Image analysis – Applications – Biomedical applications
Reexamination Certificate
1998-12-08
2001-07-03
Bella, Matthew C. (Department: 2621)
Image analysis
Applications
Biomedical applications
C382S132000
Reexamination Certificate
active
06256403
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to an image processing method for determining a noise curve relating to an image, which method includes the acquisition of an image formed by pixels having an intensity which includes a signal component and a noise component which is dependent on the signal. The invention also relates to an image processing method for reducing the noise of an image, and to a medical imaging apparatus provided with a system for carrying out such a method.
The invention is intended for use in particular in a medical X-ray imaging system. The noise in X-ray images is dependent on the signal. This means that the noise level in the dark zones and in the bright zones of the image is not identical. The noise can be defined on the one hand by a value which is called the standard noise deviation relative to the mean noise. This standard noise deviation on the other hand can be represented as a function of the signal by a so-called noise curve. Noise is linked to a probability density concerning the intensity levels. In the case of video the probability density is that of Gaussian noise, being an electronic noise which is independent of the signal. The noise curve representative of the standard noise deviation as a function of the signal is then a constant. In the case of X-rays the probability density is that of a Poisson noise distribution which stems from quantum noise and this quantum noise is dependent on the signal. The noise curve representative of the standard noise deviation as a function of the signal is normally linear and ascending in the case of a Poisson noise distribution. A problem encountered is that X-ray systems include electronic devices which modify the shape of these noise curves. In the case of Poisson noise distribution the noise curve is generally not linear and not constantly ascending. It is not that of a true Poisson noise distribution, because the noise is passed through the entire electronic chain. Therefore, this noise curve must be determined in each case.
2. Description of Related Art
An image processing method which determines, in the course of processing, the variance of the noise in a sequence of digital images in which the noise is dependent on the signal is already known from the article “An adaptive technique for digital noise suppression in on-line portal imaging” by Konad W Leszcynski, published in Phys. Med. Biol., 1990, Vol. 35, No 3,429-439, printed in the UK. The cited document describes a noise reduction method which takes into account the variance of the noise. To this end, the variance of the intensity of the pixels is measured on-line in a uniform region in an averaged image during the processing of the images. This averaged image is the average of N individual video images. The variance is represented as a function of 1/N. The variance of the noise is reduced by said factor N in the averaged image. Thus, the variance of the random noise is estimated by way of the slope of the local variance.
SUMMARY OF THE INVENTION
In a first example of an application, relating to X-ray imaging, it is very important to reduce the noise because the signal applied to the patient is generally very weak since these images are generally very noisy. In order to achieve the reduction of the noise in an image it is necessary to evaluate the noise and hence to determine the standard noise deviation curve as a function of the signal for the reasons described above. In another example of an application, relating to video, it may also be necessary to reduce the noise before encoding because this noise contains high frequencies and the encoding of high frequencies is very costly.
SUMMARY OF THE INVENTION
The present invention offers a solution to this problem by providing an image processing method as defined in claim
1
.
The method according to the invention offers the advantage that it can be used for estimating quantum noise as well as for estimating electronic noise, i.e. for estimating noise with a Poisson distribution as well as for estimating noise with a Gaussian distribution. Another advantage resides in the fact that the method is suitable for real-time execution.
An image processing method for reducing the noise in an image involves the execution of such a method as defined in claim
9
.
A medical imaging apparatus provided with means for reducing the noise in a noisy digital medical image includes means as defined in claim
10
.
REFERENCES:
patent: 5649021 (1997-07-01), Matey et al.
patent: 5673332 (1997-09-01), Nishikawa et al.
patent: 5909515 (1999-06-01), Makram
patent: 5911012 (1999-06-01), Bernard et al.
“An adaptive technique for digital noise suppression in on-line port imaging” by Konad W. Leszcynski, published in Phys. Med Biol. 1990, vol. 35, No. 3,429-439.
Charpail François R. G.
Florent Raoul
Makram-Ebeid Sherif
Soyer Christel
Bella Matthew C.
Charpail François R. G.
Choobin M B
U.S. Philips Corporation
Vodopia John F.
LandOfFree
Method of processing an image in which the noise is... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of processing an image in which the noise is..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of processing an image in which the noise is... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2520001