Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-02-09
2003-09-23
Chang, Jon (Department: 2623)
Image analysis
Applications
Biomedical applications
C382S199000, C382S224000
Reexamination Certificate
active
06625303
ABSTRACT:
FIELD OF THE INVENTION
This invention relates in general to a method for automatically locating instances of a specified image pattern in a digital image, and more specifically relates to a method for the automatic detection and classification of cylindrical bone patterns in digital projection radiographs.
BACKGROUND OF THE INVENTION
Image processing algorithms that can automatically detect and classify patterns in digital imagery have broad application in consumer photography, remote sensing, and medical imaging. The present invention describes a flexible and computationally efficient approach for detecting and classifying patterns in digital imagery that has specific application to digital projection radiography. The method is useful for image visualization processing and computer aided diagnosis (CAD). The present invention has been optimized for the detection and classification of cylindrical bone patterns in digital projection radiographs.
Finding the location of the anatomical region-of-interest (ROI) in a digital projection radiograph is a critical step for enhancing the brightness, contrast and sharpness of an image. Once the ROI has been identified, standard techniques such as tone scaling and unsharp masking can be used to optimally render the image for diagnostic interpretation. The automatic segmentation of the bone and soft tissue anatomical regions is especially useful for generating image renderings that maximize the diagnostic information content that is presented to the radiologist for various types of exams. This processing must be fully automated to facilitate workflow in busy hospital imaging departments. Automatic detection of patterns in digital radiographs also has application to CAD. For example the automatic identification of certain bone patterns can be used as a pre-processing step for computer-assisted image interpretation. Once a specific bone pattern has been located, the bone morphology can be analyzed to obtain information about bone mineral density, bone growth, and fractures. The present invention has been optimized for locating the class of “cylindrical” bones, e.g. the humerus, femur, fingers, ribs, etc. However the technique is very flexible and may be generally applied to the detection of patterns in other image types such as for the automatic detection of cancerous masses and micro-calcifications in mammography. The technique is especially useful for higher dimensional spatial or spectral imagery, e.g., 3-dimensional CT or dual energy x-ray capture. In these applications, simple features can be used to first locate candidate regions which are then examined and classified by more detailed analysis.
Measurements of bone density have become an essential criterion for evaluating a patient's risk of osteoporostic fracture. Commercially available instruments for measuring the bone mineral density (BMD) are divided into X-ray absorptiometry, quantitative computed tomography (QCT), and quantitative ultrasound (QUS). QCT allows the 3D visualization of trabecular microstructure and provides assessments beyond the basic BMD result, such as biomechanical parameters describing bone strength. However, QCT requires a CT scan and is an expensive procedure. Dual-energy X-ray absorptiometry (DXA) of the spine, femur, and total-body is a widely utilized method for predicting a patient's risk of fracture. However, in many geographic areas there are inadequate resources to meet the demand, and DXA scans are not available to all patients who might benefit. Moreover, conventional DXA scanning is perceived as costly because of the need to refer patients to hospital-based facilities. Recently, there has been a wide variety of innovative equipment available for a small, low-cost dual-energy X-ray absorptiometry device dedicated to scanning the peripheral skeleton, for example, the forearm. As discussed in Christiansen et al. (C. Christiansen, P. Ravn, P. Alexandersen, and A. Mollgaard, “A new region of interest (nROI) in the forearm for monitoring the effect of therapy,” Journal of Bone Mineral Research, 12 (suppl 1): S480, 1997.), BMD measurements at forearm sites are well proven in predicting fracture risk. There is also a single-energy X-ray device, called radiographic absorptiometry (RA), where BMD in the hand (that is, the fingers) is assessed using a radiograph calibrated with an aluminum wedge. In the most recent development, devices designed to acquire a direct digital radiographic image of the hand enable bone density analysis to be performed in a physician's office.
The challenge for computer aided diagnosis of bone disease is to position the region of interest for a highly precise measurement of BMD. Various methods for locating the bone region of interest have been proposed. For example, in order to use a posterior/anterior (PA) chest radiograph to analyze lung texture, the inter-rib bones shown in the PA chest image need to be removed. U.S. Pat. No. 4,851,984, “Method and system for localization of inter-rib spaces and automated lung texture analysis in digital chest radiographs”, issued Jul. 25, 1989, to inventors K. Doi, et al., teaches a method to locate inter-rib spaces in digital chest radiograph images. First, a lung field is defined by determining the rib cage edge boundary. A horizontal signal profile is obtained at a predetermined vertical location. The pixel location at which the second derivative of this horizontal profile is minimum is defined as the rib cage edge boundary. Then two vertical profiles in the periphery of both lungs are fitted with a shift-invariant sinusoidal function. This technique assumes that the horizontal line is always perpendicular to the spinal column. Moreover, it assumes that the relative vertical locations of the objects are known a priori. Therefore, U.S. Pat. No. 4,851,984 does not teach a fully automatic method for locating instances of an image pattern. Furthermore, the profiles of the sinusoidal function do not accurately fit the profile of the cylindrical bone structure.
Histogram methods have been used to locate the bone region of interest. U.S. Pat. No. 5,228,068, “Device and method for automated determination and analysis of bone density and vertebral morphology”, issued Jul. 13, 1993, to inventor R. B. Mazess, teaches a method to determine and analyze vertebral morphology by evaluating the approximate center location of each vertebra from a digital lateral vertebral scan. The centers are located by evaluating the horizontal and vertical histograms. The horizontal histogram is constructed along a line which crosses each anterior-posterior border of the vertebra. The vertical histogram is obtained along a line crossing the superior-inferior border, which directs the spine column, because the patient is supported in the supine position on a table so that the vertebrae of the spine are generally aligned with the scan direction. However, because of the curvature of the spine, the angle of the vertebrae, that is, the angle of the anterior border, the posterior border, the superior border, and the inferior border with respect to the scan direction will vary among vertebrae. This method requires that this variation be accommodated by the trained eye of a physician in estimating the initial positions of lines which horizontal and vertical histogram are generated from. The rigid assumption about the relative orientation of the image relative to the body makes this technique sensitive to orientation error and can not be used in automatic mode for general image orientations.
Another histogram-based method disclosed in U.S. Pat. No. 4,951,201, issued Aug. 21, 1990, “Method of automatically determining imaged body posture in medical image display”, to inventors H. Takeo et al, is used to determine the image body posture. In order to produce an optimal visualization of the anatomy, the projection must be known. For example, for a chest image, the imaged thoracic vertebrae is of a relatively low density when it is imaged from the front side, and of relatively high density when it is imaged from the lat
Lee Hsien-Che
Young Susan S.
Chang Jon
Kim Chong
Noval William F.
LandOfFree
Method for automatically locating an image pattern in... 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 for automatically locating an image pattern in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automatically locating an image pattern in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3108442