Showing papers on "Dark-frame subtraction published in 1984"

Combined homomorphic and local-statistics processing for restoration of images degraded by signal-dependent noise.

Abstract: The generalized homomorphic transformation to make signal-dependent noise independent of the signal is combined with a local-statistics image restoration technique to process images degraded by signal-dependent noise. Experimental results are given for images degraded by film-grain noise and by multiplicative noise.

Region-of-interest digital substraction angiography

Abstract: In digital subtraction angiography a low X-ray energy temporal subtraction image is displayed on a video monitor. Regions in the temporal image frame that contain motion artifacts are outlined by using a cursor. The coordinates of the pixels in the defined outlined region are stored in a processor memory. The data for a related hybrid subtraction image are developed and stored. The processor then effects substitution of the hybrid subtraction image pixels, that fall within the same defined region, into the temporal subtraction image. The combined image data are transferred to a display controller memory which controls display of the combined image on a monitor. A method and means are provided for automatically determining the optimum value of the weighting coefficient applied to the high energy temporal subtraction image that results in the most complete cancellation of everything in the two images except the contrast medium filled blood vessel when the high and low energy temporal images are subtracted. A method and means are provided for altering the gain and offset of the hybrid image data in such manner as to match the hybrid image contrast and brightness with that of the temporal subtraction image.

A system for reducing the noise in a television signal

Abstract: The amount of noise reduction of a television signal produced by averaging corresponding pixels from one frame to the next is reduced by a correction factor when motion in the picture is detected. In order to prevent very active motion from disturbing detection of the noise level, a noise level measurement signal is subjected to a hold of its value in a time briefly preceding a highly active motion in the picture until a time shortly after the end of such highly active motion. The motion signal is processed so as to provide such a hold only when motion is highly active in the picture, and that signal is prolonged in a further circuit so that it will have a duration long enough to cover build-up of that motion and its trailing off, when the noise measurement signal is delayed by about half of the prolongation, so that the noise measurement level that is held corresponds to the noise level prior to the buildup of the motion which shortly thereafter becomes highly active. The noise value thus held controls the size of the correction factor (k) produced by the presence of the motion signal.

Noise analysis of a digital radiography system

Abstract: The sources of noise in a digital video subtraction angiography system were identified and analyzed. Signal-to-noise ratios of digital radiography systems were measured using the digital image data recorded in the computer. The major sources of noise include quantum noise, TV camera electronic noise, quantization noise from the analog-to-digital converter, time jitter, structure noise in the image intensifier, and video recorder electronic noise. A new noise source was identified, which results from the interplay of fixed pattern noise and the lack of image registration. This type of noise may result from image-intensifier structure noise in combination with TV camera time jitter or recorder time jitter. A similar noise source is generated from the interplay of patient absorption inhomogeneities and patient motion or image re-registration. Signal-to-noise ratios were measured for a variety of experimental conditions using subtracted digital images. The measured signal-to-noise ratios were found to fluctua...

Digital subtraction angiography: fundamental noise characteristics.

Abstract: Detectability of low-contrast objects (vessels) in subtraction imaging depends on the noise properties of the image. There is a structure imposed upon the quantum noise that is due to the attenuation of the subtracted background object; this structure is referred to as a "noise print." It is shown to be a function of the manner of presubtraction processing (linear or logarithmic) and the local thickness of the attenuator in the x-ray beam path. The influence of the noise print on the visibility of vessels in subtraction images is demonstrated.

Short communications Digital grey-scale fluorography: a new approach to digital radiographic imaging

Abstract: In this paper we show that the presence of structure noise, shading and vignetting limits the low-contrast performance of radiological imaging systems which incorporate an X-ray image intensifier. As these imperfections are reproduced exactly at each exposure they are eliminated, along with the patient's background anatomy, during digital subtraction angiography. We have developed an imaging algorithm, based on the digital subtraction technique, which eliminates systematic imperfections from grey-scale (radiographic) images. Preliminary studies indicate that this algorithm, which we call "digital greyscale fluorography" (DGF), offers new opportunities to apply digital imaging techniques in classical radiography. DGF: THE SCIENTIFIC BASIS A radiologist's ability to discriminate low-contrast lesions is limited by, among other factors, the power of the noise in the diagnostic image. If a single static image is recorded via a digital X-ray image intensifier television system, four sources of noise can be identified: X-ray quantum mottle, electronic noise arising within the television channel, noise arising from the digitisation process and fixed-pattern or structure noise (Hay & Cowen, 1981). Structure noise arises from localised and generalised inhomogeneities in the sensitivity of the various image transfer processes within the X-ray image intensifier and in the photoconductive target layer of the television camera tube. In addition, vignetting and irregular shading due to the radiation field and the coupling optics introduce a powerful contribution of low spatial frequency noise that limits the degree to which it is possible to take advantage of display windowing techniques. Cowen and Coleman (1983) have described subjective measurements that indicate that the visual threshold detectability of low contrast details in small-format fluorography is strongly influenced by the presence of structure noise. In this study a Leeds threshold-contrast detail-detectability test object type TO 9 F was used to measure the low-contrast perceptibility of small-format fluorograms which were produced at a standard exposure level (viz. 25 nC kg" 1 measured at the input plane of the image intensifier). The experiment was repeated with a unit neutral density filter placed in the optical path in order to reduce the sensitivity of the fluorographic process and thereby increase the input exposure per image by approximately an order of magnitude. For X-ray quantum-limited imaging it is possible to predict that this should have resulted in an improvement in low-contrast sensitivity by a factor of ,yiO (Sturm & Morgan, 1949). In practice we found that contrast sensitivity improved by a factor of only v /2 , indicating that increased exposure has much less effect on information transfer than has previously been assumed. These findings were supported by objective measurements of the noise power (Wiener) spectra of fluorograms (Cowen & Coleman, 1983). Fluorographic film samples were obtained under the two conditions previously described {i.e., at the standard exposure level and with a tenfold reduction in speed). The reduction in noise power density is much less than would be predicted by quantum-limited imaging; this again implies that other sources of noise exist in the fluorographic imaging process. Unpublished results of measurements we have undertaken on other makes and types of X-ray image intensifier confirm these subjective and objective findings. A digital image of the Leeds threshold-contrast detaildetectability test object type TO 9 DSF, together with the structure noise in the X-ray image intensifier and television system used in the present investigation, is shown in Fig. 1. The test object was attached to the input window of the X-ray image intensifier. The laminar design of this test object ensured that image degradation due to scattered radiation and focal spot unsharpness was minimised. A 1.5 mm thick copper filter was placed in the beam and the digital subtraction system was programmed to produce a peak tube voltage of 75 kV. The exposure was made under the automatic control of the digital

Noise reduction methods for hybrid subtraction

Abstract: In digital subtraction angiography, hybrid subtraction provides selective vessel images free of soft-tissue motion artifacts but with a lower signal-to-noise ratio (SNR) than temporal subtraction images. An image processing method called measurement-dependent filtering has been developed to enhance the SNR of hybrid images without losing resolution or selectivity. Linear combinations of four images consisting of a pre- and postcontrast dual-energy measurement pair form both the hybrid image and a lower noise but less selective vessel image. The noise-reduced image is derived by combining the low-frequency components of the hybrid image with the high-frequency components of the lower noise image in a variety of ways. The results of the filtering method, when tested on both phantom and clinical data, display images with about the same degree of conspicuity as the hybrid image and a SNR approaching that of the temporal image.

A line transfer color image sensor with 576×462 pixels

Abstract: An image sensor that uses pixel elements consisting of both photodiodes and MOS capacitor storage, and incorporates blooming suppression, will be discussed.