fastFMT

Fluorescence molecular tomography (FMT) is an emergent technique that enables to localize selective fluorescent markers within biological tissues. FMT represents a fundamental tool for medical research by addressing the molecular bases of diseases. The numerous potentialities of this technique are applicable to the diagnosis and follow-up of cancers. In particular, FMT makes it possible to perform longitudinal studies in vivo in mice.

FMT is traditionally carried out by scanning a laser spot onto the animal's surface and collecting the emitted fluorescence light by means of optical fibers or a CCD camera. By applying a reconstruction algorithm, the 3D distribution of the markers inside the animal can be recovered.

To speed up both acquisition and reconstruction and/or improve the reconstruction quality, we have proposed a novel pattern-based approach that makes the most of the available information while limiting the number of necessary measurements. This approach is non contact, easy to handle, and low cost. The sketch of the setup is provided below.

FIG 1: Digital micromirror device (DMD) allows the projection of any light pattern onto the object's surface. The fluorescence light emitted by the markers is collected by a CCD camera. The object is rotated in order to acquire fluorescence images from different views.

FIG 2/3: Reconstruction of three fluorescence inclusions (Φ = 2 mm) drilled into a cylindrical phantom (h = 45 mm, Φ = 20 mm). Left: phantom, right: reconstruction (see Ducros et al. 2011 for details). The reconstruction time is less than 2 minutes on our workstation.

A dedicated reconstruction algorithm has been developed. It is based on the following basics:
compression of the fluorescence images by means of a wavelet transform
resolution of the diffusion equation by means of the finite elements TOAST package
building of a compressed forward model
regularization of the inverse problem

Our approach, which has first been demonstrated on a slab and on a cylinder, lies on a very general framework. It can deal with any geometry of the object, any acquisition configuration (reflectance, transmittance, etc), and any light pattern. Some reconstruction results are provided below. We are applying this method to small animal imaging.

References

Animal model with wavelet patterns, multi-view
Work in progress!

Cylinder with wavelet patterns, multi-view
N. Ducros, C. D'Andrea, A. Bassi, G. Valentini, S.R. Arridge (2012) Virtual source pattern method for fluorescence tomography with structured light. Phys. Med. Biol. 57: 12. 3811-3832 May. DOI

N. Ducros, A. Bassi, G. Valentini, M. Schweiger, S.R. Arridge, C. D'Andrea (2011) Multiple-view fluorescence optical tomography reconstruction using compression of experimental data. Opt. Lett. 36: 8. 1377-1379. DOI

Slab with wavelet patterns

N. Ducros, C. D'andrea, G. Valentini, T. Rudge, S.R. Arridge, A. Bassi (2010) Full-wavelet approach for fluorescence diffuse optical tomography with structured illumination. Opt. Lett. 35: 21. 3676-3678. DOI

Slab with Fourier patterns
C. D'Andrea, N. Ducros, A. Bassi, S.R. Arridge, G. Valentini (2010) Fast 3D optical reconstruction in turbid media using spatially modulated light. Biomed. Opt. Express 1: 2. 471-481. DOI

A. Bassi, C. D'Andrea, G. Valentini, R. Cubeddu, and S. Arridge (2009) Detection of inhomogeneities in diffusive media using spatially modulated light. Opt. Lett. 34: 2156-2158 . DOI

Time resolved propagation of Fourier patterns
A. Bassi, C. D'Andrea, G. Valentini, R. Cubeddu, S.R. Arridge (2008) Temporal propagation of spatial information in turbid media. Opt. Lett. 33: 23. 2836-2838. DOI

Presentations

  • OSA Biomedical Optics (Miami, 2012)
  • IEEE ISBI (Barcelona, 2012)
  • SPIE Photonic West BiOS (San Francisco, 2011)
  • SPIE European conference on Biomedical Optics (Munich, 2009)
  • SPIE Photonic West BiOS (San Francisco, 2009)