Search

The goal of the research effort funded under this grant is to finalize the construction of two prototype DXM-1 slot-scanning digital mammography systems, conduct a technical evaluation of these prototypes and perform two clinical studies to assess the efficacy of the systems in the clinical environment. The technical evaluation and one of the clinical studies is being conducted at the Radiology Department at the University of Arizona (UA). The objective of the technical evaluation is to determine system performance parameters such as dynamic range, modulation transfer function, sensitivity, etc. The second clinical study is scheduled to be conducted at Sharp Healthcare/Sidney Kimmel Cancer Center in San Diego, California and will establish the clinical efficacy of the system compared with conventional film/screen mammography. One thousand women are scheduled to participate. Based on preliminary data gathered thus far on the DXM-1 system. The system response is linear with exposure dose and it appears that the dominant noise source is quantum (x-ray) noise. Efforts are now underway to improve system performance by removing artifacts generated in the image reconstruction process, to eliminate parasitic capacitance on the readout IC, and to increase resolution by interleaving both banks of sensors into the displayed image.

Mammography is the most effective method for early detection of breast cancer and yet, 10% to 30% of women who have breast cancer and undergo mammography have negative mammograms. Furthermore, of the women who are sent to biopsy, only 20% to 40% actually have breast cancer. Quantitative analysis of the radiographic features of microcalcifications and masses may help radiologists improve their specificity.

According to Peter Ogle of Digital Imaging (San Francisco, CA), “Radiologists should help identify common values for the use of information technology in medicine.” Achieving a set of common values often requires a framework for organizing the thought process involved. That is the focus of this article in addressing the subject of picture archiving and communication system (PACS) technology assessment.

We have now analyzed the use and benefits of four Strategic Thinking Tools for Radiology: the Vision Statement, the High Five, the Two-by-Two, and Real-Win-Worth. Additional tools will be provided during the tutorial. The tools provided above should be considered as examples. They all contain the 10 benefits outlined earlier to varying degrees. It is extremely important that the tools be used in a manner consistent with the Vision Statement of the organization. The specific situation, the effectiveness of the team, and the experience developed with the tools over time will determine the true benefits of the process. It has also been shown that with active use of the types of tools provided above, teams have learned to modify the tools for increased effectiveness and have created additional tools for specific purposes. Once individuals in the organization become committed to improving communication and to using tools/frameworks for solving problems as a team, effectiveness becomes boundless.

The mathematical foundation for a linear lumped transistor model is developed from the nonlinear regional model proposed by Schilling. A one-dimensional model is described with simple analytic equations representing carrier density and electric field profiles and stored charge across the base and collector regions of a bipolar transistor. The regional boundaries are derived as simple functions of the device parameters, e.g., impurity profile, and operating conditions at its terminals. A one-to-one correspondence between these physical device parameters and lumped model elements is provided by the model. The linear regional model is demonstrated to be equivalent to its nonlinear counterpart, consequently to Gummel's charge control model.

In most work dealing with space‐charge‐limited currents, the diffusive contribution to the current is neglected. The classic steady‐state problem was handled by Mott and Gurney. For their problem of the perfect insulator it is well known that diffusion dominates near the injecting contact but, after some critical distance, drift dominates, and it is consistent to neglect diffusion.For the case of transient SCLC the mathematical validity of neglecting diffusion has not been shown. The purpose of this work is to show that for times greater than a critical time and distances greater than a critical distance the diffusive term can consistently be neglected. The solutions found by Mark and Helfrich and by Many and Rakavy for injection of electrons into an insulator under an applied voltage pulse have been shown to be correct under the above conditions. The interesting result is that the critical parameters are related to geometrical parameters by the ratio of the thermal voltage kT/q to the applied voltage. We therefore find that for signals large compared to kT/q, the neglect of diffusion is valid over most of the range of time and distance associated with the problem.

The transient response of the current to an applied voltage step is analyzed for the range of time approaching equilibrium, using a perturbation scheme applied to the steady‐state equations. A modal equation is developed from which an infinite set of decaying, oscillatory modes are shown to exist. The time constant for the system is shown to be approximately one‐fourth of the transit time associated with one‐carrier space‐charge‐limited currents.

A regional approach to transistor design is presented which makes possible the determination of transistor gain hFE for any transistor (doping profile specified) operating at any given values of emitter current IE and collector-to-emitter voltage VCE. This approach extends the classical diffusion theory of transistors to include conductivity modulation, voltage-controlled base motion, and current-controlled base motion (base widening). Due to the simplicity of this technique, time-sharing computer techniques were used. This allowed for user interaction, making this approach very attractive for education as well as design. The regional approach is a likely candidate for satisfying the needs for the device designer (it contains device performance versus doping profile), the circuit designer (it allows determination of equivalent circuit models from knowledge of the electron and hole densities throughout the device), and the academic world (it is simple and physical). It is anticipated that the regional approach will allow engineers to develop a "feel" for transistor behavior under wide ranges of operating conditions.

Publisher Summary This chapter discusses the development of magnetic resonance imaging (MRI) technology. MRI represents the leading edge of technology in the field of diagnostic imaging. MRI has achieved clinical significance and has become the primary modality in the investigation of head, spinal cord, pelvis, cancellous bone, pericardium, gall bladder, and lymph node masses and in the evaluation of joints. Hydrogen, which is available in high concentrations in the body, together with its associated magnetic properties, provides the key ingredient for the medical applications of MRI. The MRI requires the tissue being scanned to be exposed to three magnetic fields. The application of these fields, in coordination with the detected signals, is controlled by a computer. Powerful modern minicomputers, based on the latest very-large-scale integration (VLSI) technology, are required to handle the significant quantity of data that is generated and processed in the MRI systems. The unique approach to MRI system architecture is based on the identification of a number of tasks that can be performed in parallel. These tasks are very demanding when performed by a single computer but can be performed with a minimum of special hardware by an optimum combination of a microcomputer for control and a pair of minicomputers or super-microcomputers arranged in dual-computer architecture for data acquisition, image processing, display, and archiving.

Publisher Summary This chapter presents a discussion on current injection in solids through the regional approximation method. Very substantial progress can be made by accepting a loss of accuracy in favor of an approximation scheme, which not only brings the previously intractable problems back into the fold of analytic tractability but emphasizes the underlying physics of each problem in doing so. This scheme, which is called the “regional approximation method,” is based on the simple observation that where there are several terms in an equation “competing” with each other, each being a function of position, it is generally possible to divide up the volume of the solid into separate regions in each of which either a single term or a couple of terms dominate. Within each such region, the basic approximation is made of neglecting all terms within the competing group except the one or two terms, which dominate in that region. The chapter discusses basic, prototype injection problems and device-design problem. The chapter also discusses treat one-carrier problems with planar-flow and spherical, radial-flow geometries, respectively. The chapter discusses two-carrier, planar-flow problems, first injected plasmas and then varying-lifetime, negative-resistance problems. The chapter also describes a planar-flow transistor design problem with base-widening as the key feature.