Ania Warczyk, Alia Durrani, Shivani Shah, Dan Maxwell, Timothy Chen.

Optical Imaging

Technique used in neuroscience for detection of brain activity

Uses changes in deflection of incident light to infer hemodynamic activity

Problem Statement

Design a small wireless camera for optical imaging of the cortex which allows free movement of animal being tested

Primary Objectives

Make a design with these criteria: Scalable to fit on the head of a monkey

Small and lightweight Wireless potential High resolution and well depth Providing direct, even lighting

Performance Criteria

Desired resolution: 512 x 512 Desired frame rate: 300 fps Well depth: 12 bits Must run continuously: 5 minutes Must not impede movement of animal:

~300 grams Maximum wireless frame rate: 10 fps Maximum cable frame rate: 30 fps USB is the only way to get 300 fps Eventually, wireless frame rate: 100 fps

Not before 3 years

Solution Descriptions

Current method: Large Camera

Design 1: PillCam Design 2: Lensless Setup Design 3: Beam Splitter Setup

Current Method: Large Camera

PillCam: Hypothesis

PillCam design proves a small self-contained wireless camera can be constructed

2 cm

PillCam: Synthesis

Diagram of our design based on PillCam concepts

PillCam: Performance

Failure with this approach, therefore must try new design Illumination is uneven and inconsistent No adjustable focus

Fixed lens to chip distance (S2) Microfabrication with expensive custom parts

Proprietary information

To mediate these obstacles: Need microcontrollers for lens and chip $$$$$$

Lensless Setup: Hypothesis

Can putting lens in contact with membrane on cortical surface eliminate the need for optics?

Lensless Setup: Synthesis

(Done with different illumination techniques)Slide with thin slices of pig liver

Solid piece of liver tissue imaged through glass cover slip

Lensless Setup: Performance

Liver slide with transmitted light

Liver slide with reflected light

Liver tissue with reflected light

Liver tissue with transmitted light

Lensless Setup: Resolution Test

Image of 1mm grid taken without a lens Image of 1mm grid taken with a lens

Lensless Setup: Performance

Failure of this approach, therefore must try new design Low resolution Illumination issues

Transmitted light does not work for bulk tissue Reflected light requires moving CCD chip

away from tissue surface

To mediate these obstacles: Can implant fiber optic to illuminate from

within

Beam Splitter Setup: Hypothesis

Beam splitter can provide direct illumination with conventional optical techniques in an onboard approach

Beam Splitter Setup: Performance Metrics

Provides direct, even, controlled illumination Single source eliminates light pools

Compact design Parallel to surface of brain

High resolution due to use of lens Lens and chip can be adjusted

individually Put lens and chip on threads

Beam Splitter Setup: Synthesis

Beam Splitter Setup: Performance

Data acquisition trial 1 expected March 24th

Will use grid to determine spatial resolution

Beam Splitter Setup: Calculations Thin Lens Equation:

1/S1 + 1/S2 = 1/f Let R = S1 + S2 There are two solutions to this equation:

S1 = R/2 + sqrt(R^2-4*R*f)/2 S2 = R/2 - sqrt(R^2-4*R*f)/2

The second solution is simply the reverse of the first: S2 = R/2 + sqrt(R^2-4*R*f)/2 S1 = R/2 - sqrt(R^2-4*R*f)/2

Magnification: M = -S2/S1 To map well size onto CCD, set minimum chip

width: w = 2*r w/2r = M = -S2/S1

Conclusions

Design 1 (PillCam) failed due to illumination, focus issues, and high comparative cost

Design 2 (Lensless Setup) failed due to low resolution and problems with illumination

Design 3 (Beam Splitter Setup) resolves illumination, focus, resolution, and cost issues Can fulfill requirements for size and weight

Future Work

Validate design by image acquisition Get smaller lighter parts to

miniaturize design and make more lightweight

Insert 10/90 beam splitter Add x-y-z positioners for lens and

chip Add housing to exclude ambient light