Knife-Edge Scanning Microscope (KESM 1.5): Optics and Cameras

Knife-Edge Scanning Microscope (KESM 1.5): Optics and Cameras
KESM 1.5 Optics and Cameras
Knife-Edge Scanning Microscope (KESM 1.5):
Optics and Cameras
Bruce H. McCormick
Department of Computer Science
Texas A&M University
College Station, TX, USA 77843-3112
mccormick@cs.tamu.edu
Technical Report: tamu-cs-tr-2006-10-1
-1-
KESM 1.5 Optics and Cameras
Abstract
Knife-Edge Scanning Microscopy (KESM) is being developed for its potential
applications to biology and medicine, and specifically for the scanning and reconstruction
of whole organs at a cellular level of detail. This is the challenge that KESM 1.5, the
extensive modification to the prototype instrument, KESM 1.0, is designed to address.
In the new instrument, KESM 1.5, now under design and construction, tissue is sectioned
and imaged concurrently. The plastic-embedded tissue is mounted under water in a
specimen tank above a three-axis Aerotech stage. Only the stage, and hence specimen
block, move; the diamond knife used for sectioning and the microscope/camera used for
imaging are rigidly mounted to a granite bridge over the stage. The stage provides 20nm
encoding on the two horizontal directions, X & Y, and 25nm in the Z-axis vertical lift
stage. The microscope images the top facet of the diamond knife, and a line-scan camera
scans the sectioned tissue as it flows across the top facet of the knife. Sections are
typically cut at 0.5µm thickness. The lift stage is incremented in height by the section
thickness after each cutting stroke. Because of the encoding accuracy of the stage,
registration between images from serial sections has been excellent in KESM 1.0,
allowing reconstruction of the cellular structure of the embedded tissue, for example,
mouse brain.
This technical report describes the following facets of KESM 1.5 design:
• Objectives, tube lenses, and optical trains
• Line-scan cameras
• Line-scan camera couplers
• Observation camera and its coupler, and
• Microscope/knife mounting
Each section gives the rationale for making the specified design choices.
The design presented here accommodates both bright-field and fluorescence microscopy.
However, the present technical report does not describe the illumination system of the
KESM 1.5 instrument, except in general terms, nor does it model the flow of illumination
from light source to the line-scan camera sensor. A companion technical report, KESM
1.5. Illumination for Bright-field and Fluorescence Microscopy, now in preparation, will
return to this part of a comprehensive image capture system for KESM 1.5.
-2-
KESM 1.5 Optics and Cameras
Table of Contents
1. Introduction
1.1. Rationale for KESM 1.5 optics and cameras
1.2. Versatile design to meet changing uses and technology
1.3. Overview of the instrument
1.4. Topics discussed
2. Objectives, Tube Lenses, and Optical Trains
2.1 Rationale for hybrid optical design of microscope
2.2 Olympus and Zeiss microscope objectives
2.3 Magnification and field of view in specimen plane
2.4 Olympus and Zeiss tube lenses
2.5 Optical trains
3. Line-scan Cameras
3.1 Limits on useable number of pixels
3.2 Cameras categorized by number of pixels
3.3 Cameras categorized by color and sensitivity
3.4 Cameras categorized by data rate
3.5 Output and control
3.6 Initial KESM 1.5 cameras
4. Line-Scan Camera Couplers
4.1 Grouping line-scan cameras by coupler
4.2 Line-scan couplers
5. Observation Camera and its Coupler
5.1 Uses of the observation camera
5.2 Choice of the observation camera
5.3 Couplers for the observation camera
6. Knife/Microscope/Camera Mounting
6.1 Knife/microscope axes define a three-dimensional Euclidean coordinate system
6.2. Optical components and camera mounting
6.2.1. Objectives
6.2.2. Magnification changer (Olympus optics only)
6.2.3. Universal reflected light illuminator with lamphouse and lasers
6.2.4. Dual port with tube lenses (Olympus and Zeiss)
6.2.5. Observation camera and couplers
6.2.6. Line-scan camera couplers
6.2.7. Line-scan cameras and their mounting
6.3 Knife assembly components and adjustments
6.3.1. Diamond knife and knife module
6.3.2. Knife mounting and adjustments
6.3.3. Ribbon extractor and pump
6.4. Specimen tank
6.3.1 Specimen tank
6.3.2. Specimen ring
6.3.3. Extraction of microscope/knife from specimen tank
6.3.4. Constrained movement of tank
Appendix A. Dalsa Line-Scan Cameras
-3-
KESM 1.5 Optics and Cameras
A.1. First-generation KESM cameras
A.2. Piranha P2-series line-scan cameras
A.3. Piranha P3-series line-scan cameras
A.4. Piranha HS-series line-scan cameras
A.5. New Piranha P2 color line-scan
Appendix B. Comparison of Microscopes for Bright-field and Fluorescence Imaging
List of Tables
Table 1. Water-immersion objective parameters
Table 2. Magnifications and fields of view in specimen plane
Table 3. Tube lens parameters
Table 4. KESM 1.5 optical trains
Table 5. Useable number of pixels for KESM 1.5 optics
Table 6. Useable number of pixels for various effective objective magnifications
Table 7. Cameras characterized by choice of stain technology
Table 8. Dalsa line-scan cameras categorized by color and sensitivity
Table 9. Available cameras categorized by data rate: pixel rate/(line/frame) rate
Table10. Initial KESM 1.5 cameras
Table11. Available cameras listed by increasing sensor diameter
Table12. Coupler specifications
Table13. Line-scan camera couplers used in KESM 1.5
Table14. Observation camera parameters
Table15. Couplers for the observation camera
Table16. Dual port tube lenses
Table A.1. CT-series cameras used in KESM 1.0
Table A.2. Piranha2 P2-series cameras
Table A.3. Piranha3 P3-series cameras
Table A.4. Piranha2 HS-series cameras
Table A.5. New Piranha2 color cameras (anticipated December 2006, preliminary data)
Table B.1. Comparative study of four microscope systems
List of Figures
Fig. 1. Objective position and orientation with respect to the diamond knife
Fig. 2. Concurrent sectioning and imaging of specimen block. (a) Mouse brain in
specimen ring; (b) Tissue ribbon rolling across top facet of diamond knife
(graphic visualization)
Fig. 3. Layout of KESM 1.5 optical trains
Fig. 4. Knife/microscope axes define 3D regular Cartesian coordinate system
Fig. 5. Magnification changer
Fig. 6. Universal reflected light illuminator (URLI)
Fig. 7. Modified URLI with dual laser ports
Fig. 8. Mounting of Olympus Super 20X objective to the magnification changer (MC)
and universal reflected light illuminator (URLI)
Fig. 9. Mounting of Zeiss 63X objective to magnification changer (MC) and universal
reflected light illuminator (URLI)
Fig. 10. Dual port (side view)
Fig. 11. Observation camera and coupler (schematic)
Fig. 12. Adjustment of knife orientation
-4-
KESM 1.5 Optics and Cameras
1. Introduction
1.1 Rationale for KESM 1.5 optics and cameras
Knife-Edge Scanning Microscopy (KESM) is being developed for its potential
applications to biology and medicine, and specifically for the scanning and reconstruction
of whole organs at a cellular level of detail. No small-animal organ (e.g., brain,
cardiovascular system, kidneys, or liver, as large as several hundred cubic millimeters in
volume), has yet been scanned at submicron resolution, reconstructed in three
dimensions, and visualized. Though visualized piecemeal through microscope binoculars,
the anatomy of specimens only one-tenth this volume remains largely descriptive.
Quantitative anatomy of tissue on a cellular scale, including visualizing its microstructure
and statistically analyzing its cells and their interconnections, remains yet to be done.
This is the challenge that KESM 1.5, the extensive modification to the prototype
instrument, KESM 1.0, is designed to address.
The reconstruction of the mouse brain, which is the focus of our current research efforts,
remains a significant challenge for light microscopy. Diffraction-limited optics must be
pushed to its limits to resolve fine structure in the brain--its dendritic spines and extensive
fiber tracts of axons. The meshwork of fibers must be resolved to reconstruct the mouse
brain network. We need the finest light microscope objectives and line-scan cameras
available for such high-resolution biomedical imaging.
High resolution comes at a high cost: Every improvement by a factor of two in linear
resolution of the specimen demands an eight-fold increase in scanning time, computation,
and data storage. Therefore we also need a complementary facility to survey organs
targeted for a cellular level of analysis, with less resolution but greater speed. Microscope
objectives and line-scan cameras are needed, of course, for this end of the spectrum as
well.
To meet these diverse and complex needs, we propose a design for the optics and
cameras of the revitalized instrument, KESM 1.5. This design, based on our experience
with the prototype instrument, KESM 1.0, will achieve the goal of scanning whole smallanimal organs, and especially the mouse brain, at a submicron, or cellular level.
1.2 Versatile design to meet changing uses and technology
Biomedical imaging at a cellular level of detail is undergoing dynamic upheaval. First,
the entry of transgenic animals into biomedical research has made prominent the imaging
of fluorescent proteins, such as GFP (green fluorescent protein). Then additional stain
technologies unused even a few years ago, such as quantum dots and heavy- element
stains, have arrived on the biomedical imaging scene. Second, vastly improved waterimmersion microscope objectives have come into play, such as the Super 20X Olympus
objective (0.95 NA) for electrophysiology, and the new Zeiss 63X objective (1.0 NA) for
confocal microscopy, both employed in KESM 1.5. Third, digital imaging technology has
passed through a virtual lifetime since the introduction of our prototype instrument,
KESM 1.0; recall the consumer digital cameras available five years ago. Of course it is
-5-
KESM 1.5 Optics and Cameras
impossible to design KESM 1.5 for the ages. Nonetheless we have proposed a versatile
design, allowing great freedom to meet changing uses and needs.
1.5 Overview of the instrument
In the new instrument, KESM 1.5, now under design and construction, tissue is sectioned
and imaged concurrently. The plastic-embedded tissue is mounted under water in a
specimen tank above a three-axis Aerotech stage. Only the stage, and hence specimen
block, move; the diamond knife used for sectioning and the microscope/camera used for
imaging are rigidly mounted to a granite bridge over the stage. The stage provides 20nm
encoding on the two horizontal directions, X & Y, and 25nm in the Z-axis vertical lift
stage. The microscope images the top facet of the diamond knife, and a line-scan camera
scans the sectioned tissue as it flows across the top facet of the knife. Sections are
typically cut at 0.5µm thickness. The lift stage is incremented in height by the section
thickness after each cutting stroke. Because of the encoding accuracy of the stage,
registration between images from serial sections has been excellent in KESM 1.0,
allowing reconstruction of the cellular structure of the embedded tissue, for example,
mouse brain.
In KESM 1.5, water-immersion objectives having a 35º access angle are used. The axis of
the microscope's objective is inclined 35º from the vertical (Fig.1). Accordingly, one side
of the objective virtually scrapes along the horizontal top surface of the specimen block.1
The microscope images the top facet of the diamond knife, whose normal lies 35º above
the horizontal, or more specifically, above X-axis, the cutting axis of the instrument. Of
the 35 degrees available, 30º are taken up by the angle between the top and bottom facets
of the diamond knife, and the remaining 5º, called the clearance angle, provides
necessary clearance between the bottom facet of the knife and the newly-cut surface.
.
Fig.1. Objective position and orientation with
respect to the diamond knife
1
The access angle of the Super 20X Olympus objective, XLUMPLFL 20XW, is 31°. To achieve the
requisite 35° access angle, one side of the objective is flattened sufficiently to meet this criterion. In the
instrument the objective is oriented such that its flattened side faces the newly-cut top surface of the
specimen block. The objective is flattened by grinding, removing ceramic and metal, not glass.
-6-
KESM 1.5 Optics and Cameras
In the prototype instrument (KESM 1.0), water immersion objectives were chosen
because they alone had a large access angle – the angle between the specimen plane and
the narrowest cone, with apex at the center of the objective’s focus, which encompasses
the objective. Nikon 10X and 40X WI objectives, with access angles of 45°, were used.
Water immersion objectives provide better resolution by virtue of their higher numerical
aperture. KESM 1.5, in search of higher resolution, uses water-immersion objectives with
minimum angles of 35º. Access angle of the objective, not its working distance, is the
critical parameter.
In KESM 1.5 we image a stripe across the tissue ribbon, as the ribbon flows across the
top facet of the knife (Fig. 2). The stripe, aligned along the knife edge and 50µm or less
in width, spans the field of view of the objective. The field of view is the 1.100mm for
the super-20X Olympus objective and 0.317mm for the Zeiss 63X objective. For lowsensitivity line-scan cameras using a simple linear sensor, the stripe width is the size of a
pixel back-projected onto the specimen plane (e.g., 10µm pixel size/20X = 0.5µm for the
Olympus objective). High-sensitivity line-scan cameras use an area sensor, typically 96
TDI registers by k pixels (with k = 2, 4, 6, or 8 x 1024 pixels). Back-projected onto the
specimen plane, the area sensor views a minimum stripe of width approximately 50µm
for the 20X Olympus objective. For the Zeiss objective, the corresponding stripe widths
are 10µm/63X = 0.16µm for a single register line-scan cameras and 16µm for 96 TDI
register high-sensitivity cameras, again for a 10µm pixel size.
Fig 2 (a) Concurrent sectioning and
imaging of specimen block, showing
mouse brain in specimen ring
Fig. 2 (b) Tissue ribbon rolling
across the top facet of knife
(graphic visualization)
In the mouse brain application, sections are cut 15mm long. After each sectioning stroke,
the lift stage of the instrument is translated vertically and the specimen (plastic-embedded
mouse brain) moved upward by the section thickness, typically 0.5µm. During data
acquisition, 70% of the time is spent sectioning/scanning data and the remaining 30%
spent returning the specimen block to its home position. Obtaining maximum data rate is
all-important; for example, imaging a mouse brain at cellular level using the 63X Zeiss
objective generates 26 terabyte of data, even after throwing away image data not
containing mouse tissue. Line-scan cameras vary in their peak output rate, from160MHz
to 640MHz. Scanning times of 100 hr or more are anticipated for many of the scanning
tasks for which the instrument has been designed.
-7-
KESM 1.5 Optics and Cameras
1.4 Topics discussed
This technical report limits its attention to the following facets of KESM 1.5 design:
• Objectives, tube lenses, and optical trains (Sec. 2),
• Line-scan cameras (Sec. 3),
• Line-scan camera couplers (Sec. 4),
• Observation camera and its coupler (Sec. 5), and
• Microscope/knife mounting (Sec. 6).
Each section gives the rationale for making the specified design choices. Appendix A
provides truncated specifications for all applicable Dalsa line-scan cameras.
The design presented here accommodates both brightfield and fluorescence microscopy.
For comparison, Appendix B tabulates the parameters for the optical trains for four
comparable microscopes. However, the present technical report does not describe the
illumination system of the KESM 1.5 instrument, except in general terms, nor does it
model the flow of illumination from light source to the line-scan camera sensor. A
companion technical report, KESM 1.5 Illumination for Bright-field and Fluorescence
Microscopy, now in preparation, will return to this part of a comprehensive image
capture system for KESM 1.5.
2. Objectives, Tube Lenses, and Optical Trains
2.1 Rationale for hybrid optical design of microscope
KESM 1.5 is best thought of as a hybrid microscope merging two co-axial optical trains.
The instrument requires both 20X and 63X water-immersion objectives with access
angles ( ≥ 35°) and high numerical aperture (0.95-1.0 NA). These microscope objectives
cannot be provided by a single manufacturer: whether Olympus, Zeiss, or Nikon.. Thus
one optical train is required to support the Olympus Super 20X objective, its associated
magnification changer, and tube lens; and a second optical train for the Zeiss 63X
objective and its color-correcting tube lens. The two tube lenses, for Olympus (UIS2®
system) and Zeiss objectives (CIS® system) respectively, are not interchangeable.
Otherwise, the two optical trains share parts, for example, a universal reflected light
illuminator (URLI) drawn from the Olympus repertoire of components for the BX2 series
of microscopes.
2.2 Olympus and Zeiss microscope objectives
Table 1 below summarizes the basic parameters for the Olympus and Zeiss objectives.
Here is a glossary and some basic formulas used in the calculation of objective
parameters:2
Specimen plane: object plane of the microscope.
Intermediate image plane: image plane formed by the tube lens of the microscope.
2
This information has been drawn largely from two sources: Nikon’s MicroscopyU web site:
http://www.microscopyu.com and Olympus’s Microscopy Resource Center web site:
http://www.olympusmicro.com.
-8-
KESM 1.5 Optics and Cameras
Magnification (M): specimen length (viewed in the intermediate image plane)/specimen
length (viewed in the specimen plane). Also referred to as objective
magnification, M is marked on the objective housing.
Working distance (WD): distance from specimen plane to front element of objective,
when specimen is in focus.
Numerical aperture (NA): measure of spatial resolution of the objective: NA = nSinα ,
where α is the half-angle extended by the pupil diameter of the objective from
the center of the field-of-view in the specimen plane, and n is the index of
refraction ( n = 1.335 for water-immersion objectives).
Parfocal length: distance from the specimen plane to the rear shoulder of the objective.
Focal length (F): FObjective = FTubeLens / M where FTubeLens is 180mm for the Olympus tube
lens and 130mm for the Zeiss tube lens, and M is the objective magnification. For
example, the focal length F = 9 mm for the 20X Olympus objective, and 2.1mm
for the Zeiss objective.
Pupil diameter: Diameter of the objective pupil is given by the formula:
Pupil diameter = (2 ∗ NA ∗ F ) / n , where F is the focal length of the objective and
n is the refractive index of the immersion media. For the Olympus objective, we
compute Pupil diameter = ( 2 ∗ 0.95 ∗ 9 ) /1.335 = 12.8mm
Field number (FN): The diameter of the field of view measured in the intermediate
image plane. In conventional microscopes, the eyepiece field diaphragm
determines FN.
Field of view in the specimen plane (FoV(specimen)): computed from the formula:
FoV ( specimen ) = FN / M , where FN is the field number and M is the objective
magnification, as follows directly from the definition of the magnification.
Table 1. Water-immersion objective parameters
Optics
Olympus
Objective
XLUMPLFL 20XW
Manufacturer’s part #
Magnification
Type
Working distance (WD)
Numerical aperture (NA)
Access angle (from
horizontal)
Parfocal length
Focal length
1-UB965
20X
Infinity-focus
2.0mm
0.95
31° , flattened on one side to
35° (see footnote 1)
75mm
9mm
Pupil diameter
Field number (FN)
Changer magnifications
Manufacturer’s part #
Field of View (FoV) in
specimen plane
12.8mm
22mm
1X, 1.25X, 1.6X, 2X
U-IT110
1.1mm @ 1X mag. setting
-9-
Carl Zeiss
63X PLAN
APOCHROMAT (VIS-IR)
441470-9900-000
63X
Infinity-focus
2.1mm
1.0
35°
45mm
20mm
1X only
0.317mm
KESM 1.5 Optics and Cameras
2.3 Magnification and field of view in the specimen plane
Table 2 summarizes obtainable magnifications and their corresponding fields of view in
the specimen plane, FoV ( specimen ) , using the Olympus magnification changer where
applicable.3 Effective objective magnification is given by M = objective magnification x
changer magnification. As above, FoV ( specimen ) is the maximum size of object that
can be imaged and is given by FoV ( specimen ) = FN / M , where FN is the field number
and M is the effective objective magnification.
Table 2. Magnifications and fields of view in specimen plane
Qualitative
Effective
FoV
Objective
Magnification
Magnification
(specimen)
Low
20X
1.100mm
Olympus
20X
Medium 1
25X
0.880 mm
Olympus
20X
Medium 2
32X
0.688mm
Olympus
20X
Medium 3
40X
0.550mm
Olympus
20X
High
63X
0.317mm
Zeiss 63X
Changer
Magnification
1X
1.25X
1.6X
2X
Not used
2.4 Olympus and Zeiss tube lenses
Table 3 summarizes the basic parameters for the Olympus and Zeiss tube lenses.
Here is a glossary and the basic formulas used in the calculation of tube lens parameters.
This information has been drawn in part from the two sources cited above:
Tube lens: The lens that focuses the rays emerging from an infinity-focus objective lens
onto the intermediate image plane. Zeiss designs their tube lenses to compensate
for the residual aberrations of the objective lenses; Olympus does not. Between
the objective lens and the tube lens (called “infinity-space”), the rays are parallel.
Mirror units of a universal reflected light illuminator for fluorescence microscopy
can be introduced into that space while minimally disturbing the focus or
aberrations of the microscope.4
Tube lens focal length: distance from conjugate point of tube lens to intermediate image
plane. For thick tube lenses (e.g., as used by Olympus), the conjugate point may
lie outside the glass elements of the lens.
3
Were the magnification changer equipped with a 3X lens, the Olympus Super 20X objective, followed by
this changer setting, would behave as a 60X (0.95 NA) objective, formally not significantly different from
the Zeiss 63X (1.0 NA) objective.
4
Edited from Nikon MicroscopyU.
- 10 -
KESM 1.5 Optics and Cameras
Table 3.Tube lens parameters
Optics
Olympus
Type
U-TLU-1-2 Single port tube, tube lens,
accepts camera adapter, ECO glass.
Tube lens for Olympus BX51/BX61
microscopes (UIS2® optical system)
Manufacturer’s
3-U840EC
part number
Focal length
158mm from the front shoulder of port
tube (holding tube lens) to intermediate
image plane
Field number
22mm
Color correcting
No
∞ -space bounds
50mm-170mm
Olympus/Zeiss
Straight port tube lens
Optics
Olympus tube
U-TLU
port
Zeiss tube port
Zeiss 130mm tube lens in TLU tube
mounting
Carl Zeiss
Tube lens for Zeiss Axio
Imager microscope
(modified CIS® optical
system)
452308
130mm focal length
20mm
Yes
130-166mm (est.)
Side port tube lens
U-TLU
Zeiss 130mm tube lens
in TLU tube mounting
2.5 Optical trains
The design of the Olympus and Zeiss optical trains (Table 4, Fig. 3) satisfies three
constraints: (1) ∞ -space bounds; (2) optical components fit within ∞ -space; and (3) tube
lens units are mounted parfocal to the intermediate image plane. These constraints are:
Infinity-space bounds: Different manufactures impose different bounds on ∞ -space,
as seen in Table 3 above. Outside those bounds, optical quality at the intermediate
image plane can not be assured.
Optical components fit within ∞ -space: All components of the optical train
(Olympus or Zeiss) must fit within the ∞ -space of the appropriate optics. Each
optical train requires two optical components: (1) universal reflected light illuminator
(URLI) and (2) dual port (DP), with dichromatic mirror to feed the observation
camera. The Olympus optical train takes a third component immediately preceding
the URLI: (3) the magnification changer. which can not be used in the Zeiss optical
train as its inclusion would exceed the bound on available ∞ -space.
Tube lens units are parfocal to intermediate image plane: The Olympus and Zeiss
objectives take different tube lenses. Each tube lens fits in a distinct tube lens unit
(TLU). Two TLUs (of common type) fit, in turn, into the two ports of the appropriate,
nearly identical, dual port (DP) (Figure 10, Section 6.2.4). The appropriate dual port
is mounted to the top of the URLI by an Olympus male mount protruding from its
objective-side shoulder. Both optical trains share a common intermediate image
plane. A camera whose sensor is positioned at the intermediate image plane is then in
focus for either choice of optics (Olympus or Zeiss). Having a common intermediate
plane, in itself, imposes no constraint. However, it is highly desirable to rigidly mount
the universal reflected light illuminator (URLI) and the camera mount(s) to a
- 11 -
KESM 1.5 Optics and Cameras
common support. Co-mounting of the URLI and the camera mount was successfully
used in KESM 1.0.
Fig. 3. Layout of KESM 1.5 optical trains
- 12 -
KESM 1.5 Optics and Cameras
Table 4. KESM 1.5 optical trains
Optics
Olympus
Working distance
2mm
Parfocal length of objective 75mm (unique to
(specimen plane to
XLUMPLFL 20XW
objective shoulder)
objective)
168.5mm (to tube lens unit
Objective shoulder to tube
shoulder)
lens ( ∞ -space distance)
Magnification changer,
Optical components
mounted in ∞ -space (from universal reflected light
objective shoulder to TLU) illuminator, and dual port
Tube lens to intermediate
158mm
image plane (TLU shoulder
to image plane)
Length (specimen plane to
401.5mm
intermediate image plane)
Carl Zeiss
2.1mm
45mm
164mm (to tube lens unit
shoulder)
Magnification changer,
universal reflected light
illuminator, and dual port
158mm, by positioning
130mm Zeiss tube lens
within a TLU-like housing
367mm
The design of the optical train in Table 4 (see Fig.3 above) allows virtually no flexibility,
as (1) the upper bound on ∞ -space, and (2) the distance between the objective and tube
lens, are reached for both Olympus and Zeiss optics.
3. Line-Scan Cameras
Our criteria for choice of line-scan cameras depends the desired sampling resolution,
choice of mono/color, light sensitivity, maximum data rate, and choice of computer
interface. These choices are described below.
3.1 Limits on useable number of pixels
The optical resolution imposed by the objective places limits on the useable number of
pixels. The optical resolution of the light microscope is limited by Fraunhofer diffraction.
Traditionally the Rayleigh criterion is used to specify the minimum distance r separating
two points in the specimen plane such that their point spread functions (psf) in the image
plane are distinguishable. Approximating the point spread functions by Airy disks, the
Rayleigh criterion centers the psf for the second point at the first minimum (zero) for the
psf (Airy disk) of the first point: r = 0.61 ∗ λ / NA , where λ is the median wavelength of
the incoherent illumination, nominally 550nm, and NA is the numerical aperture of the
objective. Table 5 shows the resolution limits for the objectives used in KESM 1.5.
The Nyquest sampling theorem, as used in microscopy, asserts that the optical image,
sampled at two pixels per optical resolution displacement, r , can be perfectly
reconstructed. The useable number of sensor pixels required for Nyquest sampling is
given by p = 2 ∗ FoV / r , where FoV is the field of view in the specimen plane. The
minimum number of pixels for Zeiss optics and various effective magnifications of
Olympus optics are shown in Tables 5 and 6.
- 13 -
KESM 1.5 Optics and Cameras
Table 5. Useable number of pixels for KESM 1.5 optics56
Optics
20X Olympus
Resolution limit (r) (Rayeigh criterion, λ = 550nm ) 353nm
Field of view in specimen plane (FoV specimen)
1.100mm
Minimum number of pixels (p) (Nyquest sampling) 6232 pixels
63X Zeiss
336nm
0.317mm
1890 pixels
Table 6. Useable number of pixels for various effective objective magnifications
Qualitative
Effective
FoV (specimen) Useable No. Nearest 2k
Magnification Magnification
of Pixels
multiple
Low
20X
1.100mm
6232
6k
Medium 1
24X
0.917mm
5195
6k/4k
Medium 2
32X
0.688mm
3895
4k
Medium 3
40X
0.550mm
3116
4k
High
63X
0.317mm
1890
2k
3.2 Cameras categorized by number of pixels
The line-scan cameras have linear arrays that come in multiples of 2048 (2k) pixels. We
plan to scan specimens at sensors having four pixel counts: low (2k), medium (4k), high
(6k), and over-sampled high (8k). Table 6 shows that low, medium and high
magnifications use line-scan cameras with 6k, 4k, and 2k respectively for Nyquest
sampling. Survey studies use deliberate under-sampling to shorten scanning time. Except
for doubling the number of sectioning/scanning strokes, imaging the field of view at 20X
with a 2k camera is equivalent to imaging at 10X with a 4k camera. However, the optical
quality of the image is noticeably better in the former case: 0.95 NA (Olympus 20X)
compared to 0.3 NA (Nikon 10X in KESM 1.0), approximately a three-fold
improvement.
3.3. Cameras categorized by color and sensitivity
KESM 1.5 can scan embedded tissue stained as summarized in Table 7. Monochromatic
cameras are preferred when possible, as generally their data rate is at least three times
that of the corresponding color cameras, when the latter are available. Fluorescence
imaging requires the use of high-sensitivity line-scan cameras, i.e., cameras using timedependent integration (TDI) and multiple (up to 96) TDI registers.2
5
For a fuller exposition of optical resolution and digital camera requirements, including an interactive Java
tutorial, see “Digital Camera Resolution Requirements for Optical Microscopy,” Nikon MicroscopyU,
http://www.microscopyu.com/tutorials/java/digitalimaging/pixelcalculator/index.html.
6
The corresponding table for KESM 1.0 optics, using Nikon water-immersion objectives 10X (0.3NA) and
40X(0.8 NA), is shown below:
Optics
10X Nikon 40X Nikon
419nm
Resolution limit (r) (Rayeigh criterion, λ = 550nm ) 1118nm
Field of view in specimen plane (FoV specimen)
2.5mm
0.625mm
Minimum number of pixels (p) (Nyquest sampling)
4472 pixels 2035 pixels
- 14 -
KESM 1.5 Optics and Cameras
Table 7. Cameras characterized by choice of stain technology
Type Low sensitivity
High sensitivity
Mono Monochromatic non-fluorescent stains Single-channel fluorescence (e.g., GFP,
(e.g., Nissl, Golgi-Cox, and heavysimple immunofluorescence)
element stains)
Color Conventional multicolored histological Multi-channel fluorescence, (e.g.,
stains (e.g., as used in cellular-level
multi-FP, quantum dots,
anatomy and medical pathology)
immunofluorescence counter-stains)
The focusing requirements of the KESM 1.5 vary with the type of camera used. Lowsensitivity cameras use a single register of pixels, often one for each color.
Monochromatic low-sensitivity imaging must keep only one sensor register in focus.
Color low-sensitivity imaging, using three co-linear sensor registers, must keep all three
registers in focus. Focusing requirements for high-sensitivity cameras are more
demanding. These cameras use an area sensor with width determined by the sensor
resolution (e.g., 2048 or more pixels) and height determined by the number of TDI
registers used (i.e., up to 96 registers). The entire area sensor must be held in focus by
KESM 1.5, a difficult optical alignment issue. Also the specimen block must be stepped
along the cutting axis at a sampling interval such that successive TDI registers see exactly
the same line in the specimen plane.
We will continue to use Dalsa line-scan cameras in KESM 1.5. These cameras outperform their competitors (principally Basler and Fairchild). Dalsa technical service has
also been superior.7 Table 8 summarizes the Dalsa line-scan cameras appropriate for the
KESM by their color and sensitivity.
Table 8.Dalsa line-scan cameras categorized by color and sensitivity
Camera type Low sensitivity
High sensitivity
Mono
Piranha P2 and P3 Series Piranha HS Series, CT-F38
Color
Piranha P2-color Series
Piranha HS-color Series not available, CT-F7
3.4 Cameras categorized by data rate
Almost all research light microscopes sold today come equipped with a single digital
camera. Why then are multiple cameras employed in KESM 1.5? The answer comes
down to the all-important issue: How long does it take to scan a specimen? Almost all
biomedical imaging at submicron resolution is limited to imaging thin specimens in two
dimensions, or building aligned stacks of a few hundred images in three dimensions. For
these applications, minimizing the time to image the specimen is rarely a significant
concern.
However, scanning of an entire small animal organ (e.g., mouse brain) at submicron
resolution turns the situation is turned on its head. For KESM 1.5, scanning times of 100
8
Both the CT-F7 and the CT-F6 are cameras held over from the prototype instrument, KESM 1.0. Both
cameras are now obsolete and will be replaced when funds become available. .
- 15 -
KESM 1.5 Optics and Cameras
hours will be the norm. Camera selection, after allowing for resolution, color, and
sensitivity requirements, is dictated solely by whether a camera of adequate data rate is
available. To bracket the stain technologies of Table 7, an ideally equipped KESM
laboratory would provide cameras for three resolutions, mono/color imaging, and
low/high sensitivity: in total, twelve possible cameras. Table 9 categorizes available
Dalsa line-scan cameras by type, exhibiting for each its data rate: both pixel rate and
(line/frame) rate. Seven monochromatic cameras are available, while no color cameras
are currently available (although two are anticipated in December 2006). Also positioned
in Table 9 are the two Dalsa line-scan cameras (CT-F3 and CT-F7) carried over from
KESM 1.0.
Table 9. Available cameras categorized by data rate: pixel rate/(line/frame) rate
Resolution Series 2k
4k
6k
8k
Mono
CT-F3
160 MHz/36 kHz
P2
160 MHz/68 kHz
160 MHz/36 kHz 160 MHz/
24 kHz
P3
320 MHz/
33.7 kHz
HS
120 MHz/52 kHz
160 MHz/36 kHz
640 MHz/
68 kHz
Color
CT-F7 3 (color) x
25 MHz/ 10.7 kHz
P2
3 (color), by
3 (color), by
12/2006
12/2006
HS
3.5 Output and control
Piranha-series (P2, P3, and HS) line-scan cameras use the industry-standard
CameraLink® interface. The Camera Link interface comes in three speed ranges: Base,
Medium, and Full. All cameras listed in Table 9 can use the Medium or Full Camera Link
interface, with the latter required exclusively for the 8k, 640 MHz camera. A common
MDR26 Camera Link control is used throughout.
KESM 1.5 uses one Camera Link interface card: the Coreco Imaging X64 Full Frame
Grabber, which supports any Base, Medium, and Full Camera Link camera. Coreco is
now a subsidiary of Dalsa, and Dalsa uses Coreco cards for testing their cameras.
3.6 Initial KESM 1.5 cameras
Three Dalsa line-scan cameras will be used initially (see Table 10).
Table10. Initial KESM 1.5 cameras
Camera No. Model/Series No. Pixels
Mono/color
1
2
3
mono
color
mono
Piranha2-P2
CT-F7
CT-F3
2048 (2k)
2048 (2k)
4096 (4k)
- 16 -
Sensitivity
(low/high)
low
high
high
Coupler
(Table 11)
1 (no lens)
2
4
KESM 1.5 Optics and Cameras
Camera 1, the new Piranha2 P2-2k camera, uses a 10µm pixel size. For Zeiss optics its
sensor just covers the 20mm diameter field of view in the intermediate image plane.
Direct (no lens) optics is used to couple between the tube lens and camera, giving the best
image quality possible. No higher pixel count is meaningful for the 63X objectives, as
determined by the Rayleigh optical resolution criterion and Nyquest sampling (see Table
5 above).
Cameras 2 and 3 are being carried over from KESM 1.0. These cameras, the Dalsa CT-F7
and CT-F3 cameras respectively, are more than 5 years old, a generation ago in the fastmoving digital camera world. Neither camera is currently manufactured by Dalsa. These
cameras were designated at the time of purchase as high-sensitivity cameras (i.e., having
multiple TDI registers). Today these older cameras (when operated at 32-TDI registers,
our normal operating point) have lower responsivity than the new, single-line, Piranha2series cameras (e.g., Camera 1 in Table 10). Both older cameras will be phased out, and
replaced by newer Dalsa cameras drawn from Table 9, when funds become available.
4. Line-Scan Camera Couplers
The minimum number of optical couplers required to match the KESM 1.5 microscope to
all applicable Dalsa Piranha-series line-scan cameras is determined in Sec. 4.1 and 4.2
below (see Table 11). We show that four couplers in total are required (Tables 11 and
12). Coupler one uses 1X (no lens) direct imaging of the intermediate image plane, and
will be used with the Piranha 2k (10µm pixel) camera for high-resolution imaging with
the Zeiss 63X objective and low-resolution imaging with the Olympus Super 20X
objective. Couplers 2 and 4 are used with the CT-F7 and CT-F3 cameras, respectively.
Furthermore, all cameras for a given coupler use a common mount (M42 x 1 for couplers
1 and 2 and M72 x 0.75 for couplers 3 and 4). All cameras use a common computer
interface card (Coreco Imaging X64 Full Frame Grabber, see Sec. 3.5). The line-scan
camera couplers used in KESM 1.5 are summarized in Sec. 4.3.
4.1 Grouping line-scan cameras by coupler
Available Dalsa Piranha-series cameras and our older CT-series cameras used by KESM
1.0 (see Table 9 above) are sorted by increasing sensor diameter in Table 11. Cameras are
then partitioned by sensor diameter into four groups, each group using a common
coupler. As shown in Table 11, all cameras for a given coupler can use a common mount,
either the M42 x 1 mount for smaller-format cameras or the M72 x 0.75 mount for largerformat cameras. Table 12 summarizes the coupler specifications.
Table 11. Available cameras listed by increasing sensor diameter
Coupler Camera No.
Pixel Sensor Lens
Output
Pixels Size Dia.
Mounts (Camera
(µm) (mm)
Link)
1
Piranha2 2k
10
20.48
C, F,
M
P2
M42
2
Piranha2 2k
13
26.62
F, M42 M
HS
- 17 -
Olympus
Resolution9
LR
Zeiss
Resolution3
HR
LR
HR
KESM 1.5 Optics and Cameras
3
4
Piranha2
HS
CT-F7
Color
Piranha2
P2 Color
Piranha2
P2
Piranha2
P2 Color
Piranha2
P2
Piranha3
P3
Piranha2
HS10
CT-F3
4k
7
28.67
F, M42
M
MR
2k
14
28.67
M42
LR
HR
2k
14
28.67
F, M42
Custom
(EPIX)
--
LR
HR
4k
10
40.96
F, M72
M
MR
4k
10
40.96
F, M72
--
MR
6k
7
43.01
F, M72
M
HR
8k/6k
filled
8k/6k
filled
4k
7
43.01
M72
M. F
HR
7
43.01
M72
F
HR
13
53.25
M72
MR
7
57.34
M72
Custom
(EPIX)
M, F
HR+
7
57.34
M72
M, F
HR+
7
57.34
M72
F
HR+
Piranha2 8k
P2
Piranha3 8k
P3
Piranha2 8k
HS11
Table 12. Coupler specifications
Coupler Max. Sensor Lens
Dia. (mm)
Mount
1
20.48
M42
2
28. 67
M42
3
40.96
M72
4
57.34
M72
Olympus
Mag. ideal)12
0.93X
1.30X
1.86X
2.61X
Zeiss Mag.
(ideal)4
1.02X
1.43X
N/A
N/A
Coupler Mag.
(in KESM 1.5)
1X (direct)
1.3X
1.9X
2.6X
4.2 Line-scan camera couplers
Sources of line-scan camera couplers are given in Table 13. All couplers are direct
couplers and project to the intermediate image plane (± a differential length, which is
adjusted by the threading the camera in or out about its threaded mount. Two couplers are
1X (no lens), and serve merely to shield the optical train from external light, an important
consideration for fluorescence microscopy. All couplers are supported from the straightthrough port of the dual port (Olympus or Zeiss, see Section 6.2.4). The cameras are
supported from the camera support (Sec. 6.2.6), which provides differential
displacements in X, Y, and in θ by theta-rotation of the camera within its mount.
- 18 -
KESM 1.5 Optics and Cameras
Table 13. Line-scan camera couplers used in KESM 1.5
Optics
Magnification Source
Olympus
1X (no lens)
Olympus
Zeiss
1.3X
1.9X
2.6X
1X (no lens)
1.3X
Quioptic
Quioptic
From KESM 1.0
Micro Star Tech.
Quioptic (custom)
Part No.
U-TV1X,
U-TMAD
DT13OU
DT19OU
DT13ZZ
5. Observation Camera and its Coupler
5.1 Uses of the observation camera
The KESM 1.5 microscope does not have binoculars or a trinocular. The field of view in
the specimen plane is imaged via a digital observation camera, which under program
control can run in either still or video mode. The observation camera will be used for
coarse focusing, microscope/knife alignment, and ribbon monitoring. Ribbon monitoring
entails examining the top facet of the knife at the end of each stroke for residual
sectioning debris remaining after the knife has been flushed. Automation of knife/optical
alignment and ribbon monitoring require taking digital images of the top facet of the
knife. An area-scan digital camera has been chosen for this purpose.
5.2 Choice of the observation camera
An IEEE-1394/FireWire CMOS digital color camera (PixeLINK PL-A742, 1280 x 1024
pixels) has been selected as the observation camera (Table 14).
Table 14. Observation camera parameters
Imaging Device
CMOS Sensor (2/3” Chip)
Manufacturer and Model
PixeLINK PL-A742 CMOS color camera
Video Output
IEEE-1394/FireWire
Power Requirements
Via IEEE.a cable
Lens Mount
C-Mount
Synchronization
External via trigger
Control
Via downloadable software
Output Image Size Range (H x V) 1280 x 1024
5.3 Couplers for the observation camera
The observation camera images the entire field of view of the microscope. The knife edge
and the newly-cut tissue ribbon, illuminated by a structured light stripe across the top
facet of the knife, are visible on the computer monitor. Unlike conventional video
couplers, which match the chip diagonal to the field number (FN) of the microscope, the
coupler for the observation camera matches the horizontal width of the chip to FN. This
results in couplers of low magnification:
Chip horizontal size/FN (Olympus) = 8.6mm/22mm = 0.39X
- 19 -
KESM 1.5 Optics and Cameras
Chip horizontal size/FN (Zeiss) = 8.6mm/20mm = 0.43X
Slightly less magnification is acceptable. Both couplers (Table 15) use a C-mount to
attach to the camera. Each coupler mounts to its own dual port containing a beamsplitting dichromatic mirror.
Table 15. Couplers for the observation camera
Optics
Magnification
Manufacturer
Olympus
0.38X
Qioptic
Zeiss
0.38X
Qioptic
Part Number(s)
DC38NN13
DC38NN
6. Knife/Microscope/Camera Mounting
In this section we position and orient the knife and microscope/camera relative to the
mounting backplane fastened to the granite bridge behind the stage and parallel to the XZ plane defined by the three-dimensional stage. The basic framework, a threedimensional Cartesian coordinate system defined by the knife/microscope axes when
perfectly aligned, is introduced in Sec. 6.1. Into this framework the components of the
optical train are introduced in stack order, from objective to camera (Sec. 6.2). The knife
assembly components are then introduced in Sec.6.3. This section also explains the tradeoff between knife and the microscope/camera adjustments, which makes it possible to
bring these two assemblies into proper alignment. The specimen tank (Sec. 6.4) sits atop
the 3-axis precision stage. Its structure and allowable displacements are tightly
constrained by the prior optical and knife component mountings.
6.1. Knife/microscope axes define a 3D Cartesian coordinate
system
The microscope/camera scans the specimen ribbon as it rolls across the top facet of the
diamond knife, imaging an illuminated stripe parallel to the knife edge and displaced by
10-20µm from the knife edge (Fig. 2). Upon proper alignment, the specimen plane of the
microscope coincides with the top facet of the knife.
Fig. 4. Knife/microscope axes
define 3D Cartesian coordinate
system (Y’-axis parallel Y-axis)
- 20 -
KESM 1.5 Optics and Cameras
The knife and microscope are placed within a three-dimensional rectangular coordinate
system, whose origin is at the center of the objective’s field of view in the specimen
plane (Fig. 4). The first axis (X’) of the coordinate system is in the specimen plane (top
facet of the knife, assuming perfect alignment) and parallel to Z-X plane of the stage
(which, in turn, is parallel to the back mounting plane defined by the granite bridge of the
instrument). The second axis (Y’) is parallel to the Y-axis of the stage. The third axis
(Z’), the optical axis of the objective, also lies parallel to the Z-X plane of the stage. Axes
Z’ and X’ are rotated 35° counter-clockwise about the Y’ axis: ZÆ Z’ and X Æ X’,
respectively.
6.2. Mounting the components of the optical trains
6.2.1. Objectives
Two objectives, the Super 20X Olympus WI objective and the Zeiss 63X WI objective
are used, with parfocal lengths of 75mm and 45mm, respectively. When an objective is
positioned and oriented correctly, its field of view is centered on the illuminated stripe of
the knife’s top facet, whose normal lies along the objective’s optical axis (Fig. 4).
The Olympus Super 20X objective is mounted to the bottom (objective-side) port of the
magnification changer (MC), as shown below in Fig. 8 below.
The Zeiss 63X objective is mounted to the (objective-side) port of the magnification
changer, to a fitting attached to the 1X setting hole of the unit, as shown below in Fig. 9
below.
6.2.2. Magnification changer
The Olympus magnification changer, U-IT110, provides 1X, 1.25X, 1.6X and 2X
intermediate magnifications to the Olympus Super 20X objective. Fig. 5 is a dimensioned
drawing of the changer. The Olympus objective mounts directly to the bottom (objectiveside) mount of the changer (Fig. 8), while the Zeiss objective mounts to the 1X setting
hole of the unit (Fig. 9).
6.2.3. Universal reflected light illuminator with lamphouse and lasers
The universal reflected light illuminator (URLI) provides epi-illumination for
fluorescence microscopy. A dimensional drawing of the unit is shown in Fig. 6, while
Fig. 7 is a photograph of the custom addition to the unit showing the dual laser ports. The
URLI uses a lamphouse or either of two lasers (not supplied). We will use a conventional
Olympus lamphouse, and for the lasers we will use the Coherent 488mm laser mounted
to the rear laser port (facing toward the back mounting plane) and a ultra-violet 405nm
laser to be mounted to the front laser port.
- 21 -
KESM 1.5 Optics and Cameras
Fig. 5. Magnification Changer
- 22 -
KESM 1.5 Optics and Cameras
Fig. 6. Universal reflected light illuminator (URLI)
Fig. 7. Modified URLI with dual laser ports
- 23 -
KESM 1.5 Optics and Cameras
Because the distance from the optical axis to its rear port considerably exceeds the
corresponding distance in the Nikon URLI used in KESM 1.0, the Olympus URLI is
mounted sideways, with its controls facing up and to the right, while its long axis extends
down and to the left (Figs. 8 and 9). The lamphouse is beyond the limited travel of stage
during sectioning/scanning. The unit, in common with the camera mount support, is
attached to the Y’-axis linear stage by thick rectangular stock (Figs. 8 and 9). Amply
space is left between the back mounting plane and the URLI, when the URLI is mounted
as shown in Figs. 8 and 9. All cantilevered parts in KESM 1.0 can be shorted
accordingly, and the specimen tank positioned more centrally over the Aerotech lift stage.
Limiting this foreshortening are three constraints: (1) clearance for the observation
camera, (2) clearance for the rear-mounted laser (Coherent 488nm), and (3) maintaining a
±25mm displacement for the specimen mount from it center position.
Fig. 8. Mounting of Olympus Super 20X
objective to the magnification changer (MC)
and universal reflected light illuminator
(URLI)
Fig. 9. Mounting of Zeiss 63X objective to
the magnification changer (MC) and
universal reflected light illuminator (URLI)
6.2.4. Dual port with tube lenses (Olympus and Zeiss)
The dual port is mounted atop the URLI via its male Olympus fitting. The dual port
consists of three parts: the beam splitter (BS) and two tube lens units (TLUs) inserted into
its straight and side ports (Fig. 10). The BS contains a dichroic mirror and feeds 33% of
the light to the side port, and hence to the observation camera via its coupler. Two dual
ports are used in KESM 1.5: one each for Olympus and Zeiss optics, respectively. They
are distinguished only by the TLUs inserted into the two ports: each port holds a tube
lens unit (TLU) for the appropriate optics (Table 16). The TLUs are designed to be
parfocal with the intermediate image plane: that is, the tube lens in mounted in the TLU
such that the distance from the shoulder of the TLU to the intermediate image plane is
independent whether an Olympus or Zeiss tube lens is embedded in the TLU.
Critical to the design of the dual port is one overriding consideration: to minimize its
usage of ∞-space -- to get this distance down to 38mm or less. Only then can the 170mm
upper bound on ∞-space for Olympus optics be met; the ∞-space constraint for Zeiss
optics is equally severe.
- 24 -
KESM 1.5 Optics and Cameras
Fig. 10. Dual port (side view)
Table 16. Dual port tube lenses
Olympus/Zeiss Optics
Straight port tube lens
Olympus tube port
U-TLU
Zeiss tube port
Zeiss 130mm tube lens in
TLU tube mounting
Side port tube lens
U-TLU
Zeiss 130mm tube lens in
TLU tube mounting
6.2.5. Observation camera and couplers
The observation camera coupler extends from the side port of the dual port (Olympus or
Zeiss). The observation camera couplers are optics-dependent: different couplers are used
for Olympus and Zeiss optics (Table 14, Sec. 5.3), though physically they are almost
indistinguishable. The double doublet optics is displaced slightly between the two units to
insure that their image planes are parfocal with the camera sensor chip.
- 25 -
KESM 1.5 Optics and Cameras
The coupler to the observation camera can not intrude below the shoulder of the dual
port, and into space occupied by the URLI. The use of a coupler, stripped of its outer
housing, and only 32mm in diameter makes this possible.
A PixeLINK PL-A742-R color camera has been selected as the observation camera (Sec.
5.2). This camera uses a right-angle configuration to shorten its outreach from the side
port of the dual port (Fig. 11).
Fig.11. Observation camera and coupler
6.2.6. Line-scan camera couplers
The line-scan camera couplers (Table 13) mount to the straight-through port of the dual
port. As described in Section 4.2, these couplers do not directly support the cameras, as
the camera support allows differential movements in the X, Y, and θ to accommodate
microscope/knife misalignment.
6.2.7. Line-scan cameras and their mounting
The camera mounting in KESM 1.0 has worked well; it will be little changed in KESM
1.5, except in four minor regards: (1) The camera mount support, in common with the
URLI, will be attached to the manually-positioned Y’-axis linear stage (Figure 4); (2) the
camera mount support can be shortened (less cantilevering) in view of the sideways
mounting of the URLI; (3) finer threads will be used for differential X, Y, and θ
displacements (Fig. 12), and (4) a M42 x 1 camera mount for small format cameras
(coupler classes 1 and 2, Table 11) will be provided, in addition to the M72 x 0.75
- 26 -
KESM 1.5 Optics and Cameras
camera mount for larger format cameras (coupler classes 3 and 4, Table 11) The latter
was used in KESM 1.0. An adapter ring (72 x 0.75 to M42 x 1) can be used to down-size
the M72-mount to the M42-mount.
6.3. Knife assembly components and adjustments
6.3.1. Diamond knife and knife module
The diamond knife and knife module has undergone a number of transformations in
KESM 1.0. The major change in KESM 1.5 is reducing the facet angle of the knife to
30°, leaving a clearance angle of 5° between the bottom facet of the knife and the newlycut surface of the block (Fig. 1). Three additional, if less consequential, changes are: (1)
providing a more substantial mounting of the knife module to the knife adjustment
assembly; (2) enhancing ribbon extraction by moving the water (and ribbon) orifice
closer to the top facet of the knife; and (3) welding the knife to its mount using a
precision fixture that insures that the top facet of the knife will be at 35° above the
horizontal when installed in the instrument (see Sec. 6.3.2 below).
6.3.2. Knife mounting and adjustments
The microscope defines a rectangular coordinate system by its optical axis, its specimen
plane, and a line parallel to the Y-axis of the stage (Fig. 4). The microscope, and hence its
coordinate system is given only one degree of freedom: a focusing adjustment, which
moves the components of the optical train and attached camera along its optical axis.
The knife must be positioned and oriented relative to this microscope-defined coordinate
system. Perfect alignment would require that the knife edge (1) lie in the specimen plane,
(2) centered in the field of view, and (3) oriented such that its image is aligned with the
linear sensor of the camera. Rigid motion of the knife to attain perfect alignment would
requires, like for any rigid body, six degrees of freedom (dof). Knife adjustments,
however, must be minimized and kept extremely rigid to minimize the potential for
chatter during sectioning. Therefore, of the six degrees of freedom, the knife is assigned
only three and the remaining three degrees of freedom are compensated by differential
translation and rotation of the camera. Conceptualizing the knife as a rowboat, the knife
is assigned one translational dof (focusing-like displacement along the X’ axis, see Fig.
4), and two orientation degrees of freedom: roll and yaw (Fig. 12). Pitch is adjusted at the
time of manufacture of the knife module: the diamond knife is placed in a custom jig, and
welded to the knife module. In summary, the three degrees of freedom not provided the
knife are compensated by incorporating three dof’s in the adjustments provided by the
camera mounting.
6.3.3. Ribbon extractor and pump
The problems with the prototype ribbon extractor are well-known. These stem from
multiple sources: (1) Extraction suction upon the newly-cut ribbon is inadequate, leading
to ribbon fold-over, which in turn gives rise to occasional vertical blobs in the scanned
image; (2) Occasional debris ca be left attached to the top facet of the knife at the end of
the sectioning stroke. This debris will be removed by an end-of-stroke flush, and
validated as clean by an automated inspection using the observation camera.
- 27 -
KESM 1.5 Optics and Cameras
Fig. 12. Adjustment of knife orientation
The pump assembly was upgraded during KESM 1.0 to a larger size. Mounting the pump
to the frame of the instrument does not appear to introduce significant vibration; no
difference in degree of chatter was observed when cutting with the pump turned off.
6.4. Specimen tank
6.4.1 Specimen tank
The specimen tank, mounted atop the three axis stage, carries the specimen under water.
The specimen tank with its spill tray has worked well. The principal improvements in
KESM 1.5 are: (1) Lighten the specimen tank (including water fill), as both the specimen
tray and the Y-axis stage rest atop the lift stage, and nearly exceed its load limit; (2)
Move the specimen tank back toward the rear mounting plate so as to reduce the extent of
cantilevering required.
6.4.2. Specimen ring
The specimen is mounted to a specimen ring (Fig.2a), which is keyed to the bottom of the
specimen tank. The specimen ring, introduced in KESM 1.0, has worked well. Its home
position is with the microscope/knife centered over the specimen block, prior to
sectioning, and the lift stage at its lowest position. The lift stage raises as serial sections
are taken.
Presently mouse brains are embedded apart from the specimen ring, using a minimum of
plastic embedding compound, and then in a second stage, mounted atop the plastic-filled
specimen ring. As work expands with rat brain, which are four times larger in volume
- 28 -
KESM 1.5 Optics and Cameras
than the mouse brain, it may prove necessary to introduce a larger specimen ring, but that
decision would be premature today.
6.4.3. Extraction of knife/microscope from specimen tank
First, the specimen tank is returned to its home position prior to this movement. The knife
is then extracted from the specimen tank as follows: the manual knife stage is unlocked
and retracted along the X’ axis (Fig. 4). Likewise, the microscope manual stage is
unlocked and retracted along the Y’-axis (Fig.4).
6.4.4. Constrained movement of tank
The sidewise mounting of the URLI and its associated lamphouse constrains free
movement of the specimen tank. However, as seen in Figs. 8 and 9, there is adequate
room for the 15mm-20mm travel required for sectioning.
Appendix A. Dalsa Line-Scan Cameras
Truncated data sheet information on select Dalsa line-scan cameras is tabulated below.
Data sheets for currently available cameras, including CAD mechanical drawings and
mounting details, can be downloaded from the Dalsa website: http://www.dalsa.com.
The prototype instrument, KESM 1.0, uses two high-sensitivity line-scan cameras: CT-F7
(color, 2k) and CT-F3 (mono, 4k); for details see Sec. A.1. The Piranha2 series of linescan cameras (both the P2-series, single-register cameras and the high sensitivity HSseries, multiple TDI register cameras) are described in Sections A.2 and A.3 respectively.
In particular, the 2048-pixel Prianha2 camera, P2 40-2k, used in KESM 1.5, is described
there. New Pyranha2 color line-scan cameras, anticipated December 2006 but not yet
announced by Dalsa, are described briefly in Sec. A.4. Finally, Sec. A.5 describes the
recently announced Piranha P3-series line-scan cameras. These latter cameras, though not
directly applicable to KESM 1.5, give insight into the probable technological direction of
future Dalsa line-scan cameras, and show the impact of such manufacturing applications
as the testing of large-format LCD panels.
A.1. First-generation KESM cameras
The Dalsa high-sensitivity line-scan cameras, (CT-F7, (2k, color) and CT-F3 (4k, mono)
are used with the prototype KESM 1.0.. Neither camera is currently manufactured by
Dalsa. Table A.1 describes their imaging characteristics.
Table A.1. CT-series cameras used in KESM 1.0
Resolution
2048 x 64 TDI Color
Camera Model
CT-F7
Data Rate
3 (color) x 25 MHz
Max. Line/Frame Rate 10.7 kHz
Pixel Size
14µm
Data Format
8 bit
Output
Custom
Lens Mount
M72 x 0.75
- 29 -
4096 x 96 TDI
CT-F3
4 x 40 MHz
36 kHz
13 µm
8 bit
Custom
M72 x 0.75
KESM 1.5 Optics and Cameras
Responsivity
Sensor Aperture
Control
DN/(nJ/cm2)
180 DN/(nJ/cm2)/ 60
DN/(nJ/cm2)/ for 32 TDI
53.3 x 1.3mm
EPIX card/software
28.7 x 1.6mm
EPIX card/software
A.2. Piranha2 P2-series line-scan cameras
Table A.2. Piranha2 P2-series cameras
Resolution
2048
Data Rate
4 x 40 MHz
Max. Line/Frame Rate 68 kHz
Pixel Size
10µm
Data Format
8, 10 bit
Output
Medium Camera
Link
Lens Mount
C, F, M42 x 1
mount
Responsivity
76 DN/(nJ/cm2) @
10dB
Control
MDR26 Camera
Link
Cost (USD)
$4651
4096
4 x 40 MHz
36 kHz
10µm
8, 10 bit
Medium Camera
Link
F, M72 x 0.75
mount
76 DN/(nJ/cm2) @
10dB
MDR26 Camera
Link
$5481
6144
4 x 40 MHz
24 kHz
7µm
8, 10 bit
Medium Camera
Link
F, M72 x 0.75
mount
38 DN/(nJ/cm2) @
10dB
MDR26 Camera
Link
$5880
A.3. Piranha3 P3-series line-scan cameras
Table A.3. Piranha3 P3-series cameras
Resolution
8192
Data Rate
8 x 40 MHz
Max. Line/Frame Rate 33.7 kHz
Pixel Size
7µm
Data Format
8,12 bit
Output
Med, Full Camera
Link
Lens Mount
M72 x 0.75
Responsivity
44 DN/(nJ/cm2)
Control
MDR26 Camera Link
Cost (USD)
$6185
12288
8 x 40 MHz
23.5 kHz
5 µm
8, 12 bit
Med, Full Camera
Link
M72 x 0.75
27 DN/(nJ/cm2)
MDR26 Camera Link
A.4. Piranha2 HS-series line-scan cameras
Table A.4. Piranha2 HS-series cameras
Resolution
2048 x 64 TDI
Data rate
2 x 60/4 x 30 MHz
Max. Line/Frame
52 kHz
Rate
Pixel Size
13µm
4096 x 96 TDI
4 x 40 MHz
36 kHz
8192 x 96 TDI
320/640 MHz
34/68 kHz
7 µm
7µm
- 30 -
KESM 1.5 Optics and Cameras
Data Format
Output
Lens Mount
Responsivity
Control
Cost (USD)
8,10 bit
Base, Medium
Camera Link
M42 x1, F mount
1610 DN/(nJ/cm2)
MDR26 Camera
Link
8,12 bit
Base, Medium
Camera Link
M42 x 1, F mount
1170 DN/(nJ/cm2)
MDR26 Camera
Link
$5942
8,10 bit
Medium, Full
Camera Link
M72 x 0.75
1170 DN/(nJ/cm2)
MDR26 Camera
Link
$10,880/$14,048
A.5. New Piranha2 color line-scan cameras
Table A.5. New Piranha2 color cameras (anticipated December 2006, preliminary data)
Resolution
2048
4096
Pixel Size
14µm
10µm
Appendix B. Comparison of Microscopes for Brightfield
and Fluorescence Imaging
Table B.1 gives a comparative study of four microscope systems: (1) the Olympus BX51WI+DPMC, (2) the Olympus BX51+ BX-RFA, (3) the Zeiss Axio Imager, and (4) the
Nikon Eclipse PhysioStation E600FN. The later microscope was used in KESM 1.0. The
data in the table was derived from the four corresponding schematics (Fig. 17).
Table B.1. Comparative study of four microscope systems
Dimensions (mm)
Olympus
Olympus
Zeiss Axio
BX51+WI BX51+BX
Imager
+DPMC
-RFA
Objective
Olympus
Zeiss 63X
Super 20X
Parfocal length
75
45
45
Specimen plane to mid153.9
126
plane URLI
e∞-space to trinocular
134
124
128
Trinocular height
90.8
62.5
>86.7
URLI height (L)
87.5
87.5
URLI width (max) (W)
88
88
URLI length: optical axis 261
261
327.6
to rear port (L1)
- 31 -
Nikon
Eclipse
E600FN
Nikon 10X
& 40X
60
187.5
155
82.7
55
235
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising