published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Mathematics / Doctoral / Doctor of Philosophy

The design, construction and details of operation of an instrument for measuring Bidirectional Scattering Distribution Functions (BSDF) {i.e., Bidirectional Reflection Distribution Functions (BRDF) and Bidirectional Transmission Distribution Functions (BTDF)} at four wavelengths from the visible to the far infrared (FIR) are presented. Hardware, software, theory, calibration, accuracy and performance are all addressed. The four major interacting subsystems--optical, electronic, mechanical and computer--are discussed in various degrees of detail. A BASIC language control program for running the system via the Z-80 microprocessor is included in the appendix. The origin of BRDF is traced, and a brief theoretical summary examines it from both a physical optics and a radiometric point of view. Modeling is performed to determine the effect of large radiometer collecting apertures on accuracy. Dynamic range and resolution are determined from experimental and theoretical considerations. Finally, new measurements of Martin Black at 118.8 (mu)m are presented along with measurements at 0.6328, 3.39 and 10.6 (mu)m. The FIR BRDF measurements covered 6 orders of magnitude and values as low as 6 x 10('-4) sr('-1) have been recorded.

External nuclear medicine diagnostic imaging of early primary and metastatic lung cancer tumors is difficult due to the poor sensitivity and resolution of existing gamma cameras. Nonimaging counting detectors used for internal tumor detection give ambiguous results because distant background variations are difficult to discriminate from neighboring tumor sites. This suggests that an internal imaging nuclear medicine probe, particularly an esophageal probe, may be advantageously used to detect small tumors because of the ability to discriminate against background variations and the capability to get close to sites neighboring the esophagus. The design, theory of operation, preliminary bench tests, characterization of noise behavior and optimization of such an imaging probe is the central theme of this work. The central concept lies in the representation of the aperture shell by a sequence of binary digits. This, coupled with the mode of operation which is data encoding within an axial slice of space, leads to the fundamental imaging equation in which the coding operation is conveniently described by a circulant matrix operator. The coding/decoding process is a classic coded-aperture problem, and various estimators to achieve decoding are discussed. Some estimators require a priori information about the object (or object class) being imaged; the only unbiased estimator that does not impose this requirement is the simple inverse-matrix operator. The effects of noise on the estimate (or reconstruction) is discussed for general noise models and various codes/decoding operators. The choice of an optimal aperture for detector count times of clinical relevance is examined using a statistical class-separability formalism.

In the past decade, charge-coupled devices (CCDs) have rapidly become the astronomical imaging detector of choice for the visible and near-IR spectral regions. There are, however, several problems which have greatly reduced the availability of sufficient quality CCDs to the astronomical community. These include the low blue and ultraviolet quantum efficiency of thick devices, the lack of properly thinned devices, warped imaging surfaces, interference fringing, and the small size of the detectors themselves compared to telescope focal planes. This dissertation presents methods which can be used to optimize CCDs obtained from various manufacturers for astronomical observations. A new thinning technique which produces an optically flat surface across an entire CCD is demonstrated. A mounting technique which maintains a flat and stable imaging surface for thinned devices by bonding the CCD backside against a transparent glass support substrate is also demonstrated. Bump bonding of CCDs onto a silicon support before thinning is discussed as a future mounting/thinning technique. The design of antireflection coatings for the near-UV through near-IR spectral regions is explained and demonstrated on silicon diodes, allowing quantum efficiencies as high as 90% to be obtained. The reduction of interference fringing amplitudes by as much as 70% in the red and near-IR with AR coatings is also discussed. And finally, the design of CCD focal plane mosaics using the optimization techniques presented is discussed.

Thermoacoustic emissions induced by 2450 MHz microwave pulses in water, tissue-simulating phantoms and dog kidneys have been detected. The analytic signal magnitude has been employed in generating 'A-mode' images with excellent depth resolution. Thermoacoustic emissions have also been detected from the dose-gradient at the beam edges of a 4 MeV x-ray beam in water. These results establish the feasibility of employing thermoacoustic signals in generating diagnostic images, and in locating x-ray beam edges during radiation therapy. A theoretical model for thermoacoustic imaging using a directional transducer has been developed, which may be used in the design of future thermoacoustic imaging system, and in facilitating comparisons with other types of imaging systems. A method of characterizing biological tissues has been proposed, which relates the power spectrum of the detected thermoacoustic signals to the autocorrelation function of the thermoacoustic source distribution in the tissues. The temperature dependence of acoustic signals induced by microwave pulses in water has been investigated. The signal amplitudes vary with temperature as the thermal expansion of water, except near 4°C. The signal waveforms show a gradual phase change as the temperature changes from below 4° to above 4°C. This anomaly is due to the presence of a nonthermal component detected near 4°C, whose waveform is similar to the derivative of the room temperature signal. The results are compared to a model based on a nonequilibrium relaxation mechanism proposed by Pierce and Hsieh. The relaxation time was found to be (0.20±0.02) ns and (0.13±0.02) ns for 200 ns and 400 ns microwave pulse widths, respectively. A microwave-induced thermoacoustic source capable of launching large aperture, unipolar ultrasonic plane wave pulses in water has been constructed. This source consists of a thin water layer trapped between two dielectric media. Due to the large mismatch in the dielectric constants, the incident microwaves undergo multiple reflections between the dielectric boundaries trapping the water, resulting in an enhanced specific microwave absorption in the thin water layer. This source may be useful in ultrasonic scattering and attenuation experiments.

A study of the application of single-mode fiber optics to the multiple-beam interferometric recombination problem is presented. In the laboratory, the fibers have been used in wide bandwidth, two-arm, Mach-Zehnder test interferometers as well as a 5-telescope imaging interferometer connected to an all-fiber beam combiner. Based upon these experiments and some theoretical studies it is shown that fiber optics and fiber optic components such as directional couplers provide an excellent alternative to conventional optics such as mirrors, beamsplitters, and relay lenses. The equations describing the measurement of the complex degree of coherence in an interferometer with a single-mode fiber in each arm are derived. The equations reveal an important feature of the fibers: they filter phase fluctuations due to aberrations and turbulence at the input and convert them to intensity fluctuations at the output. This leads to a simplification of the calibration of measured visibilities. The coupling efficiency of light which has passed through a turbulent atmosphere is also studied as a function of fiber parameters and turbulence conditions for both image motion stabilized and non-stabilized cases. For the former case, coupling efficiency remains greater than 50% as long as telescope diameter is no larger than the turbulence coherence length. Beam combination architectures using arrays of directional couplers are fully discussed. Arrays accommodating up to 20 input beams are presented. The arrays require only N detector pixels for N input beams. A scheme of temporal multiplexing of the phase of each beam is used to identify individual fringe pairs. One possible scheme allows wide bandwidths even for large numbers of beams. A 5-telescope interferometer has been constructed and connected to an all-fiber beam combiner. Two extended objects were observed and reconstructed using standard radio astronomy VLBI software. The interferometer and beam combiner had good thermal and polarization stability and high throughput. Reconstructed images had dynamic ranges of about 50.

This dissertation presents realistic mathematical models of normal and diseased livers and a nuclear medicine camera. The mathematical model of a normal liver is developed by creating a data set of points on the surface of the liver and fitting it to a truncated set of spherical harmonics. We model the depth-dependent MTF of a scintillation camera taking into account the effects of Compton scatter, linear attenuation, intrinsic detector resolution, collimator resolution, and Poisson noise. The differential diagnosis on a liver scan includes normal, focal disease, and diffuse disease. Object classes of normal livers are created by randomly perturbing the spherical harmonic coefficients. Object classes of livers with focal disease are created by introducing cold ellipsoids within the liver volume. Cirrhotic livers are created by modelling the gross morphological changes, heterogenous uptake, and decreased overall uptake. Simulated nuclear medicine images are made by projecting livers through nuclear imaging systems. The combination of object classes of simulated livers and models of different imaging systems is applied to imaging-system design optimization in a psycho-physical study. Human observer performance on simulated liver images made on nine different systems is compared to the Hotelling trace criterion (HTC). The system with the best observer performance is judged to be the best system. The correlation between the human performance metric dₐ and the HTC for this study was 0.829, suggesting that the HTC may have value as a predictor of observer performance. Texture in a liver scan is related to the three-dimensional distribution of functional acini, which changes with disease. One measure of texture is the fractal dimension, related to the Fourier power spectrum. We measured the average radial power spectra of 70 liver scans. All of these scans yield straight lines when plotted on a log-log scale, a characteristic of fractal objects. The slope of the line is related to the fractal dimension of the acini. The slopes are significantly higher for normal than abnormal livers (t = 4.04, df = 29, p = 0.005). On 32 liver scans with confirmed diagnoses, receiver operating characteristics (ROC) analysis was performed using power spectral slope as a feature. Analysis of the ROC curve yielded an area under the curve of 85, suggesting that power spectral slope may be a useful classifier of disease.

Imaging in nuclear medicine involves the injection of a radioactive tracer into the body and subsequent detection of the radiation emanating from an organ of interest. Single-photon emission computed tomography (SPECT) is the branch of nuclear medicine that yields three-dimensional maps of the distribution of a tracer, most commonly as a series of two-dimensional slices. One major drawback to transaxial tomographic imaging in SPECT today is the rotation required of a gamma camera to collect the tomographic data set. Transaxial SPECT usually involves a large, single-crystal scintillation camera and an aperture (collimator) that together only satisfy a small portion of the spatial sampling requirements simultaneously. It would be very desirable to have a stationary data-collection apparatus that allows all spatial sampling in the data set to occur simultaneously. Aperture or detector motion (or both) is merely an inconvenience in most imaging situations where the patient is stationary. However, aperture or detector motion (or both) enormously complicate the prospect of tomograhically recording dynamic events, such as the beating heart, with radioactive pharmaceuticals. By substituting a set of small modular detectors for the large single-crystal detector, we can arrange the usable detector area in such a way as to collect all spatial samples simultaneously. The modular detectors allow for the possibility of using other types of stationary apertures. We demonstrate the capabilities of one such aperture, the pinhole array. The pinhole array is one of many kinds of collimators known as coded apertures. Coded apertures differ from conventional apertures in nuclear medicine in that they allow for overlapping projections of the object on the detector. Although overlapping projections is not a requirement when using pinhole arrays, there are potential benefits in terms of collection efficiency. There are also potential drawbacks in terms of the position uncertainty of emissions in the reconstruction object. The long-term goal of the research presented is dynamic SPECT imaging of the heart. The basic concepts and tasks involved in transaxial SPECT imaging with pinhole arrays are presented along with arguments for the combination of modular gamma cameras and pinhole arrays. We demonstrate by emulation two methods of tomographically imaging a stationary single object slice and present results for these two systems on object space grids of 10cm x 10cm and 20cm x 20cm.