Search PubMed⌕ Search

Biomedical subjects

C E Metz

Publications and source records attributed to C E Metz.

101 records · Page 6Linked to original sources
PubMedDate not supplied

Investigation of basic imaging properties in digital radiography. 4. Effect of unsharp masking on the detectability of simple patterns.

The signal-to-noise ratios (SNRs) of simple radiologic patterns processed by the unsharp-masking technique were calculated on the basis of a statistical decision theory model that includes both the observer's visual transfer function and a noise component internal to his eye-brain system. For a variety of processing parameters, the contrast-detail diagrams predicted from this SNR agreed qualitatively with experimental results obtained in observer performance studies. Unsharp masking with a large mask and a large weighting factor improved the detection of square objects to a level comparable with that achieved by the overall contrast enhancement technique using a factor of 4. However, unsharp masking with a small mask and a large weighting factor can substantially degrade the detectability of these objects. The potential practical utility of the unsharp-masking technique is discussed.

Humans↗
PubMedDate not supplied

Contrast enhancement of noisy images by windowing: limitations due to the finite dynamic range of the display system.

A theoretical model has been developed to explain the effects of simple linear windowing on the apparent contrasts of signals in displayed digital images. The model predicts, and experimental results demonstrate, that the effective displayed contrast of a digital radiographic signal depends in a complex way upon interactions among the endpoints of the display scale, the signal contrast and noise level of the original data, the window center and display center selected, and the contrast enhancement factor applied. The results obtained from this work apply quantitatively to the highly idealized situation in which (i) a uniform signal is superimposed on a uniform background containing Gaussian pixel-value noise, and (ii) a linear (or exponential) relationship exists between the optical density of a film display (or the light intensity of a luminous display) and pixel value in some finite range. However, the qualitative effects demonstrated here may be expected to arise in a broad variety of situations involving strong digital contrast enhancement.

Computer Simulation↗
PubMedDate not supplied

Investigation of basic imaging properties in digital radiography. 13. Effect of simple structured noise on the detectability of simulated stenotic lesions.

We investigated the effects of structured background noise on the detectability of stenotic lesions. Digital subtraction angiographic (DSA) images of stenotic blood vessels were simulated and superimposed onto uniform noise samples. Eighteen-alternative forced choice (18-AFC) experiments were employed to determine the detectability of the stenotic lesion in the structured-noise background of a blood vessel. In this study, the dependence of detectability on lesion size, vessel size, and incident x-ray exposure was examined. Our results indicate that the presence of structured noise in an image will reduce the detectability of a lesion. However, the relative performance of an observer when the lesion size and incident exposure were varied was the same with and without the presence of the structured background. Thus, conclusions obtained previously with regard to changes in the detectability of a lesion in the presence of uniform background noise can be applied directly to conditions in which simple structured anatomic background is present.

Angiography↗
PubMedDate not supplied

Determination of three-dimensional structure in biplane radiography without prior knowledge of the relationship between the two views: theory.

A novel method allows the determination of three-dimensional object structure from two projection images that are obtained at arbitrary, unknown orientations. Only minimal prior information concerning the imaging system is required. First, the image coordinates of eight or more object points that can be identified unambiguously in both views are used to determine the relative geometry of the two projections. Subsequently, the three-dimensional coordinates of the identified object points are determined, to within a scale factor, from the image coordinates of the points and the calculated imaging system geometry. A theoretical description of the overall method is provided, along with techniques for the reduction of effects of experimental errors and numerical errors that may arise in the course of the calculations. Methods to retrieve the absolute scale of the object are discussed also.

Mathematics↗
PubMedDate not supplied

Propagation and reduction of error in three-dimensional structure determined from biplane views of unknown orientation.

We are developing a technique for determination of the three-dimensional (3-D) structure of vascular objects from two radiographic projection images acquired at arbitrary and unknown relative orientations. No separate calibration steps are required with this method, which exploits an inherent redundancy of biplane imaging to extract the imaging geometry as well as the 3-D locations of eight or more object points. The theoretical basis of this technique has been described previously. In this paper, we review the method from the perspective of linear algebra and describe an improvement, not heretofore reported, that reduces the method's sensitivity to experimental error. We then examine the feasibility and inherent accuracy of this approach by computer simulation of biplane imaging experiments. The precision with which 3-D object structure may be retrieved, together with the dependence of precision on the actual imaging geometry and errors in various measured quantities, is studied in detail. Our simulation studies show that the method is not only feasible but potentially accurate, typically determining object-point configurations with root-mean-square (RMS) error on the order of 1 to 2 mm. The method is also quite fast, requiring approximately one second of CPU time on a VAX 11/750 computer (0.6 MIPS).

Angiography↗
PubMedDate not supplied

Computerized detection of pulmonary nodules in digital chest images: use of morphological filters in reducing false-positive detections.

Currently, radiologists can fail to detect lung nodules in up to 30% of actually positive cases. If a computerized scheme could alert the radiologist to locations of suspected nodules, then potentially the number of missed nodules could be reduced. We are developing such a computerized scheme that involves a difference-image approach and various feature-extraction techniques. In this paper, we describe our use of digital morphological processing in the reduction of computer-identified false-positive detections. A feature-extraction technique, which includes the sequential application of nonlinear filters of erosion and dilation, is employed to reduce the camouflaging effect of ribs and vessels on nodule detection. This additional feature-extraction technique reduced the true-positive rate of the computerized scheme by 13% and the false-positive rate by 50%. In a comparison of the scheme with and without the additional feature-extraction technique, inclusion of the additional technique increased the detection sensitivity by about half at the level of three to four false-positive detections per chest image.

False Positive Reactions↗
PubMedDate not supplied

Computerized detection of masses in digital mammograms: analysis of bilateral subtraction images.

A computerized scheme is being developed for the detection of masses in digital mammograms. Based on the deviation from the normal architectural symmetry of the right and left breasts, a bilateral subtraction technique is used to enhance the conspicuity of possible masses. The scheme employs two pairs of conventional screen-film mammograms (the right and left mediolateral oblique views and craniocaudal views), which are digitized by a TV camera/Gould digitizer. The right and left breast images in each pair are aligned manually during digitization. A nonlinear bilateral subtraction technique that involves linking multiple subtracted images has been investigated and compared to a simple linear subtraction method. Various feature-extraction techniques are used to reduce false-positive detections resulting from the bilateral subtraction. The scheme has been evaluated using 46 pairs of clinical mammograms and was found to yield a 95% true-positive rate at an average of three false-positive detections per image. This preliminary study indicates that the scheme is potentially useful as an aid to radiologists in the interpretation of screening mammograms.

Breast Neoplasms↗
PubMedDate not supplied

Gains in accuracy from replicated readings of diagnostic images: prediction and assessment in terms of ROC analysis.

From the perspective of signal detection theory, human variation in image reading degrades diagnostic accuracy by broadening the statistical distributions of perceived evidence upon which decisions were based. A new multivariate "random-effects" model formulates the total variation in a diagnostic decision variable as a sum of three uncorrelated components that represent differences among cases, readers, and repeated readings by a given reader. This model provides a basis for quantitative predictions concerning the amount by which diagnostic accuracy, as specified by ROC analysis, can be enhanced by the replication of image readings. Although these predictions apply exactly only to the hypothetical situation in which normally distributed decision variables from equivalent readers are averaged, computer-simulation studies and an analysis of mammographic image-reading data from five radiologists show that similar gains in accuracy can be achieved by averaging discrete confidence ratings.

Decision Support Techniques↗
PubMedDate not supplied

Statistical comparison of two ROC-curve estimates obtained from partially-paired datasets.

The authors propose a new generalized method for ROC-curve fitting and statistical testing that allows researchers to utilize all of the data collected in an experimental comparison of two diagnostic modalities, even if some patients have not been studied with both modalities. Their new algorithm, ROCKIT, subsumes previous algorithms as special cases. It conducts all analyses available from previous ROC software and provides 95% confidence intervals for all estimates. ROCKIT was tested on more than half a million computer-simulated datasets of various sizes and configurations representing a range of population ROC curves. The algorithm successfully converged for more than 99.8% of all datasets studied. The type I error rates of the new algorithm's statistical test for differences in Az estimates were excellent for datasets typically encountered in practice, but diverged from alpha for datasets arising from some extreme situations.

Algorithms↗