Doctoral course - Polarimetric techniques combined with machine learning algorithms for biomedical diagnosis of tissue

This course begins with a brief introduction to the theory of polarized light and Mueller polarimetry, followed by practical examples of using this optical technique for a number of biomedical applications, including early diagnosis and surgery guidance. A specific focus will be on using machine learning algorithms for polarimetric data post-processing to quantify the pathological conditions of biological tissue.

Event information

Time

-

Location

Linnanmaa

Add event to calendar

Lecturer

Tatiana Novikova, PhD, Dr Habil (Ecole polytechnique, IP Paris, Palaiseau, France)

Schedule

  • 6 hours lectures
  • 6 office hours
  • 2 hours exams

Monday 11.3 in TS133 at 9:00-17:00

Tuesday 12.3 in TS136 at 9:00-17:00

Wednesday 13.3 in TS287 at 10:30-11:00

Assessment

Attendance and exams (Pass/Fail)

Course Abstract

The studies of polarized light interaction with biological tissue may provide valuable information about the tissue pathological status. The basic advantage of polarimetric techniques consists in being relatively low-cost, fast and non-destructive, thus allowing the measurements even for in-situ applications. Recent advances of imaging Mueller polarimetry demonstrated its potential for early and accurate medical diagnosis of various diseases. Having access to the complete set of polarimetric data, namely, multi-spectral Mueller matrices proves to be the key issue for the accurate characterization of tissue samples. Post-processing of tissue polarimetric images combined with the machine learning algorithms can provide the quantitative diagnostic metric and complement the standard medical practice of tissue diagnosis.

Learning Objectives

  • Introduction to the theoretical framework for the description of fully polarized and partially polarized light
  • Getting familiar with the design, calibration and operation of imaging Mueller polarimeters in either reflection or transmission mode
  • Understanding of non-linear Mueller matrix decomposition algorithms
  • Practical examples of biomedical applications of imaging Mueller polarimetry
    • optical biopsy of bulk tissue (colon, uterine cervix, brain)
    • digital histology of tissue thin sections with transmission Mueller microscopy
  • Machine learning algorithms for diagnostic segmentation of polarimetric images

Schedule

Day 1

9:00 – 10:30 Theoretical Stokes-Mueller formalism for the description of fully polarized and partially polarized light

10:30 – 10:45 Break

10:45 – 12:15 Design, calibration and operation of imaging Mueller polarimeters. Mueller matrix decomposition algorithms

12:15 – 13:15 Lunch

13:15 – 14:45 Biomedical applications of imaging Mueller polarimetry - optical biopsy of bulk tissue and digital histology of tissue thin sections

14:45 – 15:00 Break

15:00 – 16:30 Partial Mueller polarimetry. Machine learning algorithms for diagnostic segmentation of polarimetric images

Day 2

9:00 – 10:30 Office hours

10:30 – 10:45 Break

10:45 – 12:15 Office hours

12:15 – 13:15 Lunch

13:15 – 14:45 Office hours

14:45 – 15:00 Break

15:00 – 16:30 Office hours

Day 3

10:30 – 11:00 Exams

About the Instructor

Dr. Habil. Tatiana Novikova, Fellow of SPIE and Optica, leads the Characterisation and Modeling Division at the Laboratory of Physics of Interfaces and Thin Films, CNRS, Ecole polytechnique, IP Paris, France. She is a Courtesy Professor of the Department of Biomedical Engineering, Florida International University, Miami, USA. Dr Novikova published more than 100 peer-reviewed articles on optical polarization, Mueller polarimetry, biomedical imaging, polarimetric instrumentation, optical metrology and numerical modeling of electromagnetic wave interaction with structured and random media. Dr Novikova is an Editorial Board member of SPIE Journal of Biomedical Optics. She is a recipient of 2020 SPIE G. G. Stokes Award in optical polarization and 2023 European Optical Society Prize Winner.

Last updated: 1.3.2024