Noise Patterns in industrial Micro Computed Tomography


Research Paper (postgraduate), 2010

51 Pages, Grade: 1.4


Excerpt

Table of contents

Table of Figures

Table of Tables

1 Introduction
1.1 The Natural History Museum London (NHM)
1.1.1 The beginning of the British Museum (NHM London)
1.1.2 NHM in the 21st Century
1.2 Computed Tomography(CT) - Introduction
1.2.1 Industrial Computed Tomography
1.2.2 Micro-CT
1.2.3 ImageJ
1.3 Noise/Artefacts
1.3.1 Artefacts
1.3.2 Noise
1.4 Filters
1.4.1 Reasons for using a digital filter
1.4.2 High-passFilters
1.4.3 Low-passFilters
1.4.4 Kalman Stack Filter
1.4.5 Gaussian blurFilter
1.4.6 Kuwahara Filter

2 Project Assignment: Analysing noise pattern in Micro-CT
2.1 MaterialsandMethods
2.1.1 Metrls X-Tek HMX ST 225 CT System
2.1.2 Phantom Design
2.1.3 ImageJ plug-in Code
2.2 Description ofproject objectives: Parti, Part II and Part III
2.2.1 Part I: How do scan parameters affect noise?
2.2.2 Part II: Whatis the best type of noise reduction algorithm for CT?
2.2.3 Part III: How current, exposure and noise affect each other?
2.3 Summary and Analysis
2.3.1 Analysing pattern of the noise variation
2.3.2 Optmising parameters for noise reduction
2.3.3 Analysing current vs. exposure effect

3 Conclusionandfuturework
3.1 Insightfrom experiments
3.2 Insights from post-processing
3.3 Suggestionsforfuturework:

AppendixA

Plug-in Code

How to use the Plug-in

References

Table ofFigures

FIGURE 1: PORTRAIT OF HANS SLOANE

FIGURE 2: MONTAGU HOUSE, BLOOMSBURY (HUBPAGES LONDON 2010)

FIGURE 3: THE NATURAL HISTORY MUSEUM, LONDON (WIKIPEDIA NATURAL HISTORY MUSEUM 2010)

FIGURE 4: CLASSIFICATION OF NOISE

FIGURE 5: METAL ARTEFACTS

FIGURE 6: MECHANISM OF PARTIAL VOLUME ARTIFACTS, WHICH OCCUR WHEN A DENSE OBJECT LYING OFF-CENTER PROTRUDES PART OF THE WAY INTO THE X-RAY BEAM (KEAT 2004)

FIGURE 7: CT IMAGES OF THREE 12-MM-DIAMETER ACRYLIC RODS SUPPORTED IN AIR PARALLEL TO AND APPROXIMATELY 15 CM FROM THE SCANNER AXIS. IMAGE OBTAINED WITH THE RODS PARTIALLY INTRUDED INTO THE SECTION WIDTH SHOWS PARTIAL VOLUME ARTIFACTS

FIGURE 8: MATRIX X-TEK SYSTEM (NHM KEIN DATUM)

FIGURE 9: CONE BEAM CONFIGURATION (GEOCHEMICAL INSTRUMENTATION AND ANALYSIS 2010)

FIGURE 10: PHANTOM IN VGSTUDIO MAX 2.1

FIGURE 11: PLASTIC FLASK USED IN THE CONSTRUCTION OF THE PHANTOM

FIGURE 12: THE ABOVE FIGURE SHOWS RELATIONSHIP BETWEEN VOLTAGE, CURRENT, EXPOSURE DURATION AND NOISE. THE ARROW IN THIS FIGURE INDICATES THE NOISE

FIGURE 13: OPTIMISING MICRO-CT. HIGHLIGHTED IN YELLOW SHOWS THE RESULT WITH LOWEST VALUE OF STANDARD DEVIATION. HIGHLIGHTED IN ORANGE SHOWS THE RESULT WITH LOWEST VALUE WITHOUT BEAM HARDENING AND NOISE REDUCTION 2

FIGURE 14: EFFECT OF FILTERS

FIGURE 15: KALMAN STACK FILTER GIVES BEST RESULTS (LOW STANDARD DEVIATION)

FIGURE 16: EFFECT OF CURRENT AND EXPOSURE ON NOISE. THE NUMBER INSIDE THE BUBBLES INDICATE THE STANDARD DEVIATION VALUE

FIGURE 18: SD VS RATIO OF EXPOSURE TO CURRENT (MS/MA) BETWEEN 708 AND 2000 MS EXPOSURE. THE TOP NUMBERS IN EACH REPRESENT THE CURRENT VALUE AND EXPOSURE TIME. THE BOTTOM NUMBER SHOWS THE RATIO / STANDARD DEVIATION

FIGURE 19: SD VS EXPOSURE TIME, NO TREND BETWEEN EXPOSURE TIME AND STANDARD DEVIATION

FIGURE 20: TIME = SCANNING TIME PLOTTED AGAINST SD

FIGURE 21: SET OF SCANS WITH MO TARGET AND DIFFERENT VOLTAGE, CURRENT AND EXPOSURE. FIRST NUMBER IS VOLTAGE (KV) SECOND IS CURRENT (MA), NEXT IS EXPOSURE (MS) AND TARGET (MO)

FIGURE 22: SELECTING THE REGION OF INTEREST

FIGURE 23: MANUAL MODIFICATION OF THE PARAMETERS CAN BE DONE IF NEEDED

FIGURE 24: RESULTS FROM THE SCAN

Table ofTables

TABLE 1: PATTERN OF NOISE VARIATION

TABLE 2: EFFECT OF FILTER

TABLE 3: COMPARISON OF THE THREE FILTERS WITH VARIATION IN PARAMETERS

TABLE 4: RESULTS OF NOISE REDUCTION FILTERS. THE BEST RESULTS ARE HIGHLIGHTED

TABLE 5: COMPARISON BETWEEN CTPRO NOISE REDUCTION ALGORITHM AND KALMAN FILTER

1 Introduction

The first section introduces the background and then the recent developments of the Natural History Museum, London. This is then followed by a brief description of Computer Tomography (CT), covering basic concepts of this technology with emphasis on industrial CT and Micro-CT.

The difference between noise and artefacts are discussed in detail and the problems faced - during this project - in the operation of computer tomography are explained. Additionally, fundamental principles of low and high pass filters for digital signal processing are covered.

1.1 The Natural History Museum London (NHM)

The Natural History Museum (NHM) is located in one of the biggest Natural History Museums in the (for the general public) and actively researches on:

South Kensington, London and is world. The museum both exhibits

- animals
- plants
- ecosystems
- geology
- palaeontology
- climatology(Wikipedia Natural History Museum)

1.1.1 The beginning of the British Museum (NHM London)

The NHM started with a young English boy named Hans Sloane who was interested in Botany. Sloane during his education in medicine spent some of his time in Europe. On his way to Paris he met the chemist Nicolas Lemery; in the French capital he visited the Jardin du Roi and frequented the Charite hospital. He also heard lectures on botany and anatomy. After completion of his degree, Sloan soon became famous as a physician as well as a botanist.

illustration not visible in this excerpt

Figure 1: Portrait of Hans Sloane

When Christopher Monck, second duke of Albemarle, was appointed governor of Jamaica, he appointed Sloane to accompany him to the island as his personal physician. The expedition was of great value to Sloane, not only giving him first-hand experience of a relatively little-known island but also enabling him to search for new drugs. The description of the voyage and the observations on the inhabitants, diseases, plants, animals, some of which he brought back alive, and meteorology of the West Indies make Sloane’s book on Jamaica valuable even today. The duchess of Albemarle was suffering from the first stages of the mental illness which turned to madness later. As a result the duke became an alcoholic and died within a year of arriving in Jamaica. Sloane escorted the duchess back to England in 1689. He brought with him collections of plants, animals, fossils, minerals and earth and a large quantity of note and drawings (Thackray and Press).

His house, now 3 Bloomsbury Place W.C.I, where he lived from 1695 until 1742, soon became so full that he was obliged to rent the adjoining house, No. 4, as well. In 1712 he bought the manor house at Chelsea, although he did not retire to live there until 1742 when all his collections followed him.

From the early days of the century Sloane’s museum had been an object of interest not only to British scholars and men of science but also to foreigners, many of whom were urged not to leave the country without seeing it. Sloane’s doors were always opened to the visitors. He always wanted the public to benefit from his collection(Thackray and Press).

Sloane died on 11 January 1753. In his will he offered all these treasures to the British nation, on condition that £20 000 was paid to his daughters, who might have expected to inherit the valuable collection. After Sloane’s death on January 11, 1753 the trustees whom he had appointed met; the matter was brought before parliament, which received Royal Assent for the Act which enabled Sloane’s collection to be acquired and a suitable building purchased to house it. King George II was one of the interested parties mentioned in Sloane’s will. The King decided to buy some other collections as well. On the 7th June 1753 King George II established the Sloane, Cotton and Harleian collections as the British Museum for general use and benefit of the public.

illustration not visible in this excerpt

Figure 2: Montagu House, Bloomsbury (Hubpages London)

On Monday the 15th of January 1759 the door of the British Museum were opened for the public. Many donations were made during the years and the British Museum in Bloomsbury become crowded with specimens. A decision to separate the Natural History collection from the rest was made. In 1880 new building in South Kensington was completed, where the departments of Mineralogy, Botany and Geology were move (Thackray and Press).

On the 18th of April 1881, the museum was officially named “British Museum “(Natural history)”. In 1992 the name was changed to The Natural History Museum because of the confusion as many visitors used to come expecting to see Egyptian mummies and other artefacts of ancient civilization (Thackray and Press).

The NHM has always been at the forefront of newest technologies and their applications such as for analyzing specimens. For example, the first electron microscope was constructed in the 1939s and the Museum purchased its first model, a transmission electron microscope, in 1965 (Thackray and Press).

1.1.2 NHMinthe21st Century

The Museum is an exempt charity and a non-departmental public body sponsored by the Department of Culture, Media and Sport.

The museum contains more than 70 million specimens and has a library with very rare books - with great historical and scientific value. The museum’s staff stands at 843, of whom approximately 315 are scientists that devote the majority of their time to the collections and research.

illustration not visible in this excerpt

Figure 3: The Natural History Museum, London (Wikipedia Natural History Museum)

The Museum is also a world-renowned centre of research, specialising in taxonomy, identification and conservation of specimens (Wikipedia Natural History Museum).

1.2 Computed Tomography(CT] - Introduction

Academia and users refer to computer tomography by many names:

- Computerized axial transverse scanning
- Computerized axial tomography (CAT)
- X-ray computed tomography (X-ray CT)
- Computed/computerized tomography (CT) (Karthikeyan and Deepa)

For discussions in this work, computed tomography (CT) will be used.

Computed Tomography is a widely used imaging technique, with special emphasis on medical applications. Tomography uses the radiographic images of the object obtained from different angles. Using an algorithm, called filter back projection, it is possible to reconstruct a virtual slice through the object. When different consecutive slices are reconstructed a 3D visualization can be obtained. This 3D view inside the body or specimen is extremely useful for diagnosing or for research.

The basic idea behind CT is that the internal structure of an object can be reconstructed from multiple projections of the object without destroying it.

The primary elements of X-ray tomography are:

- An X-ray source
- A series of detectors that measure X-ray intensity attenuation along multiple beam paths,and
- A rotational geometry with respect to the object being imaged

Different configurations of these components can be used to create CT scanners optimized for imaging objects of various sizes and compositions.

The great majority of CT systems use X-ray tubes, although tomography can also be done using a synchrotron or gamma-ray emitter as a monochromatic X-ray source. Important tube characteristics are the target material and peak X-ray energy, which determine the X-ray spectrum that is generated; current, which determines X-ray intensity; and the focal spot size, which impacts spatial resolution.

Most CT X-ray detectors utilize scintillators. The parameters that are usually varied are the scintillator material, size and geometry, and the means by which scintillation events are detected and counted. In general, smaller detectors provide better image resolution, but reduced count rates because of their reduced area compared to larger ones. To compensate, longer acquisition times are used to reduce noise levels. Common scintillation materials are cesium iodide, gadolinium oxysulfide, and sodium metatungstate.

1.2.1 Industrial Computed Tomography

The primary difference between Industrial CT and Medical CT is that in Medical CT the object of interest is generally between 1.5 m and 2 m and the composition is of circa 68% water with less heavy elements. However, in the industrial CT the variation of compounds is large because it is used for scanning of industrial components. The scans can vary from millimeters to meters. Additionally, the material can be homogeneous or non-homogeneous materials. All this factors make the Industrial CT-scanner more challenging however, the technology is improving very rapidly.

1.2.2 Micro-CT

Micro computed tomography (micro-CT) is primarily the same as standard CT except it uses a micro focus tube instead of a traditional tube and generally high X-ray energy and also longer exposure time. A micro-CT scan yields resolutions in microns because the focal spot of a micro focus tube is only a few microns in size. For comparison, micro-CT resolution is about 100 times better than the best CAT scan in the medical field.

1.2.3 ImageJ

ImageJ (Image Processing and Analysis in Java) is a public domain image processing program inspired by the NIH (National Institutes for Health). It can display, analyze, edit and process 8, 16, and 32-bit images and can read file formats such as TIFF and DICOM. It supports a series of images that share a single window; it can calculate distances and perform geometric transformations such as scaling, rotation and flips. Finally, it can be used as a tool to crop and measure images (Image J).

1.3 Noise/Artefacts

The foundations of imaging system performance and image quality can be traced back to the pioneering work of Albert Rose (U.S. National Library of Medicine). He showed that image quality is fundamentally limited by the statistical fluctuations in image quanta. Hence, the more image quanta used to create an image, the better the image quality.

The technology to produce images has improved dramatically during the last decades, but still researchers struggle with artefacts and noise in the image quality. Broadly speaking artefacts is an inherent property i.e. dependent on the X-ray system, whereas noise is an external influence - property that is affected by outside the X-ray system.

illustration not visible in this excerpt

Figure 4: Classification of noise

In the Figure 4 you can see in different colours, the classification of noise. To understand better what effect the different classifications could have on the quality of the data a closer look into each of the elements classified is essential.

1.3.1 Artefacts

An artefact is any error in the perception or representation of any information introduced by the involved equipment or techniques. Only the types of artefacts, which are relevant to this project assignment, are mentioned in detail.

1.3.1.1 MotionArtefacts

These artefacts are introduced by movement of the specimen. Even the slightest of movements of the specimen could produce motion artefacts. In experiments, sometimes this can be a difficult challenge as it may not possible to place a specimen so that it is very stable thus producing artefacts.

To contain this artefact a shorter scanning time should be chosen. Short scanning time could be achieved by for example using a shorter exposure time.

1.3.1.2 RingArtefacts

This artefact appears due to mis-calibration of one or more detector elements in the CT- Scanner. In this case a new calibration often helps.

1.3.1.3 MetalArtefacts

The presence of metal objects in the scan field can lead to severe artefacts known as streaking. Thisoccurs because the density of the metal is beyond the normal range that can be handled by the current computers, resulting in incomplete profiles. Streaking could be greatly reduced by use of special corrections software.

illustration not visible in this excerpt

Figure 5: Metal artefacts

1.3.1.4 BeamHardeningArtefacts

Beam hardening is a selective removal of soft X-rays from the X-ray beam. Usually a Gaussian filter is used for removing beam hardening. After removing of soft X-rays the beam becomes more penetrated. The amount of beam hardening required depends on the initial X-ray spectrum as well as on the composition of the material, which is penetrated. If no correction is made, beam hardening artefacts appear as cupping, or a reduction of the reconstructed attenuation coefficient toward the centre of the specimen (University of Texas).

1.3.1.5 PartialVolumeArtefacts

There are a number of ways in which the partial volume effect can lead to image artifacts. One type of partial volume artifact occurs when a dense object lying off-center protrudes partway into the width of the x-ray beam. The divergence of the x-ray beam as shown in Figure 6 along the z axis has been greatly exaggerated to demonstrate how such an off-axis object can be within the beam, and therefore “seen” by the detectors, when the tube is pointing from left to right but outside the beam, and therefore not seen by the detectors, when the tube is pointing from right to left. The inconsistencies between the views cause shading artifacts to appear in the image Figure 7.

illustration not visible in this excerpt

Figure 6: Mechanism of partial volume artifacts, which occur when a dense object lying off-center protrudes part of the way into the x-ray beam(Keat).

Essentially two materials are present in a voxel. Consequently the grey value is an average based on the proportions of the material.

illustration not visible in this excerpt

Figure 7: CT images of three 12-mm-diameter acrylic rods supported in air parallel to and approximately 15 cm from the scanner axis. Image obtained with the rods partially intruded into the section width shows partial volume artifacts.

1.3.1.6 Other artefacts

A variety of other artefacts can arise in some situations: stair-step artefact, zebra artefact, helical artefact etc. These are not covered in detail as they are not relevant to this project.

1.3.2 Noise

Real world signals usually contain departures from the ideal signal. Such departures are referred to as noise.

Noise arises as a result of unmodelled or unmodellable processes going on in the production and capture of the real signal.

Noise can generally be grouped into two classes:

- independent noise
- noise which is dependent on the image data

One kind of noise which occurs in all recorded images to a certain extent is detector noise. This kind of noise is due to the discrete nature of radiation, i.e. the fact that each imaging system is recording an image by counting photons. A common form of noise is data drop-out noise (commonly referred to as intensity spikes, speckle or salt and pepper noise). Here, the noise is caused by errors in the data transmission.

The corrupted pixels are either set to the maximum value (which looks like snow in the image) or have single bits flipped over. In some cases, single pixels are set alternatively to zero or to the maximum value, giving the image a 'salt and pepper' like appearance. The noise is usually quantified by the percentage of pixels which are corrupted.

1.4 Filters

Filters in general are there to block something, improving image quality by reducing image noise. There are two types of filters: analogue and digital filters.

For this project only digital filters were studied and will be discussed in detail below.

1.4.1 Reasons for using a digital filter

There are three main reasons for using digital filters: Firstly, to separate signals that have been combined. Secondly, to restore signals those have been distorted in same way. Thirdly, to block a frequency or reduce the amplitude of a frequency.

Digital filters are mainly used to suppress:

- high frequencies in the image (i.e. smoothing the image)
- low frequencies (i.e. enhancing or detecting edges in the image)

An image can be filtered either in the frequency or in the spatial domain.

1.4.2 High-pass Filters

A high-pass filter, or HPF, passes high frequencies well but attenuates (i.e., reduces the amplitude of) frequencies lower than the filter's cutoff frequency. The actual amount of attenuation for each frequency is a design parameter of the filter. It is also sometimes referred to as a low-cut filter or bass-cut filter (Smith).

1.4.3 Low-pass Filters

This is the most common filter in noise reduction. In contrast to the high-pass filter a low-pass filter passes low-frequency signals but attenuates high-frequency signals. The amount of attenuation depends on the specific low pass filter with specific objectives for example smoothing and blurring of images (Smith).

This work will focus on some of the commonly used low pass filters: the Gaussian Blur filter, Kuwahara filter and Kalman Stack filter.

1.4.4 Kalman Stack Filter

Kalman filter is an optimal estimator or an optimal recursive data processing algorithm. It incorporates all information that can be provided to it. It processes all available measurements, regardless of their precision, to estimate the current value of the variables of interest, with use of knowledge of the system and measurement devices dynamics, the statistical description of the system noise, and uncertainty in the dynamics models, and any available information about initial conditions of the variables

[...]

Excerpt out of 51 pages

Details

Title
Noise Patterns in industrial Micro Computed Tomography
College
University of Applied Sciences North Rhine-Westphalia Paderborn
Grade
1.4
Author
Year
2010
Pages
51
Catalog Number
V165584
ISBN (eBook)
9783640837892
ISBN (Book)
9783640838561
File size
3739 KB
Language
English
Tags
CT, Noise, Rauschen
Quote paper
Galina Bernhardt (Author), 2010, Noise Patterns in industrial Micro Computed Tomography, Munich, GRIN Verlag, https://www.grin.com/document/165584

Comments

  • No comments yet.
Read the ebook
Title: Noise Patterns in industrial Micro Computed Tomography



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free