Artifacts Produced on the Images During CT Imaging - Term Paper Example

Artifacts During CT Imaging Introduction Artifacts are defined as “as any structure that is seen on an image but is not representative of the actual anatomy” (Goldman, 2007, 222). In other words, artifact is “any systematic discrepancy between the CT numbers in the reconstructed image and the true attenuation coefficients of the object” (Barrett, J., and Keat, 2004, 1680). CT imaging is more prone for artifacts when compared to conventional radiography because "the image is reconstructed from something on the order of a million independent detector measurements. The reconstruction technique assumes that all these measurements are consistent, so any error of measurement will usually reflect itself as an error in the reconstructed image (Barrett, and Keat, 2004, 1680). Artifacts have a major role to play in diagnostic accuracy because their presence can degrade the quality of image of computed tomography or CT scan. In this article, various types of CT artifacts and strategies to avoid or minimize them will be discussed. Types of artifacts Artifacts can be basically classified into four major categories and they are (Yazdi and Beaulieu, 2008, 135): 1. Patient based artifacts like metallic artifacts and motion artifacts. 2. Physics based artifacts like photon starvation, beam hardening and undersampling artifacts. 3. Spiral based artifacts which arise because of spiral interpolation. 4. Scanner based artifacts which occur due to mechanical instability and sensitivity of the detectors. Based on their appearance, there are several categories of CT artifacts and they are shading artifacts, streak artifacts,ring artifacts (Goldman, 2007, 222), distortion (Barrett, and Keat, 2004, 1680), slab artifacts, step artifacts and blurring artifacts (Siemens, 2). Shading artifacts Beam hardening effects is the most common shading artifact. they are due to imperfect correction of beam hardening and occur in almost all CT images. They appear as "nonuniformities in the CT numbers of a uniform material, such as CT numbers that are lower at the center of a uniform phantom than at the periphery" (Goldman, 2007, 222). The nonuniformities are most of the time less than 5HU and usually unapparent unless viewed through a narrow window. larger hardening occurs if the scan is passing through a contrast medium or thick bone regions. Shading artifacts can also occur due to scattering. But scattering is uncommon in recent scanners. Fig.1. Beam-hardening artifact caused by unusually severe hardening of x-rays passing though thick bone (Goldman, 35, 215) Streak artifacts Streak artifacts are very common and are encountered in almost all scanners. they occur mainly due to bad detector measurements or inconsistent measurements. Such inconsistencies occur due to metals, partial volume effects, motion, insufficient intensity of X-ray and malfunctioning of the tube arc (Goldman, 2007, 222). (Goldman, 2007, 222). Figure-2. Sreak artifacts (Pillow, 2009). Ring artifacts Figure-3: Ring artifact (Dawes, 2008) Ring artifacts or partial ring or arc artifacts occur due to drift in calibration, errors, measurement inaccuracies and imbalances. The current generation CT scanners have corrective algorithms and are able to detect inaccuracies in measurements (Goldman, 2007, 222). Distortion artifacts Distortion artifacts occur due to reconstruction helically (Barrett, and Keat, 2004, 1680). Figure-4. Distortion artifacts (www.googleimages.com) Metallic artifacts Metallic artifacts are very common and occur due to metallic implants like dental fillings, surgical clips and various other prosthesis (Fig). Metallic artifacts cause streaking in the CT imaging. Chemotherapy ports can also cause metallic artifacts. Standard reconstruction methods like filtered back projection are used to reconstruct the matrix and this causes streaks which are bright and dark. The main reason why streaking appears is "inaccurate beam hardening correction" in the filtered back projection. Newer CT scan machines are infact equipped with techniques to correct the inaccurate matrix reconstruction of body organs. However, these advanced techniques yet exclude objects with high attenuation like metals. Thus, metallic artifacts degrade the quality of imaging, thus making it difficult to plan effective radiation therapy (Yazdi and Beaulieu, 2008,135- 136). Metallic artifacts cause streaking because they absorp photons heavily. If PET is used, the increase in photons leads to increased PET attenuation coefficients contributing to over estimation of the activity of PET in that region, leading to false-positive findings. Hence when metals ae present, usage of non-attenuated PET will prevent this error. Some implants like hip prosthesis have properties of attenuating PET 511-keV photons contributing to absent emission data in hat region. Application of CT attenuation correction factors to this region of photopenic emission which show diminished PET images. Hence technologists must advise the patients to remove all metallic objects in their body and documentation of the location of any unremovable metallic object must be done in order to identify and minimize artifacts in that region. (Sureshbabu and Mawlaw, 2005, 157-158). Several techniques have been developed to reduce metallic artifacts. Techniques which work on matrix are the most effective strategies and there are 2 methods with reference to this technique and they are iterative reconstruction methods and projection-interpolation based methods. In the former methods, data pertaining to projection associated with metallic substances in the projection matrix are disregarded to a large extent and reconstruction of the matrix is done by using only data which is non-corrupted. Though these methods are useful and speciafically suitable for noisy projection data and incomplete projection data, they ned to handle problems associated with convergence and thus are expensive for routine CT Scanning in a clinical setting. On the other hand, projection-interpolation based methods reconstruct by identifying missing projections manually and replacing them with non-mising projections of neighborhood. Some researchers replace msssing projections by using either linear prediction method, polynomial interpolation technique or wavelet multiresolution analysis. Besides the above described methods, another novel strategy has been developed by Mahaken et al (cited in Yazdi and Beaulieu, 2008) to compute interpolation valueby calculating "the sum of weighted nearest not-affected projection values within a window centred by the missing projection" (Yazdi and Beaulieu, 2008). In this method of computing, only distance data is used to model weights and non-affected projections projections pertaining to all directions is used to ascertain replacement values. However, continuity of the projection structures is not preserved in this strategy. There is also increased risk of noise production because of lack in continuity of the projection structures (Yazdi and Beaulieu, 2008, 136). Fig.5. (A) High-density metallic implants generate streaking artifacts and high CT numbers (arrow) on CT image. (B) High CT numbers will then be mapped to high PET attenuation coefficients, leading to overestimation of activity concentration. (C) PET images without attenuation correction help to rule out metal-induced artifacts. (Sureshbabu and Mawlawi, 2005) Photon Starvation artifact Photon starvation also lead to artifacts and they appear as streaks. Such artifacts are mainly noted near parts of body with high tissue volume. examples of such regions are hip, heart and shoulder. People with mass body have increased risk f phton starvation artifacts seen in their CT Scans. The main reason why artifacts arise is because of insufficient passage of photons through the wide parts of the body leading to noisy individual projection. Reconstruction of missing projects by standard algorithm available in commercial scanners leads to magnification of noise which appear as streaks "in the direction of widest part" (Yazdi and Beaulieu, 2008). Photon starvation artifacts can be prevented by balancing the number photons that are received by detectors in all directions. One strategy to balance photon reception is to "increase of photon flux (by increasing current (mA) through the scanner tube) through widest parts without changing the photon flux through narrower parts" (Yazdi and Beaulieu, 2008). Thus balancing of photons can be done. Another strategy to prevent photon starvation artifacts is by filtering projections in an adaptive manner. It is very esssential to use good filter parameters and the filters must be applied in the right place, in order to improve the quality of imaging (Yazdi and Beaulieu, 2008, 136- 137). Figure-6: Example of photon starvation artifact (Yazdi and Beaulieu, 2008,137) Motion artifacts In any CT scanning, patients move and this movement leads to some artifacts which appear as blurring or streaks in the image. Infact, the most prevalent artifacts in CT/PET scanning are respiratory motion artifacts. The movements can be simple like breathing, or can occur due to voluntary motion, or due to involuntary heart beat. Movement is common in children and those who are injured. Breathing artifacts can be avoided by asking the patients to hold their breath and spiral CT scans can take images in such small duration of time. These artifacts occur due to discrepancies in the positions of the chest in the CT image and the positions of the chest in the PET image. The acquisition time of PET imaging is long and hence the image is acquired when the patient is breathing freely. Thus, the final image is developed after taking the average of repeated images taken in different breathing cycles. But, the CT scan image is taken during a specified stage in the cycle of breathing. Thus breathing artifacts ensue (Sureshbabu and Mawlaw, 2005, 157-158). Patients must be asked to hold their breath in mid-inspiration or mid-expiration phase scanning (Sureshbabu and Mawlaw, 2005, 157-158). If breath holding is not possible, patients must be asked to breathe in a shallow manner (Sureshbabu and Mawlaw, 2005, 157-158). The breathing artifact which is the most common causes curvilinear cold areas. This type of artifact is seen when scanning is done in full-inspiration phase resulting in displacement of the diaphragm downwards due to expansion of the lungs (Gormally et al, 2007, 699). Such a downwards displacement causes underestimation of the coefficients of the CT attenuation because of more representation of air than soft tissue. Such artifacts have a great impact on the determination and understanding of the lesions in those with already proven lesions in the lungs. The artifact is commonly referred to as 'misregulation of lesions'. Another reason for mismatching between CT scans and PRT imaging is correction values which are inaccurate. Such mismatch values lead to erroneous standardized uptake data in PET-imaging. hence, it is very important to advise patients the techniques of breath-holding before entering the scan room (Sureshbabu and Mawlaw, 2005, 158- 159). Figure-7: Respiratory motion artifact (Zaldi, 2007, e37) Artifacts due to cardiac motion is a major problem because cardiac motion cannot be prevented. Recent CT scanners are incorporated with ECG gating which synchronizes acquisition of data with heart beat rhythm. The images of the heart are acquired by starting image acquisition in synchrony with the beginning of the cardiac cycle. other than these strategies, correction techniques can also be used to minimize motion artifacts. One such correction strategy is to correction of in-plane motion in the coordinated system by pixel-specific reconstruction. Another strategy is to overscan the cardiac region and remove the heart motion effects by repeating projections (Yazdi and Beaulieu, 2008, 137). Figure-8: Cardiac motion artifact (www.googleimages.com) Fig. 9: Example of artifacts produced by scanning a patient with two hip prostheses using a Siemens Somatom scanner, Hotel-Dieu Hospital Center, Quebec, Canada. Spiral artifacts Though the types of artifacts produced in conventional scanning and spiral scanning are similar, spiral scanning has more chances of development of additional artifacts because of the need for interpolation process for recovery of projections that are consistent and belong to individual slices. However, the severity of the artifacts and their appearance depends basically on the pitch of the scanning and the algorithm type used for interpolation (Yazdi and Beaulieu, 2008, 137- 138). Figure-10: Spiral CT artifact (www.googleimages.com) Contrast media artifacts It is very common to use oral or intravenous contrast media to improve the diagnostic value to CT scanning. The most commonly used contrast agents are either iodine based solutions or barium sulfate. These agents delineate the soft tissues and blood vessels and thus enhance the images of the CT. Though these agents improve the quality of CT images, they have a profound effects on the quality and quantitative accuracy of metallic implants. When high concentrations of contrast is used, high numbers of CT occurs leading to increased number of streaking artifacts because of absorption of photons. This in turn leads to high PET attenuation coefficients causing overestimation of tracer uptake and false positive results in PET imaging. Thus, the severity of of artifact on PET images attenuated with CT depends on the contrast agent concentration, the clearance and distribution of the agent and the time interval between administration of the contrast agent and acquisition of the CT. Increase in the concentration of the oral contrast agent occurs with increase in time because of increased water absorption. Consequently, the residual barium that has high CT numbers over corrects the data of the PET emission, mimicking and increase in the uptake of tracer, resulting in the affectation of the PET scan interpretation. This is one of the major reasons why patients who have undergone CT with oral contrast must not be subjected to PET imaging for 2 days. On the other hand, intravenous contrast administration does not affect the PET images because of high dilution and faster clearance (Sureshbabu and Mawlaw, 2005, 159- 160). Figure-11: Contrast media artifacts (www.googleimages.com) Truncation These artifacts in PET/CT scans occur due to various differences which occur in the field of view sizes between the 2 tomographs. The artifacts are usually seen in patients who are large or who have been scanned with their hands in down position for indications in head and neck and for melanoma scanning. When there is extension of the patient beyond the CT field of view, the ended anatomy part is truncated and hence is not represented in the image. Hence no attenuation correction occurs in the emission data of the PET resulting in a bias on the PET-attenuation-corrected images that underestimate the uptake of values that are standardized in these regions. Even truncation artifacts appear as streaks at the edge of the images causing an overestimation of coefficients of attenuation to correct the data of PET. This results in high activity at the edge causing misinterpretation of the scan. Thus, careful positioning of the hands is essential to reduce artifacts due to truncation (Sureshbabu and Mawlaw, 2005, 160). Fig. 12: Truncation artifact (www.googleimages.com) Partial volume effect This type of artifact is seen in 360 degree axial scan in which the same ray is sample two times but in the opposite direction. However, the beam divergence makes beam samples travel in different directions. At the edge of the bone or any such small structure may cause attenuation of beam traveling in one direction and can get missed when coming from opposite direction. this inconsistency can cause streaking in images (Goldman, 2007, 222). Figure-11: Partial-volume streaks are caused by opposing x-ray beams, which nominally pass through the same voxels but actually sample slightly different cone-shaped tissue volumes as a result of beam divergence. Small structure, such as piece of bone, is detected by beam from one direction but is missed by opposing beam. Resulting inconsistency leads to streak artifact (Goldman, 2007, 222). Out-of view artifacts These artifacts occur "by anatomy that is out side of the selected scan field of view". To avoid these artifacts, the field of the scanner must be larger than the scanned object (Luo, Texas). Figure-13. Out-of-view artifact (Luo, Texas). Edge gradient artifacts These occur when "the edges of a "sharp" high density object interface with a smooth surface. They frequently occur from the edges between bone and soft tissue" (Luo, Texas). Such artifacts can be avoided by using high frequency, sharp algorithm during mathematical reconstruction. Figure-14: Edge gradient artifact (www.googleimages.com) Conclusion CT artifacts are common both in conventional and spiral CT scans. there are many sources for the development of artifacts. Artifacts impact the image quality of the CT scan and also reduce the quality of diagnosis. usage of newer designs in the technology of scanners, appropriate positioning of the patients during the process of scanning and optimum selection of parameters of scanning help in the reduction of development of artifacts. Certain software tools also can help in the reduction of artifacts. References Barrett, J., and Nicholas Keat. 2004. Artifacts in CT: Recognition and Avoidance. RadioGraphics 24:1679–1691. Cardiac artifacts. n.d. Siemens Notes. Dawes, L. 2008. Ring artefact. Retrived on 14th August 2010 from Radiopedia.org.http://radiopaedia.org/articles/ring_artefact Gormally, J., D.A.T. Gay, N.E. Manghat and M.P. 2007. A new CT artefact - the bubble curve sign. Clinical Radiology 62, 699- 702. Goldman, L.W. 2007. Principles of CT: Radiation Dose and Image Quality. J Nucl Med Technol 35: 213–225 Luo, Q. n.d. Artifacts in X-ray CT. Research Imaging Center, University of Texas Health Science Center, TX 78229. Pillow, M.T., Robert Mulliken and Christopher Straus. 2009. Emergency Neuroradiology. Emedicine from WebMD. Retrieved on 14th August 2010 from http://emedicine.medscape.com/article/810904-overview Sureshbabu, W. and Osama Mawlawi. 2005. PET/CT Imaging Artifacts. J Nucl Med Technol 33:156–161 Yazdi, M., and Luc Beaulieu. 2008. Artifacts in Spiral X-ray CT Scanners: Problems and Solutions. International Journal of Biological and Life Sciences 4 (3:, 135- 139. Zaldi, H. 2007. Optimisation of whole-body PET/CT scanning protocols. Biomed Imaging Interv J., 3(2):e36 Read More