Editing CT scan (section)

== Interpretation of results ==

=== Presentation ===
[[File:CT presentation as thin slice, projection and volume rendering.jpg|thumb|upright=1.4|Types of presentations of CT scans: <br />− Average intensity projection<br />− [[Maximum intensity projection]]<br />− Thin slice ([[median plane]])<br />− [[Volume rendering]] by high and low threshold for [[radiodensity]]]]

The result of a CT scan is a volume of [[voxel]]s, which may be presented to a human observer by various methods, which broadly fit into the following categories:
*Slices (of varying thickness). Thin slice is generally regarded as planes representing a thickness of less than 3 [[Millimetre|mm]].<ref name="Goldman2008">{{Cite journal |last=Goldman |first=L. W. |year=2008 |title=Principles of CT: Multislice CT |journal=Journal of Nuclear Medicine Technology |volume=36 |issue=2 |pages=57–68 |doi=10.2967/jnmt.107.044826 |issn=0091-4916 |pmid=18483143 |doi-access=free}}</ref><ref name=":2">{{Cite journal |last1=Reis |first1=Eduardo Pontes |last2=Nascimento |first2=Felipe |last3=Aranha |first3=Mateus |last4=Mainetti Secol |first4=Fernando |last5=Machado |first5=Birajara |last6=Felix |first6=Marcelo |last7=Stein |first7=Anouk |last8=Amaro |first8=Edson |date=29 July 2020 |title=Brain Hemorrhage Extended (BHX): Bounding box extrapolation from thick to thin slice CT images v1.1 |journal=PhysioNet |volume=101 |issue=23 |pages=215–220 |doi=10.13026/9cft-hg92}}</ref> Thick slice is generally regarded as planes representing a thickness between 3&nbsp;mm and 5&nbsp;mm.<ref name=":2" /><ref>{{Cite journal |last1=Park |first1=S. |last2=Chu |first2=L.C. |last3=Hruban |first3=R.H. |last4=Vogelstein |first4=B. |last5=Kinzler |first5=K.W. |last6=Yuille |first6=A.L. |last7=Fouladi |first7=D.F. |last8=Shayesteh |first8=S. |last9=Ghandili |first9=S. |last10=Wolfgang |first10=C.L. |last11=Burkhart |first11=R. |last12=He |first12=J. |last13=Fishman |first13=E.K. |last14=Kawamoto |first14=S. |date=2020-09-01 |title=Differentiating autoimmune pancreatitis from pancreatic ductal adenocarcinoma with CT radiomics features |journal=Diagnostic and Interventional Imaging |volume=101 |issue=9 |pages=555–564 |doi=10.1016/j.diii.2020.03.002 |issn=2211-5684 |pmid=32278586 |s2cid=215751181|doi-access=free }}</ref>
*Projection, including [[maximum intensity projection]]<ref name="FishmanNey2006">{{Cite journal |last1=Fishman |first1=Elliot K. |author-link=Elliot K. Fishman |last2=Ney |first2=Derek R. |last3=Heath |first3=David G. |last4=Corl |first4=Frank M. |last5=Horton |first5=Karen M. |last6=Johnson |first6=Pamela T. |year=2006 |title=Volume Rendering versus Maximum Intensity Projection in CT Angiography: What Works Best, When, and Why |journal=RadioGraphics |volume=26 |issue=3 |pages=905–922 |doi=10.1148/rg.263055186 |issn=0271-5333 |pmid=16702462 |doi-access=free}}</ref> and ''average intensity projection''
*[[Volume rendering]] (VR)<ref name="FishmanNey2006" />

Technically, all volume renderings become projections when viewed on a [[Display device#Full-area 2-dimensional displays|2-dimensional display]], making the distinction between projections and volume renderings a bit vague. The epitomes of volume rendering models feature a mix of for example coloring and shading in order to create realistic and observable representations.<ref name="SilversteinParsad2008">{{Cite journal |last1=Silverstein |first1=Jonathan C. |last2=Parsad |first2=Nigel M. |last3=Tsirline |first3=Victor |year=2008 |title=Automatic perceptual color map generation for realistic volume visualization |journal=Journal of Biomedical Informatics |volume=41 |issue=6 |pages=927–935 |doi=10.1016/j.jbi.2008.02.008 |issn=1532-0464 |pmc=2651027 |pmid=18430609}}</ref><ref>{{Cite book |last=Kobbelt |first=Leif |url=https://books.google.com/books?id=zndnSzkfkXwC |title=Vision, Modeling, and Visualization 2006: Proceedings, November 22-24, 2006, Aachen, Germany |date=2006 |publisher=IOS Press |isbn=978-3-89838-081-2 |pages=185}}</ref>

Two-dimensional CT images are conventionally rendered so that the view is as though looking up at it from the patient's feet.<ref name="auto" /> Hence, the left side of the image is to the patient's right and vice versa, while anterior in the image also is the patient's anterior and vice versa. This left-right interchange corresponds to the view that physicians generally have in reality when positioned in front of patients.<ref>{{Cite journal |last1=Schmidt |first1=Derek |last2=Odland |first2=Rick |date=September 2004 |title=Mirror-Image Reversal of Coronal Computed Tomography Scans |journal=The Laryngoscope |volume=114 |issue=9 |pages=1562–1565 |doi=10.1097/00005537-200409000-00011 |issn=0023-852X |pmid=15475782 |s2cid=22320649}}</ref>

==== Grayscale ====
[[Pixel]]s in an image obtained by CT scanning are displayed in terms of relative [[radiodensity]]. The pixel itself is displayed according to the mean [[attenuation]] of the tissue(s) that it corresponds to on a scale from +3,071 (most attenuating) to −1,024 (least attenuating) on the [[Hounsfield scale]]. A [[pixel]] is a two dimensional unit based on the matrix size and the field of view. When the CT slice thickness is also factored in, the unit is known as a [[voxel]], which is a three-dimensional unit.<ref>{{Cite book |url=https://books.google.com/books?id=63xxDwAAQBAJ |title=Brant and Helms' fundamentals of diagnostic radiology |date=2018-07-19 |publisher=Lippincott Williams & Wilkins |isbn=978-1-4963-6738-9 |edition=Fifth |pages=1600 |access-date=24 January 2019}}</ref> Water has an attenuation of 0 [[Hounsfield units]] (HU), while air is −1,000&nbsp;HU, cancellous bone is typically +400&nbsp;HU, and cranial bone can reach 2,000&nbsp;HU.<ref>{{Cite book |title=Brain mapping: the methods |date=2002 |publisher=Academic Press |isbn=0-12-693019-8 |editor-last=Arthur W. Toga |edition=2nd |location=Amsterdam |oclc=52594824 |editor-last2=John C. Mazziotta}}</ref> The attenuation of metallic implants depends on the atomic number of the element used: Titanium usually has an amount of +1000&nbsp;HU, iron steel can completely block the X-ray and is, therefore, responsible for well-known line-artifacts in computed tomograms. Artifacts are caused by abrupt transitions between low- and high-density materials, which results in data values that exceed the dynamic range of the processing electronics.<ref name="...">{{Cite book |last1=Jerrold T. Bushberg |title=The essential physics of medical imaging |last2=J. Anthony Seibert |last3=Edwin M. Leidholdt |last4=John M. Boone |date=2002 |publisher=Lippincott Williams & Wilkins |isbn=0-683-30118-7 |edition=2nd |location=Philadelphia, PA |page=358 |oclc=47177732}}</ref>

==== Windowing ====
CT data sets have a very high [[dynamic range]] which must be reduced for display or printing. This is typically done via a process of "windowing", which maps a range (the "window") of pixel values to a grayscale ramp. For example, CT images of the brain are commonly viewed with a window extending from 0 HU to 80 HU. Pixel values of 0 and lower, are displayed as black; values of 80 and higher are displayed as white; values within the window are displayed as a gray intensity proportional to position within the window.<ref>{{Cite book |last1=Kamalian |first1=Shervin |last2=Lev |first2=Michael H. |last3=Gupta |first3=Rajiv |chapter=Computed tomography imaging and angiography – principles |date=2016-01-01 |title=Neuroimaging Part I |series=Handbook of Clinical Neurology |volume=135 |pages=3–20 |doi=10.1016/B978-0-444-53485-9.00001-5 |isbn=978-0-444-53485-9 |issn=0072-9752 |pmid=27432657}}</ref> The window used for display must be matched to the X-ray density of the object of interest, in order to optimize the visible detail.<ref>{{Cite book |last=Stirrup |first=James |url=https://books.google.com/books?id=SarDDwAAQBAJ&q=windowing+in+ct&pg=PA136 |title=Cardiovascular Computed Tomography |date=2020-01-02 |publisher=Oxford University Press |isbn=978-0-19-880927-2 |page=136}}</ref> Window width and window level parameters are used to control the windowing of a scan.<ref>{{Cite book |last=Carroll |first=Quinn B. |url=https://books.google.com/books?id=iTwYI5rzeRMC&dq=window+width+and+window+level&pg=PA512 |title=Practical Radiographic Imaging |date=2007 |publisher=Charles C Thomas Publisher |isbn=978-0-398-08511-7|page=512}}</ref>

==== Multiplanar reconstruction and projections{{anchor|Multiplanar_reconstruction}} ====
[[File:Ct-workstation-neck.jpg|thumb|Typical screen layout for diagnostic software, showing one volume rendering (VR) and multiplanar view of three thin slices in the [[axial plane|axial]] (upper right), [[sagittal plane|sagittal]] (lower left), and [[coronal plane]]s (lower right)]]
[[File:CT of spondylosis causing radiculopathy.png|thumb|left|Special planes are sometimes useful, such as this oblique longitudinal plane in order to visualize the neuroforamina of the vertebral column, showing narrowing at two levels, causing [[radiculopathy]]. The smaller images are axial plane slices.|148x148px]]
Multiplanar reconstruction (MPR) is the process of converting data from one [[anatomical plane]] (usually [[Transverse plane|transverse]]) to other planes. It can be used for thin slices as well as projections. Multiplanar reconstruction is possible as present CT scanners provide almost [[isotropy|isotropic]] resolution.<ref name="ref3">{{Cite book |last1=Udupa |first1=Jayaram K. |url=https://books.google.com/books?id=aR6PHYluq4oC&q=3D+Imaging+in+Medicine%2C+2nd+Edition |title=3D Imaging in Medicine, Second Edition |last2=Herman |first2=Gabor T. |date=1999-09-28 |publisher=CRC Press |isbn=978-0-8493-3179-4}}</ref>

MPR is used almost in every scan. The spine is frequently examined with it.<ref>{{Cite journal |last1=Krupski |first1=Witold |last2=Kurys-Denis |first2=Ewa |last3=Matuszewski |first3=Łukasz |last4=Plezia |first4=Bogusław |date=2007-06-30 |title=Use of multi-planar reconstruction (MPR) and 3-dimentional &#91;sic&#93; (3D) CT to assess stability criteria in C2 vertebral fractures |url=http://www.jpccr.eu/Use-of-multi-planar-reconstruction-MPR-and-3-dimentional-3D-CT-to-assess-stability,71238,0,2.html |journal=Journal of Pre-Clinical and Clinical Research |volume=1 |issue=1 |pages=80–83 |issn=1898-2395}}</ref> An image of the spine in axial plane can only show one vertebral bone at a time and cannot show its relation with other vertebral bones. By reformatting the data in other planes, visualization of the relative position can be achieved in sagittal and coronal plane.<ref>{{Cite journal |last=Tins |first=Bernhard |date=2010-10-21 |title=Technical aspects of CT imaging of the spine |journal=Insights into Imaging |volume=1 |issue=5–6 |pages=349–359 |doi=10.1007/s13244-010-0047-2 |issn=1869-4101 |pmc=3259341 |pmid=22347928}}</ref>

New software allows the reconstruction of data in non-orthogonal (oblique) planes, which help in the visualization of organs which are not in orthogonal planes.<ref>{{Cite web |title=CT imaging: Where are we going? (Proceedings) |url=https://www.dvm360.com/view/ct-imaging-where-are-we-going-proceedings |access-date=2021-03-21 |website=DVM 360|date=April 2010}}</ref><ref>{{Cite book |last1=Wolfson |first1=Nikolaj |url=https://books.google.com/books?id=8Y5FDAAAQBAJ&q=Modern+software+allows+reconstruction+in+non-orthogonal&pg=PA373 |title=Orthopedics in Disasters: Orthopedic Injuries in Natural Disasters and Mass Casualty Events |last2=Lerner |first2=Alexander |last3=Roshal |first3=Leonid |date=2016-05-30 |publisher=Springer |isbn=978-3-662-48950-5}}</ref> It is better suited for visualization of the anatomical structure of the bronchi as they do not lie orthogonal to the direction of the scan.<ref>{{Cite journal |last1=Laroia |first1=Archana T |last2=Thompson |first2=Brad H |last3=Laroia |first3=Sandeep T |last4=van Beek |first4=Edwin JR |date=2010-07-28 |title=Modern imaging of the tracheo-bronchial tree |journal=World Journal of Radiology |volume=2 |issue=7 |pages=237–248 |doi=10.4329/wjr.v2.i7.237 |issn=1949-8470 |pmc=2998855 |pmid=21160663 |doi-access=free}}</ref>

Curved-plane reconstruction (or curved planar reformation = CPR) is performed mainly for the evaluation of vessels. This type of reconstruction helps to straighten the bends in a vessel, thereby helping to visualize a whole vessel in a single image or in multiple images. After a vessel has been "straightened", measurements such as cross-sectional area and length can be made. This is helpful in preoperative assessment of a surgical procedure.<ref>{{Cite journal |last1=Gong |first1=Jing-Shan |last2=Xu |first2=Jian-Min |date=2004-07-01 |title=Role of curved planar reformations using multidetector spiral CT in diagnosis of pancreatic and peripancreatic diseases |journal=World Journal of Gastroenterology |volume=10 |issue=13 |pages=1943–1947 |doi=10.3748/wjg.v10.i13.1943 |issn=1007-9327 |pmc=4572236 |pmid=15222042 |doi-access=free}}</ref>

For 2D projections used in [[radiation therapy]] for quality assurance and planning of [[external beam radiotherapy]], including digitally reconstructed radiographs, see [[Beam's eye view]].

{| class="wikitable"
|+Examples of different algorithms of thickening multiplanar reconstructions<ref>{{Cite journal |last1=Dalrymple |first1=Neal C. |last2=Prasad |first2=Srinivasa R. |last3=Freckleton |first3=Michael W. |last4=Chintapalli |first4=Kedar N. |date=September 2005 |title=Informatics in radiology (infoRAD): introduction to the language of three-dimensional imaging with multidetector CT |journal=Radiographics |volume=25 |issue=5 |pages=1409–1428 |doi=10.1148/rg.255055044 |issn=1527-1323 |pmid=16160120}}</ref>
!Type of projection
!Schematic illustration
!Examples (10&nbsp;mm slabs)
!Description
!Uses
|-
|Average intensity projection (AIP)
|[[File:Average intensity projection.gif|frameless]]
|[[File:Coronal average intensity projection CT thorax.gif|frameless|118x118px]]
|The average attenuation of each voxel is displayed. The image will get smoother as slice thickness increases. It will look more and more similar to conventional [[projectional radiography]] as slice thickness increases.
|Useful for identifying the internal structures of a solid organ or the walls of hollow structures, such as intestines.
|-
|[[Maximum intensity projection]] (MIP)
|[[File:Maximum intensity projection.gif|frameless]]
|[[File:Coronal maximum intensity projection CT thorax.gif|frameless|118x118px]]
|The voxel with the highest attenuation is displayed. Therefore, high-attenuating structures such as blood vessels filled with contrast media are enhanced.
|Useful for angiographic studies and identification of pulmonary nodules.
|-
|[[Minimum intensity projection]] (MinIP)
|[[File:Minimum intensity projection.gif|frameless]]
|[[File:Coronal minimum intensity projection CT thorax.gif|frameless|117x117px]]
|The voxel with the lowest attenuation is displayed. Therefore, low-attenuating structures such as air spaces are enhanced.
|Useful for assessing the lung parenchyma.
|}

==== {{anchor|3D}} Volume rendering ====
{{Main|Volume rendering}}
[[File:12-06-11-rechtsmedizin-berlin-07.jpg|thumbnail|3D human skull from computed tomography data]]

A threshold value of radiodensity is set by the operator (e.g., a level that corresponds to bone). With the help of [[edge detection]] image processing algorithms a 3D model can be constructed from the initial data and displayed on screen. Various thresholds can be used to get multiple models, each anatomical component such as muscle, bone and cartilage can be differentiated on the basis of different colours given to them. However, this mode of operation cannot show interior structures.<ref>{{Cite journal |last1=Calhoun |first1=Paul S. |last2=Kuszyk |first2=Brian S. |last3=Heath |first3=David G. |last4=Carley |first4=Jennifer C. |last5=Fishman |first5=Elliot K. |date=1999-05-01 |title=Three-dimensional Volume Rendering of Spiral CT Data: Theory and Method |url=https://pubs.rsna.org/doi/full/10.1148/radiographics.19.3.g99ma14745 |journal=RadioGraphics |volume=19 |issue=3 |pages=745–764 |doi=10.1148/radiographics.19.3.g99ma14745 |issn=0271-5333 |pmid=10336201}}</ref>

Surface rendering is limited technique as it displays only the surfaces that meet a particular threshold density, and which are towards the viewer. However, In volume rendering, transparency, colours and [[Phong shading|shading]] are used which makes it easy to present a volume in a single image. For example, Pelvic bones could be displayed as semi-transparent, so that, even viewing at an oblique angle one part of the image does not hide another.<ref>{{Cite journal |last1=van Ooijen |first1=P. M. A. |last2=van Geuns |first2=R. J. M. |last3=Rensing |first3=B. J. W. M. |last4=Bongaerts |first4=A. H. H. |last5=de Feyter |first5=P. J. |last6=Oudkerk |first6=M. |date=January 2003 |title=Noninvasive Coronary Imaging Using Electron Beam CT: Surface Rendering Versus Volume Rendering |url=http://www.ajronline.org/doi/10.2214/ajr.180.1.1800223 |journal=American Journal of Roentgenology |volume=180 |issue=1 |pages=223–226 |doi=10.2214/ajr.180.1.1800223 |issn=0361-803X |pmid=12490509}}</ref>

=== Image quality ===
[[File:CT-Low-Dose-2.5-LUNG.ogg|thumb|Low-dose CT scan of the thorax]]
[[File:Standard dose high resolution chest CT (HRCT).ogg|thumb|Standard-dose CT scan of the thorax]]

==== Dose versus image quality ====
An important issue within radiology today is how to reduce the radiation dose during CT examinations without compromising the image quality. In general, higher radiation doses result in higher-resolution images,<ref name="Crowther">{{Cite journal |last1=R. A. Crowther |last2=D. J. DeRosier |last3=A. Klug |year=1970 |title=The Reconstruction of a Three-Dimensional Structure from Projections and its Application to Electron Microscopy |journal=Proc. R. Soc. Lond. A |volume=317 |issue=1530 |pages=319–340 |bibcode=1970RSPSA.317..319C |doi=10.1098/rspa.1970.0119 |s2cid=122980366}}</ref> while lower doses lead to increased image noise and unsharp images. However, increased dosage raises the adverse side effects, including the risk of [[radiation-induced cancer]] – a four-phase abdominal CT gives the same radiation dose as 300 chest X-rays.<ref>{{Cite journal |last1=Nickoloff |first1=Edward L. |last2=Alderson |first2=Philip O. |date=August 2001 |title=Radiation Exposures to Patients from CT: Reality, Public Perception, and Policy |url=http://www.ajronline.org/doi/10.2214/ajr.177.2.1770285 |journal=American Journal of Roentgenology |volume=177 |issue=2 |pages=285–287 |doi=10.2214/ajr.177.2.1770285 |issn=0361-803X |pmid=11461846}}</ref> Several methods that can reduce the exposure to ionizing radiation during a CT scan exist.<ref name="ata">Barkan, O; Weill, J; Averbuch, A; Dekel, S. [http://www.cv-foundation.org/openaccess/content_cvpr_2013/papers/Barkan_Adaptive_Compressed_Tomography_2013_CVPR_paper.pdf "Adaptive Compressed Tomography Sensing"] {{webarchive |url=https://web.archive.org/web/20160313133222/http://www.cv-foundation.org/openaccess/content_cvpr_2013/papers/Barkan_Adaptive_Compressed_Tomography_2013_CVPR_paper.pdf |date=2016-03-13}}. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2013 (pp. 2195–2202).</ref>

# New software technology can significantly reduce the required radiation dose. New [[Iterative reconstruction|iterative]] [[tomographic reconstruction]] algorithms (''e.g.'', [[SAMV (algorithm)|iterative Sparse Asymptotic Minimum Variance]]) could offer [[Super-resolution imaging|super-resolution]] without requiring higher radiation dose.<ref>{{Cite book |url=https://books.google.com/books?id=hclVAAAAMAAJ&q=iterative+construction+gives+super+resolution |title=Proceedings |date=1995 |publisher=IEEE |page=10 |isbn=978-0-7803-2498-5}}</ref>
# Individualize the examination and adjust the radiation dose to the body type and body organ examined. Different body types and organs require different amounts of radiation.<ref>{{Cite web |title=Radiation – Effects on organs of the body (somatic effects) |url=https://www.britannica.com/science/radiation |access-date=2021-03-21 |website=Encyclopedia Britannica}}</ref>
# Higher resolution is not always suitable, such as detection of small pulmonary masses.<ref>{{Cite journal |last=Simpson G |year=2009 |title=Thoracic computed tomography: principles and practice |journal=Australian Prescriber |volume=32 |issue=4 |page=4 |doi=10.18773/austprescr.2009.049 |doi-access=free}}</ref>

==== Artifacts ====
Although images produced by CT are generally faithful representations of the scanned volume, the technique is susceptible to a number of [[artifact (error)#Medical imaging|artifacts]], such as the following:<ref name="ref1" /><ref>{{Cite journal |last1=Bhowmik |first1=Ujjal Kumar |last2=Zafar Iqbal, M. |last3=Adhami, Reza R. |date=28 May 2012 |title=Mitigating motion artifacts in FDK based 3D Cone-beam Brain Imaging System using markers |journal=Central European Journal of Engineering |volume=2 |issue=3 |pages=369–382 |bibcode=2012CEJE....2..369B |doi=10.2478/s13531-012-0011-7 |doi-access=free}}</ref><sup>Chapters 3 and 5</sup>

;{{Visible anchor|Streak artifact}}: Streaks are often seen around materials that block most X-rays, such as metal or bone. Numerous factors contribute to these streaks: under sampling, photon starvation, motion, beam hardening, and [[Compton scatter]]. This type of artifact commonly occurs in the posterior fossa of the brain, or if there are metal implants. The streaks can be reduced using newer reconstruction techniques.<ref name="P. Jin and C. A. Bouman and K. D. Sauer 2013">{{Cite journal |last1=P. Jin |last2=C. A. Bouman |last3=K. D. Sauer |year=2013 |title=A Method for Simultaneous Image Reconstruction and Beam Hardening Correction |url=https://engineering.purdue.edu/~bouman/publications/pdf/mic2013.pdf |url-status=dead |journal=IEEE Nuclear Science Symp. & Medical Imaging Conf., Seoul, Korea, 2013 |archive-url=https://web.archive.org/web/20140606234132/https://engineering.purdue.edu/~bouman/publications/pdf/mic2013.pdf |archive-date=2014-06-06 |access-date=2014-04-23}}</ref> Approaches such as metal artifact reduction (MAR) can also reduce this artifact.<ref>{{Cite journal |vauthors=Boas FE, Fleischmann D |year=2011 |title=Evaluation of Two Iterative Techniques for Reducing Metal Artifacts in Computed Tomography |journal=Radiology |volume=259 |issue=3 |pages=894–902 |doi=10.1148/radiol.11101782 |pmid=21357521}}</ref><ref name="mouton13survey">{{Cite journal |last1=Mouton, A. |last2=Megherbi, N. |last3=Van Slambrouck, K. |last4=Nuyts, J. |last5=Breckon, T.P. |year=2013 |title=An Experimental Survey of Metal Artefact Reduction in Computed Tomography |url=http://www.durham.ac.uk/toby.breckon/publications/papers/mouton13survey.pdf |journal=Journal of X-Ray Science and Technology |volume=21 |issue=2 |pages=193–226 |doi=10.3233/XST-130372 |pmid=23694911 |hdl=1826/8204 }}{{Dead link|date=November 2023 |bot=InternetArchiveBot |fix-attempted=yes }}</ref> MAR techniques include spectral imaging, where CT images are taken with [[photons]] of different energy levels, and then synthesized into [[monochromatic]] images with special software such as GSI (Gemstone Spectral Imaging).<ref name="PessisCampagna2013">{{Cite journal |last1=Pessis |first1=Eric |last2=Campagna |first2=Raphaël |last3=Sverzut |first3=Jean-Michel |last4=Bach |first4=Fabienne |last5=Rodallec |first5=Mathieu |last6=Guerini |first6=Henri |last7=Feydy |first7=Antoine |last8=Drapé |first8=Jean-Luc |year=2013 |title=Virtual Monochromatic Spectral Imaging with Fast Kilovoltage Switching: Reduction of Metal Artifacts at CT |journal=RadioGraphics |volume=33 |issue=2 |pages=573–583 |doi=10.1148/rg.332125124 |issn=0271-5333 |pmid=23479714 |doi-access=free}}</ref>

;Partial volume effect: This appears as "blurring" of edges. It is due to the scanner being unable to differentiate between a small amount of high-density material (e.g., bone) and a larger amount of lower density (e.g., cartilage).<ref>{{Cite journal |last1=González Ballester |first1=Miguel Angel |last2=Zisserman |first2=Andrew P. |last3=Brady |first3=Michael |date=December 2002 |title=Estimation of the partial volume effect in MRI |journal=Medical Image Analysis |volume=6 |issue=4 |pages=389–405 |doi=10.1016/s1361-8415(02)00061-0 |issn=1361-8415 |pmid=12494949}}</ref> The reconstruction assumes that the X-ray attenuation within each voxel is homogeneous; this may not be the case at sharp edges. This is most commonly seen in the z-direction (craniocaudal direction), due to the conventional use of highly [[isotropic|anisotropic]] voxels, which have a much lower out-of-plane resolution, than in-plane resolution. This can be partially overcome by scanning using thinner slices, or an isotropic acquisition on a modern scanner.<ref>{{Cite journal |last1=Goldszal |first1=Alberto F. |last2=Pham |first2=Dzung L. |date=2000-01-01 |title=Volumetric Segmentation |journal=Handbook of Medical Imaging |pages=185–194 |doi=10.1016/B978-012077790-7/50016-3 |isbn=978-0-12-077790-7}}</ref>

;Ring artifact: Probably the most common mechanical artifact, the image of one or many "rings" appears within an image. They are usually caused by the variations in the response from individual elements in a two dimensional X-ray detector due to defect or miscalibration.<ref name="Jha">{{Cite journal |last=Jha |first=Diwaker |date=2014 |title=Adaptive center determination for effective suppression of ring artifacts in tomography images |journal=Applied Physics Letters |volume=105 |issue=14 |pages=143107 |bibcode=2014ApPhL.105n3107J |doi=10.1063/1.4897441}}</ref> Ring artifacts can largely be reduced by intensity normalization, also referred to as flat field correction.<ref name="vvn15">{{Cite journal |last1=Van Nieuwenhove |first1=V |last2=De Beenhouwer |first2=J |last3=De Carlo |first3=F |last4=Mancini |first4=L |last5=Marone |first5=F |last6=Sijbers |first6=J |date=2015 |title=Dynamic intensity normalization using eigen flat fields in X-ray imaging |url=http://www.zora.uzh.ch/id/eprint/120683/1/oe-23-21-27975.pdf |journal=Optics Express |volume=23 |issue=21 |pages=27975–27989 |bibcode=2015OExpr..2327975V |doi=10.1364/oe.23.027975 |pmid=26480456 |doi-access=free |hdl=10067/1302930151162165141}}</ref> Remaining rings can be suppressed by a transformation to polar space, where they become linear stripes.<ref name="Jha" /> A comparative evaluation of ring artefact reduction on X-ray tomography images showed that the method of Sijbers and Postnov can effectively suppress ring artefacts.<ref name="jsap">{{Cite journal |vauthors=Sijbers J, Postnov A |date=2004 |title=Reduction of ring artefacts in high resolution micro-CT reconstructions |journal=Phys Med Biol |volume=49 |issue=14 |pages=N247–53 |doi=10.1088/0031-9155/49/14/N06 |pmid=15357205 |s2cid=12744174}}</ref>

;Noise: This appears as grain on the image and is caused by a low signal to noise ratio. This occurs more commonly when a thin slice thickness is used. It can also occur when the power supplied to the X-ray tube is insufficient to penetrate the anatomy.<ref>{{Cite book |last1=Newton |first1=Thomas H. |url=https://books.google.com/books?id=2mxsAAAAMAAJ&q=noise+in+computed+tomography |title=Radiology of the Skull and Brain: Technical aspects of computed tomography |last2=Potts |first2=D. Gordon |date=1971 |publisher=Mosby |isbn=978-0-8016-3662-2 |pages=3941–3950}}</ref>

;Windmill: Streaking appearances can occur when the detectors intersect the reconstruction plane. This can be reduced with filters or a reduction in pitch.<ref>{{Cite book |last1=Brüning |first1=R. |url=https://books.google.com/books?id=ImOlZNOk25sC&q=windmill+artifact+ct&pg=PA44 |title=Protocols for Multislice CT |last2=Küttner |first2=A. |last3=Flohr |first3=T. |date=2006-01-16 |publisher=Springer Science & Business Media |isbn=978-3-540-27273-1}}</ref><ref>{{Cite book |last=Peh |first=Wilfred C. G. |url=https://books.google.com/books?id=sZswDwAAQBAJ&q=windmill+artifact+ct&pg=PA49 |title=Pitfalls in Musculoskeletal Radiology |date=2017-08-11 |publisher=Springer |isbn=978-3-319-53496-1}}</ref>

;Beam hardening: This can give a "cupped appearance" when grayscale is visualized as height. It occurs because conventional sources, like X-ray tubes emit a polychromatic spectrum. Photons of higher [[photon energy]] levels are typically attenuated less. Because of this, the mean energy of the spectrum increases when passing the object, often described as getting "harder". This leads to an effect increasingly underestimating material thickness, if not corrected. Many algorithms exist to correct for this artifact. They can be divided into mono- and multi-material methods.<ref name="P. Jin and C. A. Bouman and K. D. Sauer 2013" /><ref>{{Cite journal |vauthors=Van de Casteele E, Van Dyck D, Sijbers J, Raman E |year=2004 |title=A model-based correction method for beam hardening artefacts in X-ray microtomography |journal=Journal of X-ray Science and Technology |volume=12 |issue=1 |pages=43–57 |citeseerx=10.1.1.460.6487}}</ref><ref>{{Cite journal |vauthors=Van Gompel G, Van Slambrouck K, Defrise M, Batenburg KJ, Sijbers J, Nuyts J |year=2011 |title=Iterative correction of beam hardening artifacts in CT |journal=Medical Physics |volume=38 |issue=1 |pages=36–49 |bibcode=2011MedPh..38S..36V |citeseerx=10.1.1.464.3547 |doi=10.1118/1.3577758 |pmid=21978116}}</ref>