near real time deconvolution for 3d fluorescence
play

Near Real-Time Deconvolution for 3D Fluorescence Microscopy Marc - PowerPoint PPT Presentation

Near Real-Time Deconvolution for 3D Fluorescence Microscopy Marc Bruce, PhD CEO, Microvolution GPU Accelerated Deconvolution Software After 3-D Deconvolution Other Deconvolution 16 Seconds 10.7 Min Thin Filaments Are Missing Preserves Thin


  1. Near Real-Time Deconvolution for 3D Fluorescence Microscopy Marc Bruce, PhD CEO, Microvolution

  2. GPU Accelerated Deconvolution Software After 3-D Deconvolution Other Deconvolution 16 Seconds 10.7 Min Thin Filaments Are Missing Preserves Thin Filaments: 16 sec Microvolution offers faster and more accurate deconvolution Image Courtesy of Molecular Devices

  3. Basics of f Fluorescence Microscopy Sample Objective

  4. Basics of f Fluorescence Microscopy Dichroic Illumination mirror

  5. Basics of f Fluorescence Microscopy Emission

  6. Basics of f Fluorescence Microscopy • Damage to sample from high intensity light • Haze from out-of-plane fluorescence • Blurriness due to diffraction reduces resolution of small objects Derived from Wikipedia, “Light sheet fluorescence microscopy”

  7. Diffraction Limits 3D Resolution Y Z X XY

  8. Point Spread Function

  9. The Solution: Deconvolution Software Z XY http://www.leica-microsystems.com/science-lab/deconvolution

  10. Richardson Lucy Algorithm Iterative algorithm operating under the assumptions of Poisson- distributed noise

  11. Real-time Deconvolution is Now Possible 1000 900 • GPU yields speeds up to 200X 900 Old methods faster than old methods 800 GPU accelerated • Allows for real-time 700 571 600 deconvolution Seconds 500 • Optimize image capture 400 300 215 200 145 100 7.6 4.8 0.42 1.5 0 10 MB 190 MB 890 MB 1246 MB 253x210x50 1080x1080x50 2160x2160x50 2160x2160x70

  12. Higher Signal-to to-Noise Ratio Yields Better Data • Analysis of raw FRET images showed low contrast • 4x effective increase in contrast post-deconvolution • Expected cell features can now be seen Miskolci, V., Hodgson, L. & Cox, D. Using Fluorescence Resonance Energy Transfer-Based Biosensors to Probe Rho GTPase Activation During Phagocytosis. Methods in Molecular Biology 1519, 125-43 (2017)

  13. Diffraction-Limited Resolution Limit

  14. In Increased Resolution to 180 nm 180 140 160 120 140 100 120 80 100 80 60 60 40 40 20 20 0 0 0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3 Tong Zhang and Puifai Santisakultarm, Salk Institute

  15. Microscopy in Dim Light Be Before Aft fter 3D D De Deconvolution 2048 x x 2048 x x 51 8.5 .5 sec econds Image Courtesy of Technical Instruments

  16. Widefield Microscopy Wid idefie ield Im Image Be Before Aft fter 3D D De Deconvolution 1024 x x 1024 x x 34 1.6 .6 sec econds Dennis Hughes, MD, PhD, Assoc Prof of Pediatrics, Jared Mortus, Lead Researcher, Laurence Cooper, MD, PhD, Professor of Pediatrics, George McNamara, PhD, Senior Research Scientist, The Children’s Cancer Hospital, MD Anderson Cancer Center, Houston, TX

  17. Light Sheet Microscopy Derived from Wikipedia, “Light sheet fluorescence microscopy”

  18. Light Sheet Microscopy Light Shee Ligh eet Im Image Be Before Aft fter 3D D De Deconvolution 946 x x 768 x x 321 22.5 .5 sec econds Dong-Yuan Chen in the Bilder lab, UC Berkeley

  19. Lattice Light Sheet Microscopy LLS LLSM Be Before Aft fter 3D D De Deconvolution 1167 x x 512 x x 401 10.2 .2 sec econds Dr. Boyd Butler using 3i Lattice LightSheet

  20. In Instant SIM IM In Instant SIM IM Im Image Be Before Aft fter 3D D De Deconvolution 1446 x x 1131 x x 14 4.9 .9 sec econds VisiTech

  21. FINCH (Fresnel in incoherent corr rrelation holography) Wid idefie ield Im Image Aft fter Propagation and 3D D De Deconvolu lution 2048 x x 2048 x x 21 10 sec econ onds Siegel, N., Lupashin, V., Storrie, B. & Brooker, G. High-magnification super-resolution FINCH microscopy using birefringent crystal lens interferometers. Nature Photonics 10, 802-808 (2016)

  22. Usable on All Tiers of f GPUs • HPC clusters • Split large multi-channel, multi-timepoint images onto independent nodes • Let peer GPUs collaborate • Desktops • Run alongside acquisition • Embedded systems • Currently developing for TX1

  23. Usable on All Tiers of f GPUs 35 30 25 TX1 Seconds 20 K2200 15 970M 10 K40 5 P5000 0 GP100

  24. Questions? Marc Bruce marc@microvolution.com

Recommend


More recommend