digital imaging in microscopymicroscopy.berkeley.edu/resources/talks/digital imaging.sm.pdf ·...

of 42 /42
Digital Imaging in Microscopy

Author: others

Post on 05-Jul-2020

5 views

Category:

Documents


0 download

Embed Size (px)

TRANSCRIPT

  • Digital Imaging in Microscopy

  • Hand DrawingThe Camera Lucida

  • Film at focal plane of projection lens

    Photo tube

    Projection Lens

    • 4x5" film• 35 mm film

    • Black & White filmTechPan 100

    • Color filmK25, K64

    Film Photography

  • First PassSecond Pass

    ∆T = 1/60 s (fields)

    Full frame image = 1/30s

    Video Imaging — Analog TechnologyDynamic visualization

    SignalCurrent

    Video SignalVoltage

    Target VoltageSupply

    10-100v

    Target Electrode Photoconductive Layer

    Electron beam

    Cathode

    Filament 525 l

    ines

    Interlacing

  • TV Standards

    International Television Standards Committee1950s

    NTSC, PAL, SECAM

    • 525 TV lines• 60 Hz field rate

    Realized resolution = 350 horizontal lines

  • Digital Imaging using Analog Cameras

    rateanalog-to-digital

    NTSC analog video

    Frame grabber

    Resultant digital image is 640x480 pixels

  • Imaging Requirementsfor

    Fluorescence Microscopy

    Specimen

    Excitation filter

    Dichroic mirror

    A B

    Objective

    Barrier filter

    Fluorescentlight fromsample

    Arc light source

  • Epifluorescence microscope with color image capture

    Low-light Image Capture

  • Increasing Light Sensitivity of Video Detectors

    Silicon Intensified Target (SIT) Camera

    Vidicon image intensifier

    NTSC video

  • Charge Coupled Device (CCD) Detectors

    Front-illuminated CCD

  • Solid State image detectors

  • Color Imaging — 1 CCD

    Analog (0.6 MPix) 1995

    1/3” Sony frame transfer CCD

    Bayer color filter

  • Color Imaging — 3 CCDs

    1

    2

    3

    CCDsFiltersDichroics

    Analog (640x480) Technology1.2MPix

  • CCD size

  • Bit Depth and Dynamic Range

  • 8-bit

    12-bit

    Greater Bit-Depth and Dynamic Range Records More Information

  • Quantum Efficiency and Spectral Response

    QE vs. Impinging wavelength

    QE = photons/electrons

  • Backside-illuminated CCD

    Increasing Quantum Efficiency (QE) of CCD

    Etching uniformly thins a CCD to a thickness of approximately 10 µm. The image is focused on the backside of the CCD register where there is no gate structure. Thinned CCDs exhibit high sensitivity to light from the soft x-ray to the near-infrared regions of the spectrum.

  • Coating the CCD can improve the spectral sensitivity of the CCD (top: coated)

    Increasing Spectral Response of CCD

  • Increasing Signal Readout Rate Interline Transfer CCD

    Serial Clocks

    ParallelClocks

    Output Amp

    direction of parallel shift

  • Dark Current and CCD Temperature

    Single-Stage Peltier Cooling-10°C ambient

    Optronics DEI 450

  • Photometrics CH350 digital camera

    Two-Stage Cooling

    Heat from Peltier drawn off by recirculating liquid

    Two-Stage Peltier Cooling-35°C to -50°C

  • Resolution and Sampling

    amplitude quantization

    NN

    Sampling rate (1/T)

    Nyquist sampling

    1/T > 2f

    The rate of sampling must be at least twice the frequency of the resolvable desired structure (signal)

    300 dpi 72 dpi 20 dpi

  • Matching microscope resolution and CCD pixel size

    1. Determine microscope resolution d=0.61λ/NA

    where NA is the average of the objective and condenser

    2. Match the microscope resolution with (a minimum) of two pixels on the CCDM• 0.61λ/NA = 2.0p

    where M=the objective magnification; p=pixel sizeor

    M = 2p/(0.61λ/NA)

    e.g., CCD with 6.8µm pixels and green light (500nm) illumination:M = 2•6.8μ/(0.61•0.5μ/1.3)

    M=58Thus, a 60x, 1.3 N.A. microscope objective provides a diffraction-limited image

    for the many CCDs (e.g., KAF1401E CCD, QImaging).

  • Photometrics QuantixKAF 1401E

    1035x1317 pixels6.8µm x 6.8µm pixels 12 bit pixel depth5MHz readoutTwo-Stage cooling to -50C

    Frame readout CCD

  • Photometrics CoolSnap HQfor Fast Imaging

    1040x1392 pixels12-bit pixel depth6.45µm x 6.45µm pixels 20MHz readoutSingle-stage cooling to -30C

    Interline transfer CCD

  • Sony ICX282 Progressive Scan Interline CCD, Color2560 x 19203.4µm x 3.4µm10-bit20, 10, 5, 2.5MHzPeltier thermoelectric cooling to -10°C below ambient

    QImaging MicroPublisher

    5 MPixBayer filter CCD

  • Digital Images

    (0,0)

    (1024,1024) Blue

    Red

    Green

  • E. herbico -la

    299R (P fruB -gfp

    [AAV])

    -3

    -2

    -1

    0

    1

    2

    3

    0 2 4 6 8 10 12

    cube root [mean pixel intensity]

    t=0 (inoculum)

    t=4 (leaf #2)

    t=25 (leaf #1)t=50 (leaf #1)t=75 (leaf #1)

    50

    25

    105

    10.5

    75

    9095

    9999.5

    Digital Image Processing and Analysis

    Digital Segmentation

    and Image Analysis

  • Some Applications Require 12-16 Bit Dynamic Range

    Sarah Swanson, Jones lab PMB

    ABA

    GA

    DIC 440nm fluor ratio

  • Confocal MicroscopesPoint-Source Detectors (PMT)

    Ch 1 (32 PMTs)

    Ch 2

    Ch 3

    pinholesDichroics

    Sample Plane

    Laser

    TransmittedLight detector

    diffraction grating

  • C o n f o c a l M i c r o s c o p y G e n e r a t e s Vo l u m e t r i c D a t a

  • Software Generates 3D Projections through Volumetric Data

    Green: DAPI-stained ChromosomesRed: 5S rDNA

    5S sequence marks the end of chromosome 5

  • 3D Projection Using Voxel Data

    Ring canals between nurse cells and oocytes

    Red:Rhodamine phalloidinBlue: DAPIGreen: FITC-Ab

  • V o l u m e t r i c D a t a C o l l e c t i o n A l l o w s f o r 3 D I n v e s t i g a t i o n

    leech segmented central nervous

    system

    The nuclei of all the progeny,

    neurons and glial cells, are labeled

  • GFP Expression in Arabidopsis roots

    Arabidopsis mutant from J. Haseloff Lab

    Feldman lab, PMB

  • Deconvolution Microscopy

    PSF

  • 3D Image Processing and Analysis

    Yeast Golgi

  • Pixel Colocalization

    2D histogram

    Binary channel

  • T h e G o l g i A p p a r a t u s

    Localization of Golgi and trafficking molecules

    Sebastien Carreno, Drubin Lab

  • Giardia as Imaged using Deconvolution Microscopy

    Isosurface modeling

    Deconvolved voxel projection

    Red: nucleiGreen: fluorescein anti-tubulin