Basic principles of quantification using optical techniques

Transcription

1 Contents Basic principles of quantification using optical techniques Adrian Taruttis Helmholtz Zentrum München Chair for Biological Imaging Technische Universität München Light/ tissue interactions Planar optical imaging Molecular Multispectral Optoacoustic Imaging depth 0 0.1mm 1mm 1cm 10cm Imaging with light aka Optical Imaging Contrast? Multispectral Non-ionizing Plate reader Planar TIRF Confocal Multi-photon Economical Resolution MICROSCOPY 0.1 m 1 m 100 m 1mm Absorption of light in tissue Beer-Lambert Law absorption coefficient µ a Source: Weissleder, Ntziachristos, Nature Medicine, 2003 WMIC "Quantification using OI" 1

2 Scattering of light in tissue Scattering Scattering coefficient: [cm -1 ] Mean free path: (mean distance between scattering events) Anisotropy: g Highly scattering media isotropic (Diffusion) approximation: Reduced scattering coefficient: 1 Tissue Reduced scattering coefficient (NIR) Muscle 9 cm -1 Brain 16 cm -1 Breast 12 cm -1 Lung 30 cm -1 Scattering complicates reconstruction X-ray CT reconstruction works because X-rays travel in straight lines (insignificant scattering) Contrast: This does not apply to optical imaging because scattering in tissue is usually very high Cy5.5 Niedre et al., PNAS 2008 Fluorophore-tagged selective targeting Activatable fluorescent agents (dequenching) Fluorescent proteins Imaging aim: determine the distribution of fluorophores in tissue Planar (photographic) Imaging Quantification? Quantification is hindered by: Absorption and scattering of light below the surface: surface weighted Heterogeneous absorption properties Excitation Ntziachristos V. et al., JBO 2005 WMIC "Quantification using OI" 2

3 Normalization in planar fluorescence detection modes Themelis et al., JBO 2009 Detection modes Molecular (FMT) illumination input output Simulations: μ s =10cm 1, μ a =0.3cm 1, slab thickness 1cm Objects at a depth of 3mm from either side, distance 6mm FMT: Normalized measurement data FMT: Image reconstruction Multiple projections from all sides (360 ) Discretized forward-model NIR laser CCD camera Measurement data Matrix formulation : = Matrix inversion ill-posed problem Excitation Normalized data = /Excitation Output: fluorophore distribution in volume WMIC "Quantification using OI" 3

4 FMT: Forward model (1) Diffusion approximation 2 Diffusion Equations illumination Excitation light propagation: D x r x r ax r x r Sx r FMT: Forward model (2) Green s functions Provides solutions to diffusion equations for point source illumination Emission light propagation: excitation fluorophore Dm r m r am r m r x r n r FMT: Deep tissue results Murine model of myocardial infarction MNP CLIO-Cy5.5 in 2007 Hybrid FMT-X-ray CT 3D volume XCT slice Seamless coregistration FMT reconstruction infarct control FMT Coronal slices through heart Sosnovik et al, Circulation 2007 Quantitative fluorophore distribution in 3D Improve FMT performance by structural prior information ability to assign different optical properties to CT regions Hybrid FMT-CT FMT Instrumentation FMT-XCT: Osteogenesis Imperfecta NIR Lasers: 680 nm and 750 nm XCT Instrumentation Schulz et al., IEEE TMI, 2010 Ale, A. et al., Nature Methods 2012 WMIC "Quantification using OI" 4

5 resolution 1mm Resolution degrades with depth fluorescence molecular tomography Pulsed Light Optoacoustic Imaging Ultrasound detector 100μm r 10μm 1µm confocal microscopy 10µm multiphoton microscopy 100µm 1mm 1cm penetration depth Absorber Tissue Nd:YAG + OPO Imaging based on photoacoustic effect optical contrast ultrasound resolution deep inside tissue pressure light fluence optical absorption coefficient Multispectral Optoacoustic (MSOT) oxyhemoglobin deoxyhemoglobin ICG (in plasma) mouse subcutaneous tumor 4T1 breast cancer model MSOT Tumor Imaging ICG i.v. 800nm Hemoglobin day 6 day 13 ICG multispectral 1.0 ICG MSOT channel [a.u.] 0.1 Specific imaging of: intrinsic tissue absorbers: hemoglobin oxygenation (functional) optical agents: fluorochromes and nanoparticles Herzog, Taruttis, Beziere, Lutich, Razansky, Ntziachristos, Radiology 2012 Summary IBMI Planar Contrast Quantitative accuracy Spatial resolution Temporal resolution no undefined Real-time Vasilis Ntziachristos Daniel Razansky... Molecular yes ~1mm ~20 min Multispectral Optoacoustic Absorption yes ~150µm Real-time WMIC "Quantification using OI" 5