sweep and indicate signals location
    • carried out a prototype for data clusterization based on mixture models
  • skills:
    • Digital signal processing, Fourier analysis
  • technologies:
    • C++, MATLAB, Qt, Intel IPP, Visual Studio

August 2010 - February 2014: Software engineer, OOO Bodynova, Saint Petersburg, Russia

  • Objectives:
    • create a server-client software complex to model a patient body shape and results of breast implant and expander surgery procedures
    • simulate the biomechanical properties of skin, adipose, muscles, bones, and implants realistically
  • Results:
    • developed software that simulates implant placement under soft tissues, estimated its accuracy and computation time
    • created a graphical user interface for the simulation software that allows selecting an implant from a catalogue and choosing its placement parameters
    • constructed a multilayer, nonlinear, and anisotropic mass-spring system to describe soft tissues deformation
    • proposed and employed a modification of a voxel model to accelerate the processing of collision between implants and soft tissues
  • skills:
    • Virtual reality, Numerical simulation, Mass-spring system, 3D computer graphics
  • technologies:
    • C++, WxWidgets, JavaScript, Wolfram Mathematica, Visual Studio

July 2008 - July 2010: Intern, Associate software engineer, EMC corporation, Saint Petersburg, Russia

  • Objectives:
    • participate in a new storage system development
  • Results:
    • contributed to the code that writes data directly to a disk, without using random access memory
    • contributed to the module responsible for cache data saving in case of emergency shut off
    • configured hardware of four storage systems for remote access connections
  • skills:
    • Storage device drivers, Linux development
  • technologies:
    • C, Perl

Back