The localization of neural electromagnetic sources from measurements at the head surface requires the solution of an inverse problem; that is, the determination of the number, location, spatial configuration, strength, and time-course of the neuronal currents that give rise to the magnetic field or potential distribution. In most general form, the neuromagnetic and electrical inverse problems are ill-posed and have no unique solution; however, approximate solutions are possible if assumptions are made regarding the shape and conductivity of the head and the number and configuration of neuronal currents responsible for the surface distributions. To help resolve ambiguities and to reduce the number and range of free parameters required to model complex neuromagnetic sources, the authors are investigating strategies to constrain the locations of allowable sources, based on a knowledge of individual anatomy. The key assumption, justified by both physiological evidence and theoretical considerations, is that the dominant neuromagnetic sources which contribute to surface field distributions reside within the cortex. It is demonstrated that anatomically constrained source modeling strategies can produce significant improvements in source localization; however, the conclusion is that additional improvements in model fitting or source reconstruction procedures are required.

The acquisition of radial lines in 2-D Fourier space (k space) allows the realization of the extremely short echo times useful in the imaging of short-T2 species and flow using Magnetic Resonance (MR). The disadvantage of this imaging method, which is also known as projection reconstruction, is that in order to prevent aliasing at a given resolution, (pi) times as many excitations are required as the conventional 2-D Fourier transform or spin-warp acquisition. The authors propose an acquisition and reconstruction method that halves the required number of radial lines and excitations by using the properties that projections are real or have slowly varying phase and that radial acquisition methods are over-sampled at the center of k space. This method preserves resolution while reducing imaging time at the expense of signal-to-noise ratio (SNR). This has been verified with both phantom and human subjects.

A method for Magnetic Resonance Imaging (MRI) was investigated, whereby an object is put under a homogeneous magnetic field, and the image is obtained by applying inverse source procedures to the data collected in an array of coil detectors surrounding the object. The induced current in each coil due to the precession of the magnetic dipole in each voxel depends on the characteristics of both the magnetic dipole frequency and strength, together with its distance from the coil, the coil direction in space and the electrical properties of the coils. By calculating the induced current signals over an array of coil detectors, a relationship is established between the set of signals and the structure of the body under investigation. Based on the proposed method, a computer simulation demonstrates the feasibility of this new modality. An improved method of multicoil recording is also suggested, whereby it is combined with the conventional zeugmatographic method with read and phase gradients, to result in a novel method of magnetic resonance imaging. In the combined method an equivalent number of coils is used instead of encoding gradients. The number of coils is thus reduced many times in comparison with the method where only a multicoil array is used. An experimental setup with a 9 coils detector array was built to give a coarse resolution of 3X3 pixels. By measuring the induced current signals over this array of coil detectors, a relationship is established between the set of signals and the structure of the body under investigation. The linear relation can then be represented in matrix notation, and inversion of this matrix will produce an image of the body.

A real-time volumetric ultrasonic imaging system has been improved through the use of parallel processing in two dimensions. The imaging system uses pulse-echo phased array principles to steer a two-dimensional transducer array in a pyramidal scan format. The transducer array consists of 96 transmit elements and 32 receive elements, and operates at a center frequency of 2.5 Mhz. Parallel processing allows reception of 16 unique view directions distributed over a small pyramidal volume during receive. Information from the 16 directions is obtained simultaneously after the transmission of an ultrasonic pulse. Images consisting of 4096 lines are produced at a rate of over 15 frames/second. The images are analogous to those produced by an optical camera or the human eye and supply more information than conventional sonograms. Echo data for the scanned volume is presented as projection images with depth perspective, stereoscopic pairs, multiple tomographic images, or C-scans. Potential medical applications include improved anatomic visualization, tumor localization, and better assessment of cardiac function.

Interest in 3D medical imaging continues to increase. However, in ultrasound, real-time imaging is an indispensable strength; and real-time 3D ultrasonic imaging is not practical when conventional steered, focused beam techniques are used. This is because the speed of sound severely limits the size of the volume that can be imaged in real time. For real-time 3D imaging, approaches like simultaneous multiple beams or holography have been considered but never commercially implemented for, in part, the following reasons: A new 3D ultrasound technology should provide the convenience of a hand-held scan head, should yield real-time 3D images, and should provide 2D images with quality equal to, or greater than, presently available 2D ultrasound images. Convenient size and a reasonable price are also requirements. In this paper, a 3D ultrasonic imaging method with the potential to meet the above criteria is described. It may also provide even higher quality 2D ultrasound images than are presently available. The new method relates more closely to computed tomography than to focused steered beams. It, however, uses projections and back-projections over 3D ellipsoids rather than straight lines; and it does this in a simple straight forward manner. Implementation in software of filtered ellipsoidal back-projection is described, resolution and side lobes are discussed, and examples of the 3D point image (re. point spread function) are given.

Ultrasonic echoes, backscattered from an inhomogeneous medium have the character of a random signal, which is mainly responsible for the observed speckle in medical images. Such a medium can be modeled as a uniform matrix with scattering bodies distributed randomly. When the number of density of scatterers is high, the individual scatterers are not resolved by the imaging process, and a speckle pattern is produced as a result of interference of waves from many scatterers within the resolution cell volume. This cell volume depends on the beam profile and the pulse width of the interrogating pulse. The authors have used a 3D simulation phantom that takes into account the 3D distribution of scatterers and the 3D nature of the resolution cell volume. Several simulations were performed to study the effect of scatterer number density (SND) and resolution cell volume on the backscattered signal. Assuming the process is linear and the stochastic signal is ergodic and stationary, Kurtosis (K), which involves 2nd and 4th moments, was estimated in each case. It was found that Kurtosis varies linearly with another parameter Fs that depends on the resolution cell volume. The results are analyzed in the light of theoretical predictions. Reasonable estimates of SND can be derived from the slope of Kurtosis vs. parameter Fs graph.

This paper addresses the problem of imaging a dynamic object with an imaging modality that utilizes a single illumination instead of sequential sector scanning used in conjunction with B-scanners. Two data collection strategies and their associated reconstruction algorithms are presented for dynamic object imaging. One scheme utilizes a single unfocused wave transmitted by an element of an array. In the other method, the elements of an array simultaneously illuminate the object with unfocused multiplexed waves. The data base acquired in either of these methods is sufficient for the recovery of the object's monostatic synthetic aperture data and reconstruction.

Recent advances in the fabrication of sensors and transistors from hydrogenated amorphous silicon are allowing the creation of flat panel, large area, radiation-damage-resistant arrays for radiotherapy and diagnostic imaging. A straightforward application of available a-Si:H technology involves the configuration of photodiode sensors coupled with field effect transistors into regular 2-dimensional pixel patterns. While the specifications of the array design must be tailored to the demands of the imaging application, the design is at the same time constrained by various array parameters. Considerations affecting choices for these parameters and how they relate to the signal, noise, and readout properties of radiotherapy arrays are presented and some initial data reported. Implications for a-Si:H diagnostic imagers are discussed.

The signal-transfer and noise behavior of cesium iodide (CsI) layers, which are used as input screens of X-ray image intensifiers, have been investigated. Experimentally the performance of the CsI screen has been studied with a laboratory-type X-ray image intensifier as well as with a medical device. The scintillation spectrum (gain distribution), the spatial resolution (MTF), the image noise and the corresponding signal-to-noise (S/N) ratios have been determined. The measurements show that the DQE of the X-ray image intensifier is almost exclusively determined by the input screen performance for a wide range of spatial frequencies. The experimental data are compared with the results of a simulation. The screen model includes the primary X-ray absorption process, the effects of K-fluorescence escape and reabsorption, as well as the optical properties of the dedicated (needlelike) CsI screen structure. The simulated signal-to-noise (S/N) ratios are in good agreement with the experimental results. The analysis shows the influence of the various physical processes on the S/N-performance of the screen. The observed drop of the DQE below the absorption limit, which is most pronounced at higher spatial frequencies, is strongly related to the escape and the reabsorption of K-fluorescence.

A linear self-scanning diode x-ray imaging system for the chest has been modified so that its characteristics for mammographic imaging can be measured. The system consists of an x-ray tube and self-scanning linear diode array assembly which moves in synchrony above and below a phantom or patient. The advantage of this type of system is that scattered radiation is rejected through the use of the line scanning geometry, resulting in improved image quality and less radiation than with area imaging systems. The line scanning digital system is characterized for resolution, MTF, linearity, and contrast sensitivity. Resolution in the scan direction is 6 lp/mm; perpendicular to the scan direction it is 5 lp/mm at 55 kVp and 10% MTF. As currently configured, the S/N ratio of the digital mammographic images is too low for it to be satisfactorily implemented as a mammographic system. S/N could be increased through the use of a more powerful x-ray tube, a multi-slit system, and longer scan times, which would make it more viable for mammography.

By local variation of exposure, the wide absorption-range of the thorax can be equalized. Normally, the X-ray spectrum remains constant during this process. Further improvement seems possible if the radiation quality is included in the equalization process. A system has been built to investigate the merits of such a process. An overall improvement of contrast is demonstrated.

The authors investigate the potential application of a novel 2048X2048-matrix image intensifier (II)/TV based real-time digital radiography system for GI-tract examinations. The basic imaging properties of the system and clinical image evaluations have been studied. Imaging properties measured have been compared with those of the conventional screen/film system, computed radiography system (CR), and 100-mm roll-film system. It was found that this DR system can be applied not only to mass screening but to accurate imaging in gastrointestinal examinations.

The imaging chain of a bi-planar fluoroscopic system is described for a new neurosurgical technique: the Video Tumor Fighter (VTF). The VTF manipulates a small intracranially implanted magnet, called a thermoseed, by a large external magnetic field gradient. The thermoseed is heated by rf-induction to kill proximal tumor cells. For accurately guiding the seed through the brain, the x-ray tubes are alternately pulsed up to four times per second, each for as much as two hours. Radio-opaque reference markers, attached to the skull, enable the thermoseed's three dimensional position to be determined and then projected onto a displayed MRI brain scan. The imaging approach, similar to systems at the University of Arizona and the Mayo Clinic, includes a 20 cm diameter phosphor screen viewed by a proximity focused microchannel plate image intensifier coupled via fiberoptic taper to a solid state camera. The most important performance specifications are magnetic field immunity and, due to the procedure duration, low dosage per image. A preliminary arrangement designed in the laboratories yielded usable images at approximately 100 (mu) R exposure per frame. In this paper, the results of a series of studies of the effects of magnetic fields on microchannel plate image intensifiers used in the image detection chain are presented.

The authors previously described how image reconstruction algorithms can be evaluated on the basis of how well binary-discrimination tasks can be performed using the reconstructions. The test statistic in the detection task was the estimated activity within the object, also known as the non-prewhitening matched-filter output. This approximation to the likelihood function was used because a full characterization of the posterior probability function had not yet been performed. A more complete approach is possible when the reconstruction procedure is founded on the Bayesian method. In this case the reconstruction is chosen to maximize the posterior probability and task performance involves using the posterior probability of the various alternatives as the decision variable. This full Bayesian approach should lead to optimal results because the posterior probability incorporates the full dependence on the measurements and constraints, yet is based on the relatively simple likelihood and prior probability distributions. The commercially available code MEMSYS 3 provides Bayesian image reconstructions based on an entropy prior. This paper details a method of image reconstruction evaluation based on the full posterior probability ratio, and describes results obtained using images derived from the MEMSYS 3 algorithm for a simple binary detection task. The results demonstrate the improvement in detection performance that can be achieved when the full posterior probability function is incorporated into the decision variable.

A methodology for measuring observer internal noise has been extended to measure display perceptual noise. Observer performance experiments were performed in which the observer's task was to detect and locate several test objects (dots) at random locations in a noise field. From this Free-Response Receiver Operating Characteristic (FROC) experiment the ROC detectability parameter d' was extracted. By performing this experiment at two levels of the externally added noise, one can determine the noise added by the display and the observer. The method takes into account the contrast of the display, the pixel size and the viewing distance.

This paper gives a color analysis of the nonlinear phase modulation for density pseudocolor encoding technique in medical application. Theoretically, instead of the traditional rectangular model, a trapezoid model is used for calculating chromaticity of the image. Thus a color specification using this technique can be set up as a guide for medical diagnosis. Some medical color images and a new sample modality are also presented.

A new compact, manual-load computed radiography system is suitable for siting at remote nodes on a digital radiology network. The storage phosphor plates are permanently installed in a standard-size cassette and removed only by the system during reading and erasing.

This paper deals with methods of reducing the total time required to acquire the projection data for a set of contiguous CT images. Normally during the acquisition of a set of slices, the patient is held stationary during data collection and translated to the next axial location during an inter-scan delay. The authors will demonstrate, using computer simulations and scans of volunteers on a modified scanner, how acceptable image quality is achieved if the patient translation time is overlapped with data acquisition. If the concurrent patient translation is ignored, structured artifacts significantly degrade resulting reconstructions. A number of algorithms are presented to minimize the structured artifacts through the use of projection modulation using the data from individual and multiple slices. Comparison is made of the methods with respect to structured artifacts, noise, resolution and susceptibility to motion. Review of preliminary clinical feedback by a panel of radiologists has indicated that the residual image degradation is tolerable for selected applications when it is critical to acquire more slices in a patient breathing cycle than is possible with conventional scanning. The method is a useful protocol when some image quality can be traded for increased scan rate. Applications include increased contrast utilization and minimization of registration artifacts.

In this paper, a new approach to the Filtered BackProjection (FBP) algorithm is presented. The method is based on the reconstruction stability in Sobolev spaces and B-spline functions which define a Pixel Intensity Distribution Model (PIDM-n) according to the spline degree n of the desired reconstruction. It is shown that PIDM-n reconstructions can be efficiently obtained. Angular sampling is studied and comparison with standard FBP shows the superiority of the algorithm presented. Moreover, simulation studies of noise degradation and blur in the projections show the algorithm to be superior to FBP in this more realistic case.

Simulated and real Positron Emission Tomography (PET) images are reconstructed using the iterative EM algorithm. The data is split up into independent parts. The EM algorithm is applied to each part and stopped according to a cross-validation procedure. For a variety of simulated and real data sets, stopping points were reached. For simulated data, the average of the reconstructions from a four-way split of the data was visually superior to the reconstruction obtained by applying the EM algorithm to the full data and stopping at a subjectively chosen iteration. For real data, such an improvement was not observed. To remove point artifacts, the reconstructions were filtered.

A digital angiographic system has been modified into a prototype of an image intensifier-based volume CT imager for angiography. The modified system has been tested by a vascular phantom. The system consists of two x-ray tubes and two image intensifiers that are separately mounted on a gantry. The two tube-detector sets can be rotated over 360 degree(s) in the gantry. To explore the imaging performance of the system for reconstructing a three-dimensional (3D) vascular structure a set of nonsubtraction projections of a vascular phantom, acquired over 36 projection angles, were digitized. These data were reconstructed using an iterative algorithm specially designed for 3D vascular structures. The quality of the reconstructed vascular images indicates that the system can offer adequate signal-to-noise ratio (SNR) for direct 3D vascular reconstruction when only a few projections are used without subtraction procedure, assuming intra-arterial injection of contrast. Also, a pincushion distortion correction algorithm has been developed and the results suggest that the algorithm works well for the image intensifier-based volume CT imager.

The authors use a Toshiba LX-40A Radiation Therapy Simulator with a 14' (35.6 cm) image intensifier and a 1' saticon camera to collect data for computed tomography (CT) imaging. The custom designed data acquisition system is interfaced with a 386-PC and the LX-40A to allow the LX-40A to perform as a CT scanner under PC control. The motion versatility of the simulator allows fields of view (FOV) greater than 40 cm with a single 360 degree(s) rotation of the gantry. Distortion and other corrections are applied to give relatively artifact-free images. A visible resolution performance of 7 lp/cm is obtained throughout the FOV. One percent contrast targets are visible down to 3 mm for head-sized objects, though contrast sensitivity depends, of course, on many scan parameters. An objective of this research is to give CT functionality to a radiation therapy simulator; this eliminates the need for a conventional, diagnostic CT scanner for radiation therapy planning. Extensions to multislice and volume CT are possible.

This work is concerned with truly 3D X-ray tomography. The method consists in the acquisition of an object's radiographs for different positions of an X-ray cone beam source. The image is then obtained by solving a 3D reconstruction problem from cone beam projections. When considering a series expansion approach, the problem is equivalent to the resolution of a linear system, presenting very particular characteristics in size and sparseness. The authors investigate the use of block iterative techniques which allow an efficient implementation of the algorithm on a parallel computer. Three different block iterative reconstruction schemes are developed. They can be used with or without simple constraints on the solution (positivity, amplitude, support...). Results obtained on simulated images allow comparison to the convergence properties of the different methods. Contrary to the conventional case in truly 3D X-ray tomography, different trajectories of the cone beam source are considered and the first results obtained on simulated objects are discussed.

Three-dimensional Computed Tomography (CT) with cone-beam geometry will take the place of a conventional CT because 3-D CT is superior to conventional CTs in terms of both data acquisition time and spatial resolution, particularly along the rotation axis. The small bone structures are reconstructed using magnified projections of the region of interest. This paper describes the solution for some problems in practical use. Experimental results show 3-D CT will be feasible.

The ocular lens is an import variable refractive element in the eye; its transparency is vital for vision. Conventional ophthalmic imaging devices, such as MRI, ultrasound, and optical instruments are unable to image the ocular lens at submicron resolution. Confocal light microscopy can image the in situ ocular lens and produce high contrast images of the semi-transparent structures. The transverse resolution is less than one micron and the range resolution (important for optical sectioning) is submicron. The studies were performed on freshly enucleated rabbit eyes using a real-time confocal microscope developed by Xiao and Kino at Stanford University. This microscope is based on a rotating Nipkow disk. Other confocal studies were performed on a laser scanning optical microscope. The following structures were imaged: lens capsule, lens epithelium, lens fibrils, nuclei and intrafibril structures. The two-dimensional data sets obtained from the optical sectioning properties of the confocal microscope were used to reconstruct the three-dimensional voxel data sets of the lens. Volume rendering techniques were used for the reconstruction. The three-dimensional views of the in-situ ocular lens are used for visualization of the ocular lens. Clinical potential of the three-dimensional visualization of the ocular lens will be reviewed.

A method of imaging through highly scattering media has been applied to the reconstruction of a transaxial slice across a cylindrical phantom containing a low absorbing, high scattering solution. The imaging method involves recording anddiscriminating between the times-of-flight of transmitted photons. Images of the phantom are presented which were generated using light transmitted through the phantom with the shortest flight times.