The Claude Research Group Hyperspectral Microscope based on Imaging Fourier transform Spectrometry Dr. Li Jianping, Claude jp.li@siat.ac.cn June 2017 Outline Spectral Imaging Fourier transform Imaging Spectroscopy A general-purpose FTIS instrument 1
Spectral Imaging (Imaging Spectroscopy) 2D-Imaging 1D-Spectroscopy Acquisition time Time Resolution Why Spectral Imaging? 2
Applications of Spectral Imaging Astronomy Remote sensing Meteorology Oceanography Mining geology Mineralogy Agriculture Surveillance Applications in Medicine, Biology and Chemistry 3
Modalities of Spectral Imaging Imaging Microscopy Photography Telescopy Spectroscopy Reflection Absorption Transmission Scattering Raman Fluorescence Electroluminescence Chemoluminescence Bioluminescence Methods for Spectral Imaging 4
Fourier Transform Spectrometer (FTS) 0 0 I ( x) I ( x) I( x) B( )cos2 xd 2 exp 2 B I x j x dx Fourier Transform Imaging Spectrometer (FTIS) I ( x, ) B ( ) cos(2 x cos ') d d p c p 2 p, exp 2 B I x j xdx 5
Throughput (Jacquinot) Advantage FTSs can have more than 60 times higher energygathering capability than grating spectrometers for the same resolving power and similar instrument size. 6
Multiplex (Fellgett) Advantage FTS simultaneously observes all the spectral information from the entire range of a given spectrum during a scan period. Mixed FT Wavenumber (Connes) Advantage The wavenumber scale of an FTIR is derived from a He-Ne laser fringe that acts as internal references for sampling positions in each scan. Therefore, the wavenumber calibration of FTS is much more accurate and has much better long-term stability than the calibration of dispersive instruments. 7
High and Variable Spectral Resolution Instrumental lineshape: 1 1/2 L(cm ) Wide Spectral Range Nyquist limit wavelength min 2 x System spectral transmission response Detector spectral response Optics spectral characteristics 8
A UV-Vis-NIR FTIS Setup of the UV-Vis-NIR FTIS based on the beam-folding position-tracking technique Jianping. Li, Robert. Chan, and Xuzhu. Wang, Tests of a practical visible-nir imaging Fourier transform spectrometer for biological and chemical fluorescence emission measurements. Opt. Express, 2009. 17(23). 9
Prototyping and testing (Gen1) Testing Summary Mode Determinants/ Typical value Spectral Resolution Spectral Range Image Resolution Spatial Resolution Temporal Determinant Typical value Max. OPD Max.~10cm -1 Tunable Δδx + Camera QE ~360-900nm Camera speed + PZT speed Max. 300x300pixels Hyperspectralimaging Diffractionlimited Camera specs 500nm 50x, NA=0.55 Camera speed Image resolution Spectral resolution SNR requirement 20s 200x200pixels, 512frames Only-imaging Determinant Camera QE Camera Pixel # Typical value 200-1100nm 1004x1002pixels (Cascade 1k, Photometrics) Diffractionlimited Camera specs 500nm 50x, NA=0.55 Camera Speed 10fps @ full frame Jianping. Li, Robert. Chan, and Xuzhu. Wang, Tests of a practical visible-nir imaging Fourier transform spectrometer for biological and chemical fluorescence emission measurements. Opt. Express, 2009. 17(23). 10
Instrumental Reconfigurability Transmission Reflection EPI-Fluorescence TIR-Fluorescence Prototype Gen2 11
Absorption Beer s law: I( ) A( ) log[ T( )] log[ ] I ( ) 0 Reflection Mixed SCCBs with red, green and blue color and common glass beads JP. Li and Robert Chan. Towards a UV-Vis-NIR Hyperspectral Imaging Platform for Fast Multiplex Reflection Spectroscopy, Opt. Lett., 2010. 35(20): p. 3330-3332. 12
Reflection: SCCBs 502nm, 550nm and 630nm 512 frames, 100 100pixels, binning =8. total acquisition time is only about 10s. SCCB decoding speed of about 100beads/s. JP. Li and Robert Chan. Towards a UV-Vis-NIR Hyperspectral Imaging Platform for Fast Multiplex Reflection Spectroscopy, Opt. Lett., 2010. 35(20): p. 3330-3332. Fluorescence 50x, 100x100 pixels, binning=8, 5ms exposure time, 1k frames, total acquisition time ~40s Dye-loaded fluorescent polystyrene nano-beads (φ200nm) Li, J., R.K.Y. Chan, and X. Wang, Tests of a practical visible-nir imaging Fourier transform spectrometer for biological and chemical fluorescence emission measurements. Opt. Express, 2009. 17(23). 13
TIRF Jianping Li and Robert Chan, Two-Mode Total Internal Reflection Fluorescence Hyperspectral Microscopy, Focus on Microscopy, March, 2010, Shanghai. Towards a general-purpose Hyperspectral Microscopy System (Gen3) 14
Design Considerations 1 Integration with standard microscopes 2 Equip with faster and larger-chipped camera 3 Equip with GPU based parallel FFT processing 4 User-friendly software GUI development 5 Versatile SI platform for analytical sciences The real HSM system Hardware Software Jianping Li and Yi Xiao, GPU accelerated parallel FFT processing for Fourier transform hyperspectral imaging, Applied Optics (IF: 1.784), 54(13): p. D91-D98, 2015. 15
Performance and specs Optical Physical Spectral resolution (max):~0.4nm Spatial resolution (max):~0.5m Spectral range:~360-1000nm Pixel number: 2560x2160 Instrument size: 30 50 28mm (L W H) Objective lenses: 5x,10x,20x,50x,80x BDF-RT-PL 10x, 20x, 40x Pha Host computer: Intel i7-3820cpu NVIDIA GeForce GTX650 GPU 64GB RAM 2TB HDD Applications Bright-field (transmission) Dark-field (scattering) Bright-field (phase-contrast) Electroluminescence Bright-field (reflection) 16
Technological Comparison Features Excellent spatial & spectral resolving High measurement throughput Operational convenience High & adjustable spectral resolution and wide spectral range High & adjustable spatial multiplex and resolution No spatial resolution degradation High optical throughput Rapid image acquisition Parallel FFT processing Multi-modality imaging Flexible hardware adaptability WYSIWYG preview function Good accuracy 17
Acknowledgement Dr. Robert Chan, Dr. XZ Wang Dr. JX Fu, Mr. R Li Mr. ZN Xu @HKBU Dr. XW Zhao and Prof. ZZ Gu @SEU Prof. JF Wang @CUHK Prof. B Ren @XMU 18