Transcript Optical Imaging Instrument for BioBED - CenSSIS

Confocal Reflectance Theta Line-Scanner for
Intra-operative Imaging
Peter J. Dwyer*, Charles A. DiMarzio*, and Milind Rajadhyaksha *s
*The Center for Subsurface Sensing and Imaging Systems (CenSSIS),
Northeastern University, Boston, MA
sDermatology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY
Abstract
A confocal reflectance theta line-scanner is being developed for
imaging nuclear, cellular and architectural detail in human tissues
in vivo. Preliminary design and experimental results show lateral
resolution of 1-5 mm and sectioning (axial resolution) of 2-5 mm
is possible within living human skin; thus, nuclear and cellular
detail has been imaged in epidermis in vivo. Experimental results
are being verified with an analytic model and optical design
analysis. An immediate clinical application will be intraoperative imaging of basal cell cancers to guide Mohs
micrographic surgery for the precise excision of these cancers.
Small
Aperture
BEAM
SPLITTER
Plane in Focus
OPTICAL
SYSTEM
TISSUE
Legend:
Small Aperture
In-focus light
Out-of-focus
light
DETECTOR
Figure 1. – Confocal Microscopy,
“Optical Sectioning”
ILLUMINATION
PATH
Introduction
Table 1. - Instrument Specifications
Line Scanner Specifications
Imaging: 5-20 frames/sec (real-time)
Lateral Resolution: 1-3 mm
Vertical Section Thickness: 2-5 mm
Max Depth: 300 mm
Objective: 10x, NA = 0.80
Field of view: 1.0 mm
Laser illumination: 830 nm (Diode)
Detector Board
Objective Lens
Slit Mount
Laser
Diode
Figure 2. – Confocal RTLS Microscope
linear
polarizer
DETECTION/
IMAGING PATH
Fold Mirror
Scanning Mirror
analyzer
diode laser
Results
• Design and Development;
- Axial line spread function measurements show that the
sectioning is 1.6 – 9.8 mm for detection slits of width 5-100
um under nominal conditions (Figure 7).
- Lateral resolution of 0.8 – 2 mm have been measured with 25,
50, 100 mm and no slit widths (Figure 8). Note: Lateral
resolution = 5 mm at the detector.
- Preliminary images of epidermis appear comparable to those
with point scanning confocal microscopy.
detection slit
width 7.5 mm
scanner
u
• Confocal microscope performs “optical sectioning:”
- A thin section within tissue is imaged non-invasively,
- mm-level three-dimensional resolution and high contrast is
achieved (Figure 1).
• Line scanning is simpler and may offer resolution, contrast and
depth of imaging performance comparable to point scanning.
• Basal cell cancers (BCCs) are among the fastest growing skin
cancers (>1.2 million new cases every year in the USA alone);
- Precise microsurgical excision is performed by a procedure
known an Mohs micrographic surgery.
- Mohs surgery is guided by frozen histology of multiple
excisions, and is thus inefficient and slow.
• The confocal theta line-scanner may enable
- Imaging of BCCs directly on patients in real-time,
- Guidance of Mohs surgery, thus reducing the number of
excisions and avoiding frozen histology.
State of the Art
• Present: Point scanning confocal microscopes are complex,
expensive and difficult to manufacture.
• In-progress Research: Theta line-scanner may be simple, robust,
inexpensive and easy to manufacture.
• Present Standard of Care: Surgical removal of skin cancers with
unknown margins and potentially more scarring; surgery is
guided by frozen histology performed parallel: processing
requires 45 mins per excision, 2-20 excisions per patient.
• Proposed: Confocal theta line-scanner for real-time,
intraoperative imaging and surgical guidance.
r
qmax
OBJECTIVE
LENS
qmin
CONFOCAL
SECTION
THICKNESS
Figure 3. – Confocal Theta Line-Scanning.
The intersection of the illumination and
detection paths creates confocal probe volume
(PSF), (a). Figure 3b illustrates the system
layout.
DIVIDER
STRIP
2h
FOCAL (OBJECT)
PLANE
focussing lens
f = 125 mm
cylindrical lens
f = 200 mm
objective lens
f = 17.5 mm
quarter-wave
plate
linear detector
pixel size 7.5 mm 1024
pixels
divider strip (divides
objective lens aperture
into two halves)
object plane
wdet
Figure 3b. – Optical Layout
Figure 3a. – Geometrical Optics Concept
Figure 4. – Line scanner confocal images in
human skin in vivo using a 50 mm slit. The
upper layers of the epidermis contain the
stratum corneum (a) and the granular cells (b).
100 um
100 um
Figure 4b – Granular Cells
Figure 4a – Stratum Corneum
Figure 8. – (a) Lateral LSF Measurement; 50 mm slit. (b) Lateral
line spread function measurements for various detection slit
widths.
Figure 5. - Line scanner confocal images in
human skin in vivo using a 50 mm slit. The
middle layers of the epidermis containing the
spinous cells. The dark nuclei are surrounded
by the bright, grainy cytoplasm.
100 um
Conclusions/Future Work
100 um
Optical Design
• Confocal theta microscopy was developed by Koester (1980),
Stelzer (1995) and Webb (1999).
• We designed and built a theta line-scanning confocal microscope
(Figure 2) with specifications as shown in Table 1.
• Optimized design based on diffraction for best lateral and axial
(section) resolution.
• The optical design and system set-up are shown in Figure 3.
• In vivo and excised skin specimens were imaged (Figure 4 - 6).
• Measurements of the axial and lateral resolution were made.
Figure 7. – (a) Axial LSF measurement; 10 mm Slit. (b) Axial line
spread function measurements for various detection slit widths.
Figure 6. - Line scanner confocal images in
human skin in vivo using a 50 mm slit. The
deeper layers of the epidermis contain basal
cells. The bright basal cells are clustered
around dermal papillae (dark circles).
100 um
100 um
Acknowledgement: This work is supported in parts by NIH/NCI (Award
Number 1R43CA93106 and 2R44CA93106) and the NSF Partnerships in
Education and Research Program (Award Number EEC-012931).
• Resolution (sectioning) of line-scanner compares well to that
provided by current point-scanning technology.
• Experimental results of axial and lateral line spread function
measurements agree well with theoretical values.
• Further work needed;
- Aberrated LSF at pupil edges and within skin
- Contrast and SNR
- Illumination scan / detection de-scan mismatch.
- Signal drop-out (illum./detect. path mismatch)
- CMOS Detector Sensitivity
- Correlation of images to histology