Rheumatoid arthritis (RA) is an autoimmune disease that causes chronic inflammation of the joints. X-rays can show bony erosions typical of rheumatoid arthritis in the joints. Joint X-rays can also be helpful in monitoring the progression of disease and joint damage over time. In order to diagnose arthritis in early stage, image segmentation is very useful. Image segmentation is the process of separating out mutually exclusive homogeneous regions of interest. Image segmentation may be thought of as consisting of two related processes-recognition and delineation. Recognition is the high-level process of determining roughly the whereabouts of an object of interest in the image. Delineation is the low-level process of determining the precise spatial extent and point-by-point composition of the object in the image.
An algorithm is used for automatically detecting bone contours from hand radiographs using active contours. Initial contours have to be located for the snakes inside each bone of interest. Next, an adaptive snake algorithm is applied so that parameters are properly adjusted for each bone specifically. A novel truncation technique to prevent the external forces of the snake from pulling the contour outside the bones boundaries is used. The Snake algorithm is a feature extraction technique, also known as active contours. It works like stretched Elastic Band being released. The Initial Points defined around Feature to be extracted, explicitly has to be defined then pre-defined number of Points generated. Points are moved through an Iterative Process. "Energy Function" for each point in the Local Neighbourhood is calculated.
TABLE OF CONTENTS
CHAPT TITLE PAGE NO
LIST OF TABLES
LIST OF FIGURES
LIST OF ABBREVATIONS
1.1 GENERAL INTRODUCTION
Get your grade
or your money back
using our Essay Writing Service!
List of figures
List of Abbreviations
RA - Rheumatoid arthritis
CR - Conventional Radiography
DXA- Digital X ray absometry
US - Ultrasound
MRI- Magnetic Resonance Imaging
1.1 GENERAL INTRODUCTION
RheumatoidÂ arthritisÂ is a form of arthritis that causes pain, swelling, stiffness and loss of function in your joints. It can affect any joint but is common in the wrist and fingers. More women than men get rheumatoid arthritis. It often starts between ages 25 and 55.Doctors use X-rays and blood tests to diagnose rheumatoid arthritis, an autoimmune disorder that causes chronic inflammation and damage to the joints. While X-rays alone are not usually sufficient to diagnose rheumatoid arthritis, they can provide doctors with good clues so they can order blood tests to confirm the diagnosis. In order to detect the arthritis in early stage, automatic image segmentation algorithms are used in X-ray radiographs.
1.2 X - RAYS
X-rays are produced when electrons strike a metal target. The electrons are liberated from the heated filament and accelerated by a high voltage towards the metal target. The X-rays are produced when the electrons collide with the atoms and nuclei of the metal target.
1.2.1 DISCOVERY OF XRAY
A German physicist, W. C. Roentgen discovered X-rays on November 8, 1895. He produced and detectedÂ electromagnetic radiationÂ in aÂ wavelengthÂ range today known asÂ X-raysÂ or Roentgen rays, an achievement that earned him the firstÂ Nobel Prize in PhysicsÂ in 1901.He also made the first radiographic images of human anatomy. Roentgen's first experiment late in 1895 was a film of his wife Bertha's hand with a ring on her finger fig 1.2. Radiographic image is a picture of this x-ray distribution. In his discovery Roentgen found that the X-ray would pass through the tissue of humans leaving the bones and metals visible. In early 1896, X-rays were being utilized clinically in the United States for such things as bone fractures and gun shot wounds.
Roentgen was working with tubes similar to our fluorescent light bulbs. He evacuated the tube of all air, filled it with a special gas, and passed a high electric voltage through it; the tube would produce a fluorescent glow. Roentgen shielded the tube with heavy black paper, and found that a green colored fluorescent light could be seen coming from a screen setting a few feet away from the tube. It produced a previously unknown "invisible light," or ray, that was being emitted from the tube; a ray that was capable of passing through the heavy paper covering the tube. Through additional experiments, he also found that the new ray would pass through most substances casting shadows of solid objects on pieces of film. He named the new ray X-ray, because in mathematics "X" is used to indicated the unknown quantity.
Always on Time
Marked to Standard
X-rays have wavelengths much shorter than visible light, but longer than high energy gamma rays. Their wavelength is well suited to study crystal structures and details of the human body. In addition, several objects and processes in the Universe emit X-rays. These X-rays are the messengers revealing information of the cosmos.
1.2.2 XRAYS IN MEDICAL DIAGNOSIS
Radiography is defined as the fields of radiology, and gave raise to radiologists, physicians who specialize in the interpretation of medical images. A short duration pulse of X- rays is emitted by the X ray tube, a large fraction of the x-rays interacts the patient, and some of the x-rays passes through the patients and reach the detectors, where the radiographic image is formed. The detector used in radiography can be photographic film or an electronic detector system. Radiography is performed with an x-ray source on one side of the patient and a x-ray detector on the other side. Roentgen is considered the father ofÂ diagnostic radiology, the medical specialty which uses imaging to diagnose disease.
FIG 1.1Spectrum showing xray
FIG 1.2 A, FIG 1.2 B,
1.3 HUMAN HAND ANATOMY
The human hand is one of the most complex appendages in all of nature. It is used to both grasp large objects firmly and small objects softly. Hands are located at the ends of the arms, connected by the wrists.
There are 27 bones in the hand: eight carpal bones at the bottom of the palm; five metacarpal bones in the middle of the hand; five proximal phalanges at the bases of the fingers and thumb; five distal phalanges at the tips of the fingers and thumb; and four intermediate phalanges as the middle bones of the fingers.The three parts of the hand are the digits, palm and dorsum. There are five digits on each hand, four fingers and one thumb that are located at the sides of the hand.
Finger pads--located at the finger tips just below the nails--enable humans to have a strong sense of touch. Fingerprints are unique to every human. Nails are located at the top tip of every digit on the hand. The palmar side of the hand (the palm) has a cushiony surface.The back side of the hand (dorsum) is bonier and has thicker veins than the palmar side.
1.3.3 Abductor Muscles
Abduction refers to movement of a limb away from the central line of the body or of a digit away from the axis of a limb. Muscles, then, that carry out this type of movement are called "abductor muscles."
1.3.4 Adductor Muscles
Adductor muscles move a limb toward the central line of the body or a digit toward the axis of a limb. It comes from the Latin word, "adductus," meaning "one that draws to". An adductor is also a muscle that closes the valves in a clam or other bivalve mollusk, but it isn't the one that snaps the human mouth shut.
1.3.5 Carpal Bones
The skeleton of the wrist consists of eight small "carpal bones" that are firmly bound in two rows of four bones each. The resulting mass is the "carpus." The eight bones are the "pisiform," "triangular" or "triquetrum," the "pisiform," "lunate," and "scaphoid" on the upper end of the wrist, where it connects with ligaments and the lower arm bones, and the "hamate," "capitate," "trapezoid," and "trapezium" on the lower side of the hand by the "metacarpals," or first joint of the fingers.
1.3.6Joints in the Hand
The joints where the phalanges attach to the metacarpal joints are called the metacarpophalangeal (MCP) joints. The joints between the proximal and middle segments are the proximal interphalangeal (PIP) joints and the segments between the middle and distal segments are the distal interphalangeal joints.
1.3.7 Fibrous Sheath
The digital synovial sheath is a membranous sheath enclosing the flexor tendons and serve to prevent bowstringing of the tendons during flexion.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
The phalanges are the small bones that make up the skeleton of the fingers, thumb and toes. Each finger and smaller toe has three phalanges; the thumb and big toe each have two. The phalange nearest the body of the hand or foot is call the "proximal" phalange; the one at the end of each digit is the "distal" phalange; and, of course, when there are three, the middle one is called the "middle" phalange.
1.3.9 Extensor Carpi Radialis Brevis
The extensor carpi radialis brevis is a companion of the extensor carpi radialis longus and extends toward the middle of it. This muscle runs from the humerus (upper arm bone) to the bones in the hand and functions to extend the wrist. It also assists in abducting the hand.
1.3.10Extensor Carpi Ulnaris
The extensor carpi ulnaris is located along the outer surface of the ulna (longer forearm bone) and connects the humerus (upper arm bone) to the hand. It acts to extend the wrist and assists in abducting it.
Muscle extension occurs when the angle between the bones is increased - as when the forearm is aligned with the arm. An "extensor", then, is any muscle serving to extend a bodily part (usually a limb) away from the body or when a digit is extended from its base.
A number of muscles that function to move the ankle, foot, and toes are located in the lower leg. They attach the femur (upper leg bone), tibia, and fibula (lower leg bones) and to various bones in the foot and are responsible for a variety of movements - moving the foot upward (dorsiflexion) or downward (plantar flexion), and turning the sole of the foot inward (inversion) or outward (eversion). The "dorsal flexor muscles" include the following: The "tibialis anterior" is a dorsal flexor and is an elongated, spindle-shaped muscle located on the front of the lower leg. It arises from the surface of the tibia to pass on the inside over the far end of the tibia, and attaches to bones of the ankle and foot. Contraction of the tibialis anterior causes dorsiflexion and inversion of the foot.
The tibialis posterior muscle helps stabilize the foot while walking. The "peroneus tertius" is a muscle of variable size that connects the fibula to the outside of the foot. It functions in dorsiflexion and eversion of the foot.
The"extensor digitorum longus" is situated along the outside of the lower leg, just behind the tibialis anterior. It comes from close to the midline of the tibia and the shaft of the fibula. Its tendon divides into four parts as it passes over the front of the ankle. These parts continue over the surface of the foot and attach to the four smaller toes. The actions of this muscle include dorsiflexion of the foot, eversion of the foot, and extension of the toes.
1.3.13Extensor Indices and Extensor Retinaculum
The extensor indicis muscle (not tendon) extends the index finger. A structure called the "extensor retinacula" consists of a group of heavy connective fibers in the tissues of the wrist. It connects the lateral margin of the radius (shorter forearm bone) with the inside border of the ulna (longer forearm bone) and with certain bones of the wrist. The retinaculum gives off branches of connective tissue to the underlying wrist bones, creating a series of sheath-like compartments through which the tendons of the extensor muscles pass to the wrist and fingers. As in the case of the wrist, the tissue formations in various regions of the ankle is thickened to form retinacula.
1.3.14 Flexor Digitorum Profundis
The flexor digitorum profundis is a large muscle that connects the ulna (longest forearm bone) to the bones of the fingers. It acts to flex the distal joints of the fingers, as when a fist is made.
1.3.15 Condyles of the Humerus
At the lower end of the humerus (upper arm bone) and the femur, there are two smooth condyles (rounded processes of the bone): a knob-like "capitulum" on the lateral side and a pulley-shaped "trochlea" in the middle. The capitulum unites with the radius (smaller lower arm bone) at the elbow, and the trochlea is a notch which joins ligaments to the head of the ulna (larger lower arm bone). Above the condyles on either side are "epicondyles," which provide attachments for muscles and ligaments of the elbow. The one toward the center of the arm is the "medial epicondyle," and the one to the side is the "lateral epicondyle."
1.3.16 Lumbrical Muscles
The lumbrical muscles are four small, fleshy muscles that are associated with tendons between the four small toes of the foot and the four fingers of the hand. Each lumbrical muscle flexes the corresponding digital joint to extend the finger or toe in question.
The metacarpal is one of five, long cylindrical bones in the body of the hand. The bones run from the carpal bones of the wrist to the base of each digit of the hand. On the palm of the hand, these are padded by a thick layer of fibrous, connective tissue; on the back of the hand, they can be seen and felt through the skin. The heads of the metacarpal bones form the knuckles.The opponens pollicis muscle flexes and adducts the thumb.The palmar interosseous muscles are adductors of the digits in the hand.
1.3.18 Ligaments of the Hand and Wrist
On the radial shaft of the smaller forearm bone, just below its head, is a process called the "radial tuberosity." It serves to attach the biceps brachi muscle, which bends the arm at the elbow. At the lower end of the radius, a lateral "styloid process" provides attachments for the "palmar radio carpal ligament" (on the palm of the hand) and the "dorsal radio carpal ligament" (on the back of the hand) from the radius into the wrist. At the lower end of the larger forearm bone (ulna), its knob-like head articulates with a notch of the radius (ulnar notch) laterally and with a disk of fibro cartilage below. This disk, in turn, joins a wrist bone (the triquetrum).
A medial "styloid process" at the lower end of the ulna provides attachments for ligaments ("palmar ulnocarpal ligament" and "dorsal ulnocarpal ligament") into the wrist. The skeleton of the wrist is made up of eight small "carpal bones" that are firmly bound in two rows of four bones each.
The mass that results from these bones is called the "carpus." The carpus is rounded on its nearest surface, where it articulates with the radius and with the fibro cartilaginous disk on the ulnar side.
The carpus is rounded convexly in the front, forming a canal (retinaculum) through which tendons, ligaments and nerves extend to the palm. Its distal surface articulates with the metacarpal bones, which are joined to the carpus by the "palmar carp metacarpal ligaments."
1.4 RHEUMATOID ARTHRITIS
Rheumatoid arthritis (RA) is an autoimmune disease that causes chronic inflammation of the joints. Rheumatoid arthritis can also cause inflammation of the tissue around the joints, as well as in other organs in the body. Autoimmune diseases are illnesses that occur when the body tissues are mistakenly attacked by its immune system. Because it can affect multiple organs of the body, rheumatoid arthritis is referred to as a systemic illness and is sometimes called rheumatoid disease. While rheumatoid arthritis is a chronic illness, meaning it can last for years, patients may experience long periods without symptoms. Rheumatoid arthritis is a progressive illness that has the potential to cause joint destruction and functional disability.
1.4.1 RISKS OF RHEUMATOID ARTHRITIS
Rheumatoid arthritis is a common rheumatic disease, affecting approximately 1.3 million people in the United States, according to current census data. The disease is three times more common in women as in men. It afflicts people of all races equally. The disease can begin at any age, but it most often starts after age 40 and before 60. In some families, multiple members can be affected, suggesting a genetic basis for the disorder.
Juvenile rheumatoid arthritis (JRA) is arthritis that causes joint inflammation and stiffness for more than six weeks in a child aged 16 or younger. It affects approximately 50,000 children in the United States. Inflammation causes redness, swelling, warmth, and soreness in the joints, although many children with JRA do not complain of joint pain. Any joint can be affected, and inflammation may limit the mobility of affected joints
1.4.2 CAUSES OF ARTHRITIS
The cause of rheumatoid arthritis is unknown. Even though infectious agents such as viruses, bacteria, and fungi have long been suspected, none has been proven as the cause. The cause of rheumatoid arthritis is a very active area of worldwide research. Some scientists believe that the tendency to develop rheumatoid arthritis may be genetically inherited. It is suspected that certain infections or factors in the environment might trigger the immune system to attack the body's own tissues; resulting in inflammation in various organs of the body such as the lungs or eyes. Environmental factors also seem to play some role in causing rheumatoid arthritis. Recently, scientists have reported that smoking tobacco increases the risk of developing rheumatoid arthritis
1.4.3 SYMPTOMS OF ARTHRITIS
Symptoms of RA include fatigue, lack of appetite, low-grade fever, muscle and joint aches, and stiffness. In the morning and after periods of inactivity, muscle and joint stiffness are usually most notable. Arthritis is common during disease flares. Also during flares, joints frequently become red, swollen, painful, and tender. This occurs because the lining tissue of the joint which is called as synovium, becomes inflamed, resulting in the production of excessive joint fluid called synovial fluid. The synovium also thickens with inflammation which is called as synovium.
Fig 1.5 NORMAL AND ARTHRITIC JOINTS
Fig 1.6 IMAGES OF HAND OF PATIENTS AFFECTED BY RHEUMATOID ARTHRITIS
Image segmentation is distinguishing objects from its background. Threshold techniques, edge-based methods, region-based techniques, and connectivity-preserving relaxation methods are the four basic approaches in image segmentation for intensity images. Various algorithms are used for segmentation of different medical images.
Threshold techniques, makes decisions based on local pixel information, which are effective when the intensity levels of the objects fall squarely outside the range of levels in the background.
A connectivity-preserving relaxation-based segmentation method usually referred to as theÂ active contour model. The main idea is to start with some initial boundary shape represented in the form of spline curves, and iteratively modifies it by applying various shrink/expansion operations according to some energy function.
2.1 A FULLY AUTOMATIC ALGORITHM FOR CONTOUR DETECTION OF BONES IN HAND RADIOGRAPHS USING ACTIVE CONTOURS
Authors : Rodrigo de Luis-GarcÂ´Ä±a, Marcos MartÂ´Ä±n-FernÂ´andez, Juan Ignacio Arribas, Carlos Alberola-LÂ´opez -. 2003 IEEE
In this paper automatic algorithm is used for detecting bone contours from hand radiographs. Prior knowledge is first used for locating the initial contours for the snakes in each bone of interest and an adaptive snake algorithm is used for further processing and parameters are properly adjusted for each bone of interest. This paper mainly presents an algorithm for automatic detection of bones of interest in hand radiographs, namely phalanges and metacarpal. Determination of bone age assessment has been done in this work
Knowledge about initial contours and application of snake algorithm, in hand radiographs has been obtained from this paper
2.2 AN ADAPTIVE SNAKE ALGORITHM FOR CONTOUR DETECTION
Author : Qin Zhoungyuan, Mou xuanqin , Wang Ping, Cai Yuanlong, 2002 IEEE
In this paper an adaptive snake algorithm is presented. To every point in the initial position, the energy of the point in its neighbors is calculated by greedy algorithm. If the target contour is not included in its neighbors, the radius of its neighbors can be increased. And the energy of all the points until target contour is calculated.
The target contour can achieve by iterating once. This algorithm can be easily adapted to recognition scheme.
2.3 SKETCH INITIALIZED SNAKES FOR RAPID, ACCURATE AND REPEATABLE INTERACTIVE MEDICAL IMAGE SEGMENTATION
Arthur: T. McInerney, M.R. Akhavan Sharif
In this paper they have combined a pen and pressure sensitive tablet input device, and a sketch based user initialization process, with a general subdivision curve snake. Which is used to create an inutive, fast, accurate, interactive model based segmentation method. Using the pen input device, the is snake is quickly and precisely initialized with a series of sketch lines such that it is extremely close to the position and shape of the target object boundary, making the Snake's task much simpler and hence more likely to succeed in noisy images with minimal user editing. The user may also use pen pressure levels to easily impart knowledge of object edge strength to the model, and the low degree-of-freedom subdivision curve Snake provides powerful control and editing capabilities. Snake is applied to the segmentation of several 2D medical images.
2.4 AN ACTIVE CONTOUR MODEL FOR IMAGE SEGMENTATION BASED ON ELASTIC INTERACTION
Arthur: Yang Xiang, Albert C.S. Chung , Jian Ye
In active contour method of segmentation, the boundaries of the objects are detected by evolving curves. In this paper, a new edge based active contour method is used for orientation dependent interaction between the image boundaries and moving curves by maintaining, edge fidelity. Fast Fourier Transform is used for calculating the velocity field for moving curves caused by elastic interaction. This method is based on the elastic interaction between line defects in solids. This feature can be used to examine medical images.
2.5 A FAST AND STABLE SNAKE ALGORITHM FOR MEDICAL IMAGES
Arthur: Dong Joong Kang, 8 December 1998
A dynamic programming algorithm is used to optimize an active contour. The external energy is derived from the image features; internal energy of model depends on the local behavior of the contour. This algorithm detects convex and concave objects rapidly even when the mage quality is poor. The curvature of the snake contour is only permitted in a predefined mean and variance. The proposed method maintains the stability of the snakes vertices after and before object motion from local similarity of the vertex motion.
2.6 THE VELOCITY SNAKE: DEFORMABLE CONTOUR FOR TRACKING IN SPATIO- VELOCITY SPACE
Arthur : Natan Peterfreund , August 28, 1998
In this paper velocity control is applied to the class of elastodynamical contour model for boundary tracking and prediction of position of non rigid objects. This control term minimizes an energy dissipation function which measures the differences between the contour velocity and apparent velocity of images.
2.7. COMPUTER ASSISTED BONE AGE ASSESSMENT BASED ON FEATURE AUTOMATICALLY EXTRACTED FROM A HAND RADIOGRAPHS
Arthur : Ewa Pietka, 14 September 1994
In this paper bone age assessment of pediatric patient is done using computer aided algorithm, which is based on the feature extraction from two regions of computer radiography (CR) left hand wrist images, phalangeal regions of interest (PROI) and carpal bone region of interest (CROI). A classification rule based on a max-sum operator processes the matrix assessing the bone age. Since both regions are analyzed independently, two bone age assessments are obtained. They reflect the phalangeal and carpal bones maturity individually.
2.8 AUTOMATIC LOCALIZATION AND TRACKINGOF MOVING OBJECTS USING ADAPTIVE SNAKE ALGORITHM.
Arthur: Jong Ryul Kim, Young Shik Moon, 18 December 2003.
In this paper, an algorithm for automatic localization and tracking of moving objects using adaptive snake model is proposed. The moving objects are automatically localized by using global motion estimatiod compensation and difference image. an adaptive snake algorithm is proposed in order to extract the exact object shape. The snake energy function adaptively changes according to the perpendicular minimum distance. In the final step, the object tracking is performed by using the initial template and the motion vectors. This algorithm can automatically extract multiple moving objects without any user-assisted initialization. It can also segment and track the objects more accurately than other existing algorithms.
2.9 DIAGNOSIS OF PERIARTICULAR OSTEOPOROSIS IN RHEUMATOID ARTHRITIS USING DIGITAL XRAY RADIOGRAMMETRY.
Arthur :Joachim Böttcher,Â and Alexander Pfeil
Osteoporosis can manifest in two ways in rheumatoid arthritis: generalized bone. Digital X-ray radiogrammetry (DXR) is an effective and sensitive modality for monitoring Periarticular osteoporosis, which is among the earliest features of rheumatoid arthritis, preceding bone erosions. DXR is a promising technique, which can provide quantitative data that allow early diagnosis. During the course of rheumatoid arthritis it can be deployed in combination with established X-ray scoring methods to inform decisions regarding the optimal therapy to prevent joint destruction.
2.10 PREDICTIVE EROSIVE DISEASE IN RHEUMATOID ARTHRITIS. A ONGIUDINAL STUDY OF CHANGES IN BONE DENSITY USING DIGITAL X-RAY RADIOGRAMMETRY: A PILOT STUDY
Arthur : A. StewartÂ ,Â L. M. Mackenzie, A. J. Black HYPERLINK "http://rheumatology.oxfordjournals.org/content/43/12/1561.abstract#target-2"Â andÂ D. M. Reid, July 23, 2004
Periarticular osteoporosis is one of the first radiological signs of rheumatoid arthritis (RA). Osteoporosis is now quantified using dual-energy X-ray absorptiometry (DXA), although it was originally assessed by radiogrammetry. A new updated system of radiogrammetry has been developed: digitized X-ray radiogrammetry (DXR). We used this DXR system to identify whether changes seen in hand X-rays of RA patients can predict those who subsequently develop erosions.
2.11 DETECTIION OF BONE EROSION IN EARLY RHEUMATOID ARTHRITIS: ULTRASONAGRAPHY AND CONVENTIONAL RADIOGRAPHY VERSUS NON- CONTRAST MAGNETIC RESONANCE IMAGNG
Arthur :Maryam Rahmani; Hosein Chegini; Seyed Reza Najafizadeh; Mohammad Azimi; Peiman Habibollahi; Madjid Shakiba
The early diagnosis and treatment of RA can be done in early stage by detecting signs of bone erosion using MRI. In this paper the aim of the study is to compare the specificity and sensitivity of US and CR to MRI. By several findings its been concluded that US and MRI , best suits for detection of bone erosion in patients with RA but not CR. US will be a valuable tool for early detection of bone erosion especially when MRI is not affordable.
From this we can infer that US and MRI suits for early detection of RA. As MRI is not affordable and US is more reliable, when the disease is more active, we can also prefer CR with some more improvisation in processing methodologies.
2.12. AUTOMATED SEGMENTATION AND QUANTIFICATION OF INFALAMMATORY TISSUE OF THE HAND IN RHEUMATOID ARTHRITIS PATIENTS USING MAGENTIC RESONANCE IMAGING DATA
Arthur: Evanthia E. Tripoliti , Dimitrios I. Fotiadis, Maria Argyropoulou ,
In this paper an automated method of segmentation and quantification of inflammatory tissue of the hand in patients suffering from rheumatoid arthritis is done using contrasting enhanced TI - weighted MRI. Preprocessing is of images is done. Cluster identification can be done by using minimizing the appropriate validity index. Segmentation is done using fuzzy c means algorithm. This method performs very well and results in high detection and quantification accuracy.
2.13. RETROSPECTIVE CORRECTION OF THE HEEL EFFECT IN HAND RADIOGRAPHS
Arthur : Gert Behiels, Frederik Maes1, Dirk Vandermeulen, Paul Suetens
In this paper a method for the retrospective correction of intensity in homogeneities induced by the heel effect in digital radiographs is presented. This method is based on a mathematical model for the heel effect derived from the acquisition geometry. The method iterates between background segmentation and heel effect correction until convergence.
OERVIEW OF THE PROJECT
3.1 AIM OF THE PROJECT
The main aim of the project is to detect Rheumatoid Arthritis in early stage by detecting the bone erosion by automatically segmenting the hand X-ray image using various algorithms.
Hand radiographic images from RA patients of different age groups and some of the age matched normal patients has been collected and segmented using automatic segmenting algorithms. For first phase, segmentation is done with snake algorithm. Segmentation is done by various algorithms, and the results were compared to detect which algorithm suits the best for detecting the bone erosion in RA patients in the early stages.
3.3 MATERIALS REQUIRED
Hand xray radiographs of both normal and a RA patients.
Image segmentation algorithm: snake algorithm.
Software used for processing: MATLAB.
3.4.1 IMAGE SEGMENTATION
Segmentation is mainly done to simplify the representation of an image for easier analysis. It is used to extract the set of contours from the image and to cluster pixels into salient image regions. Image segmentation is vital tool in the field of medical diagnosis
3.4.2 SNAKE ALGORITHM
Snake algorithm is a feature extraction technique, also known as active contours. It works like stretched Elastic Band being released. The Initial Points are defined around feature to be extracted. The points then move through an iterative process to move to a point with the lowest energy function.
It is explicitly defined as an approximation of an ellipse and a predefined number of points are generated. "Energy Function" for each point in the Local Neighbourhood is calculated. Points are moved with the lowest energy function, and it is repeated for every point. Iteration is done until it meets the termination condition.
3.4.3 DETERMINATION OF INITIAL CONTOURS
Determination of initial contours has to be done by removing the unexposed background by applying various levels of threshold. Adaptive threshold method is applied. Thresholding is used to segment an image by setting all pixels whose intensity values are above a threshold to a foreground value and all the remaining pixels to a background value. Adaptive thresholding will perform a binary thresholding i.e, it creates black and white image. In this algorithm each pixel is analyzed with respect to its local neighborhood and considered in more adaptive environment.
3.4.4 ACTIVE CONTOURS
The active contour model or snakes is defined as an energy minimizing spline, the snakes energy depends on its shape and location within the image. Snakes may be understood as a special case of a more general technique of matching a deformable model to an image by means of energy minimization. They depend on other mechanisms such as interaction with a user, interaction with some higher level image understanding process. The snake is active , always minimizing its energy functional, therefore exhibiting dynamic behavior.
The energy functional which is minimized is a weighted combination of internal and external forces. The internal forces emanate from the shape of the snake, while the external forces come from the image or from higher-level image understanding processes.
Snake growing can be done based on previous knowledge, and by estimating the desired contour position and curve. This curve is used to initialize the conventional snake algorithm, this yields a contour, for elimination of the high energy segments, segmentation of the contour is done. Which results in a number of initial shorter low energy contour segments, each segments is allowed to grow in the directions of the tangents, this yields a new estimate for each contour segments. For each contour segment, run the conventional snake algorithm using new estimate as an initial estimate to get a new contour segment.
3.4.5 READING OF XRAY FILMS TO CHECK FOR RA
There are several things to look for when reading an X-ray for rheumatoid arthritis.
Doctors use X-rays and blood tests to diagnose rheumatoid arthritis, an autoimmune disorder that causes chronic inflammation and damage to the joints. While X-rays alone are not usually sufficient to diagnose rheumatoid arthritis, they can provide doctors with good clues so they can order blood tests to confirm the diagnosis. Reading an X-ray for rheumatoid arthritis involves looking at the joints and surrounding tissues to determine if swelling and deterioration are present.
The spaces between the joints in hands or feet are checked, if the space between the bones is bigger or smaller than the space on a previous X-ray. The swelling in the joints is indicated by increased joint space, it is a common symptom of rheumatoid arthritis.
The color of the white area on the X-ray has to be examined. The white areas showing your bones should be even through each bone. Areas that vary in intensity can indicate eroding or deteriorating bones. Bone abnormalities and deformities have to be checked. Advanced stages of rheumatoid arthritis can cause extensive deformities, especially in the wrists and hands. Current X-rays has to be compared with previous ones to check for bone deformities and examine the differences between X-rays of each hand to spot any differences.
3.5 RESULTS AND ANALYSIS
A hand radiograph of rheumatoid arthritis patients is processed in MATLAB. Adaptive snake algorithm is applied to the digitized x-ray images. Adaptive threshold is applied and initial contours have been determined. Then snake algorithm is applied and thousand iterations have been done to extract the particular area of interest. Our aim is segmenting the eroded region can be achieved by further iterations.
CONCLUSION AND FUTURE WORK
The snake algorithm is applied to the hand radiographic images and background is eliminated using adaptive thresholding method. Bone erosions presented in the bone can be segmented by using snake algorithm. Initial contours and active contours are used to extract the features or regions of interest. Thus the bone erosions can be detected from the x ray images of the patients affected by RA.
4.2 FUTURE DIRECTIONS
In the extension of this project, two more algorithms are used for segmenting abnormal and normal images and all the three algorithms are compared for better result as which detects the bone erosion faster and in early stage.