En poursuivant votre navigation, vous acceptez l'utilisation de cookies destinés à améliorer la performance de ce site et à vous proposer des services et contenus personnalisés.


Biomedical applications





Optical Diagnosis

In spite of the huge efforts made to improve the optical diagnostic of many lesions by various innovative techniques (such as fluorescence, OCT, Raman, confocal and nonlinear microscopies, to name a few) there is still a real need in this field.
Polarimetric imaging has been comparatively less investigated, at least on real tissues (both ex vivo and in vivo).
There is thus a strong opportunity for scientifically interesting research with potentially high societal impact.
The interpretation of polarimetric images of real tissues is challenging, as tissues are obviously highly complex systems which cannot be “exhaustively” interpreted with a simple optical model.
There is thus room for several approaches, including matrix decompositions as well as Monte Carlo simulations.
We realized such simulations by a state-of the-art code developed by Tatiana Novikova [M.R. Antonelli et al, Opt. Express 10201 (2010)], who supervised Maria Rosaria Antonelli’s thesis, defended in 2011


Preliminary studies, performed on a few samples, suggested that cervical tissue was purely depolarizing, with lower depolarization in (pre-) cancerous regions.
We thus built the ANR-RNTS “Polarimétrie” project (December 2005 – june 2009) around two essential components:
  1. A clinical trial, in vivo, of a simple OSC (Orthogonal State Contrast) imager, sufficient to characterize pure depolarizers.
  2. Full Mueller imaging of surgical samples including tissues other than cervix.
With 145 patients enrolled, the study invalidated our initial hypothesis. The polarimetric response of these tissues is much more complex than that of a pure depolarizer. [A. Nazac, JGOB 42, 464-472 (2013)];
This result was confirmed by full Mueller imaging of ex vivo samples (See figure below).
The healthy malpighian epithelium is characterized by a strong retardation (up to 60°) which disappears in both the dysplastic and the glandular tissues.
However, the three types of tissue can be distinguished by decreasing levels of depolarization from healthy malpighian to glandular epithelium.
All these contrasts are almost impossible to visualize in ordinary intensity.
Figure. Left part: usual non-polarized reflectance image of two surgical samples of uterine cervix. The top sample is healthy, while the bottom sample exhibits a healthy region, at the bottom left of the figure, a severe dysplasia (CIN3), at the top right, and a benign lesion (visible glandular tissue). The diagnosis of the pathologist is reported on the lines materializing the cuts where this diagnosis is available, and shown by the color code at the left of the figure. Middle and right panels : retardation and depolarization images (adapted from Pierangelo et al. Opt. Express 21, 14120-30 (2013))

In Sumary that polarimetric imaging has the potential to:

  • To distinguish dysplastic from healthy tissues for uterine cervix.
  • To identify malignant polyps in colon and provide a first, quick evaluation of the degree of invasion of the tumor, a possibility which would be extremely helpful in colposcopy to optimize the subsequent treatment [A. Pierangelo et al. Opt. Express 19, 1582-1593 (2011)].
  • To detect possible residual cancer after radiochemotherapy.



Future Projects and Perspectives

These results have been considered sufficiently promising to justify a funding by INCa, of our project PAIRGYNECO (2012-2015)
The goal of the PAIRGYNECO project is to boost our activity on uterine cancer management. The project includes three hospital partners:

The work to be done at the CHU de Bicêtre will be focussed on the detection of dysplasias (on a large number of samples > 200)
The work ar the IGR will be focussed on the detection of residual cancers after radiochemotherapy of locally advanced tumors. This is a crucial issue for the patient survival in good conditions, and no currently available imaging technique is really satisfactory for this purpose.