Bennett R. Groshong

Which media are most likely to solve the archival problem

The clinical application of quantitative methods for coronary arteriography remains limited, due in large part to the absence of a suitable replacement for cinefilm as the procedure record. The extension to the clinical environment of the validated objective methods which have found such widespread acceptance in clinical research studies is difficult to implement if the time-consuming and variable process for digitization of selected cinefilm frames is required. In addition, the complete integration of the angiographic procedure record with other patient records and procedures stored in a digital data format requires that the angiographic data eventually be converted to a digital format as well. Replacement of cinefilm requires that the media chosen for the task provide at least the same capabilities and preferably improved functions as those provided by cinefilm as a display, transport, and archival media. The demanding set of requirements imposed on the replacement options include high capacity, high acquisition rate, high transfer rate, application in a distributed environment, portability between institutions, and low expense. A true digital solution should also provide immediate access to the results of the angiographic procedure, transfer of image data over digital networks, multiple-user viewing capability, and quantitative analysis on a routine basis for all patients. In fact, a single media may not provide all the capabilities listed above but, rather, different media may need to be used for specialized tasks, i.e. the solution for archival may not be the same that will be employed as the portable patient record. Separation of the archival function from the acquisition/display and portable transfer functions increases the likelihood that cinefilm can be replaced in the imminent future by reducing the demands on a single media. Among the archival options available today are: (1) magnetic disks; (2) analog laser optical disks; (3) digital laser optical disks; (4) digital file-based magnetic tape; (5) digital video magnetic tape. In evaluating each of these alternatives, an accounting is required of how each meets the archival requirements along with an approximate breakdown of cost and readiness for implementation as a clinical solution today.

Eddy Current Image Restoration by Constrained Gradient Descent

We present in this paper an image restoration method for recovering the geometry of a flaw in a conducting surface from an eddy current image. Image restoration is formulated as a maximum likelihood estimation problem and is solved using constrained iterative gradient descent. A novel frequency-domain constraint relaxation algorithm is used to control the iterative restoration process. The technique is applied to eddy current images generated by sampling the output of an absolute transducer on a dense grid, creating smoothly blurred images of crack-like flaws.The following results were obtained: (1) Images of a range of sizes of electrical discharge machined (EDM) slots in the surface of a non-ferrous block that encompassed the transducer coil diameter were successfully restored. (2) Excellent results were also obtained for synthetic flaw images with very low signal-to-noise ratios. (3) Synthetic images of multiple flaws spaced less than a coil diameter apart were restored. (4) synthetic images of “V” and pit-shaped flaws were successfully restored.The principal conclusions drawn from this work are: (1) Representing a nonlinear system with layers of linear blurs and non-linear point transformations is general, permitting efficient gradient descent on any analytic system function. (2) Smooth blurring functions in the gradient improve the stability of the algorithm. (3) The accuracy of the estimate is improved dramatically by restoring portions of estimate spectrum which correspond to the highest signal-to-noise ratio band in the observation spectrum first.

An eddy current transducer model for image restoration

We describe in this short paper a new imaging model for the spatial magnitude response of an absolute eddy current transducer to a flaw. This model is appropriate for image restoration purposes in that it captures the transducer response with sufficient accuracy for image restoration, yet is simple enough to be computationally practical. The model described in this paper is based on a simple resistive loop approximation to the transducer impedance changes induced by a flaw, and is efficiently implemented as layers of linear blurring functions and nonlinear point operations. The model is shown to accurately reflect the magnitude response of an absolute transducer to an EDM slot in a non-ferrous alloy. Furthermore, the model is shown to produce good restoration results for both synthetic and real images of flaws. The model may be adapted to a particular combination of absolute transducer and surface flaw type by optimizing the model parameters, either by forming the partial derivatives with respect to the parameters and minimizing by gradient descent, or by a straightforward implementation of a neural net back-propagation algorithm. Other types of eddy current transducers may be modeled by simply modifying the spatial layers to combine the local terms appropriately for the geometry of the transducer.

Quantitative analysis of transesophageal echocardiograms for the intraoperative setting: clinical need and initial experience

Two computerized analysis systems that provide objective interpretation of transesophageal echocardiography (TEE) images are described. A semi-automated system has been used to study systolic wall thickening (SWT) as a means of assessing the effects of anesthetics on TEE. Representative data are presented, showing marked decrease in SWT as a result of halothane administration. Initial experience with this system has demonstrated persistently high interexamination variability of SWT measurements, which is largely the result of its reliance on manual, user-dependent description of the cardiac outlines. In order to correct this problem, as well as to be able to utilize the analysis process into the real-time interoperative environment, development of a more fully automated system that incorporates computed detection and estimation of cardiac boundaries has begun. The implementation of the system and the procedure it uses are described. A number of steps to improve the automated boundary detection process are given.<<ETX>>

Estimating simple closed contours in images

A technique for estimating the boundary of objects that may be described by a simple closed contour is presented. The problem is posed as one of maximum a posteriori (MAP) optimization. The method employs an elliptical Fourier series to describe the contour. A scaled derivative-of-Gaussian description of the object boundary cross-section is employed, coupled with guided gradient descent estimation of the contour coefficients. The technique is applied to X-ray images of the left ventricle, but is easily extensible to a wide variety of images of similar objects. It is shown that a few elliptical harmonics accurately model the ventricle outline. Accurate, robust estimation of the left ventricle outline from a single image is shown for a set of images.<<ETX>>

Design and clinical evaluation of a high-capacity digital image archival library and high-speed network for the replacement of cinefilm in the cardiac angiography environment

An economical and practical digital solution for the replacement of 35 mm cine film as the archive media in the cardiac x-ray imaging environment has remained lacking to date due to the demanding requirements of high capacity, high acquisition rate, high transfer rate, and a need for application in a distributed environment. A clinical digital image library and network based on the D2 digital video format has been installed in the Duke University Cardiac Catheterization Laboratory. The system architecture includes a central image library with digital video recorders and robotic tape retrieval, three acquisition stations, and remote review stations connected via a serial image network. The library has a capacity for over 20,000 Gigabytes of uncompressed image data, equivalent to records for approximately 20,000 patients. Image acquisition in the clinical laboratories is via a real-time digital interface between the digital angiography system and a local digital recorder. Images are transferred to the library over the serial network at a rate of 14.3 Mbytes/sec and permanently stored for later review. The image library and network are currently undergoing a clinical comparison with cine film for visual and quantitative assessment of coronary artery disease. At the conclusion of the evaluation, the configuration will be expanded to include four additional catheterization laboratories and remote review stations throughout the hospital.

Maximum a posteriori optimization-a method for calculation of dynamic changes in ventricular contours from angiographic image sequences

A robust and accurate method for automatic determination of dynamic left ventricular function has been developed for application to sequences of angiographic images. The method employs a maximum a posteriori optimization approach. The shape of the ventricle boundary is modeled by an elliptical Fourier series. The intensity cross section of the ventricle boundary is modeled using a derivative of a Gaussian. Prior knowledge of the probability of a given ventricle shape and temporal behavior is encoded using a distribution on the parameters. This prior knowledge is coupled with the spatial and temporal information in the image sequence using a guided gradient descent technique to estimate the dynamic contour coefficients. The results illustrate the performance of the optimization algorithm on two examples of digital subtraction X-ray ventriculograms.<<ETX>>

Multimodality cardiac image review station using dynamic decompression of JPEG images

A cardiology review station for clinical use has been developed which can decompress, zoom, and display full resolution JPEG-encoded cardiac angiograms and echocardiograms in real- time (30 frames/second). The review station is installed in a network which includes a digital image archival system throughout Duke University Medical Center. The review station consists of a DEC 3000/600 AXP workstation with a DEC J300 sight and sound multimedia board. The J300 is used to decompress up to 45 true color JPEG-encoded frames/sec. at normal (512 X 512) resolution and 30 frames/sec. at 1.5 X zoom. Digitally acquired 512 X 512 X 8 bit monochrome angiograms as well as 600 X 430 X 8 bit color and monochrome echocardiograms are transferred to the workstation where they are JPEG-encoded in real time using the J300 board. A compression factor of approximately 15:1 is being used. A graphical user interface (GUI) developed using OSF/Motif 1.2 enables a clinical user to simultaneously display and control several image sequences. A sequence can be retrieved in under three seconds and displayed dynamically in forward or reverse directions with instantaneous speed control. Utilizing a commercial relational database (Sybase), the GUI organizes image sequences for a patient by image modality, location, time, and view. A schematic representation of cardiac anatomy allows a user to view specific angiographic image sequences by selecting appropriate objects in the anatomic diagram.