Guide to artifacts in MRI: classification, explanation, countermeasures.

Similar documents
M R I Physics Course. Jerry Allison Ph.D. Chris Wright B.S. Tom Lavin M.S.M.P. Department of Radiology Medical College of Georgia

Abstract. Learning Objectives 8/1/2017

RAD 465 (MRI) Lecture one (Pulse Sequences) Ruba Khushaim MSc

MSK Imaging Fundamentals

4/14/2009. The Big Picture of Quality. MRI Quality Assurance and ACR MRI Accreditation Program. Basic Elements for Image Quality.

Phantom Test Guidance for Use of the Small MRI Phantom for the MRI Accreditation Program

Abdominal medical devices: what the radiologist needs to know

EPI. Thanks to Samantha Holdsworth!

Breast MR Imaging and Quality Control

Artifacts in Abdominopelvic MR A Pictorial Review

True comfort and flexibility with the power of 3T.

Image quality in non-gated versus gated reconstruction of tongue motion using Magnetic Resonance Imaging:

Premium 1.5T MRI System

HITACHI S FAST SPIN ECHO TECHNOLOGY

Magnetic resonance imaging phase encoding:

Overview of All Pixel Circuits for Active Matrix Organic Light Emitting Diode (AMOLED)

RSNA 2006 November 26 to December 1 Chicago. Guest author for ImPACT Dr. Koos Geleijns, Medical Physicist, Leiden University Medical Center.

Troubleshooting Guide. Prep, Scan Errors, and Artifacts

Development of Simple-Matrix LCD Module for Motion Picture

Nuclear Associates and


Light. Engineered for Performance- Esaote MRI, designed to make a difference.

DVM-3000 Series 12 Bit DIGITAL VIDEO, AUDIO and 8 CHANNEL BI-DIRECTIONAL DATA FIBER OPTIC MULTIPLEXER for SURVEILLANCE and TRANSPORTATION

Multiparametric MRI Prostate Imaging Protocol November 2015 Full Acquisition Protocol with Parameters GE 3T Magnet with Software Version DV25

Chapter 7. Scanner Controls

UNIT-3 Part A. 2. What is radio sonde? [ N/D-16]

CM3106 Solutions. Do not turn this page over until instructed to do so by the Senior Invigilator.

More Info at Open Access Database Process Control for Computed Tomography using Digital Detector Arrays

Request for Proposals

Chapter 6: Real-Time Image Formation

Brain-Computer Interface (BCI)

Quick reference guide

Colour Reproduction Performance of JPEG and JPEG2000 Codecs

Peacefully quiet. Remarkably fast.

White Paper. Uniform Luminance Technology. What s inside? What is non-uniformity and noise in LCDs? Why is it a problem? How is it solved?

DIGITAL COMMUNICATION

Procedure Manual for MRI of the Brain

Peacefully quiet. Remarkably fast.

Experiences in ACR MRI Accreditation Vendor Nuances That Every Clinical MRI Physicist Should Know

Comparison of Robarts s 3T and 7T MRI Machines for obtaining fmri Sequences Medical Biophysics 3970: General Laboratory

AN IMPROVED ERROR CONCEALMENT STRATEGY DRIVEN BY SCENE MOTION PROPERTIES FOR H.264/AVC DECODERS

Understanding CT image quality

Understanding Compression Technologies for HD and Megapixel Surveillance

Introduction: Overview. EECE 2510 Circuits and Signals: Biomedical Applications. ECG Circuit 2 Analog Filtering and A/D Conversion

2

Multimedia Systems Video I (Basics of Analog and Digital Video) Mahdi Amiri April 2011 Sharif University of Technology

Pre-processing of revolution speed data in ArtemiS SUITE 1

UNIVERSAL SPATIAL UP-SCALER WITH NONLINEAR EDGE ENHANCEMENT

Microincrements IP67-related solutions

ENGINEERING COMMITTEE

Calibrate, Characterize and Emulate Systems Using RFXpress in AWG Series

Assessing and Measuring VCR Playback Image Quality, Part 1. Leo Backman/DigiOmmel & Co.

Innovative Rotary Encoders Deliver Durability and Precision without Tradeoffs. By: Jeff Smoot, CUI Inc

T ips in measuring and reducing monitor jitter

PATIENT POSITION IMAGING PARAMETERS

Torsional vibration analysis in ArtemiS SUITE 1

Temporal Error Concealment Algorithm Using Adaptive Multi- Side Boundary Matching Principle

Vascular. Development of Trinias FPD-Equipped Angiography System. 1. Introduction. MEDICAL NOW No.73 (2013.2) Yoshiaki Miura

BitWise (V2.1 and later) includes features for determining AP240 settings and measuring the Single Ion Area.

COMPRESSION OF DICOM IMAGES BASED ON WAVELETS AND SPIHT FOR TELEMEDICINE APPLICATIONS

Practical Application of the Phased-Array Technology with Paint-Brush Evaluation for Seamless-Tube Testing

DELTA MODULATION AND DPCM CODING OF COLOR SIGNALS

MIE 402: WORKSHOP ON DATA ACQUISITION AND SIGNAL PROCESSING Spring 2003

Technology White Paper Plasma Displays. NEC Technologies Visual Systems Division

Joint ICTP/IAEA Advanced School on Dosimetry in Diagnostic Radiology and its Clinical Implementation May 2009

System Quality Indicators

Financial disclosure statement. Fluoroscopic Equipment Design: What s s Different with Flat Panel? Concept of flat panel imager

FRAME RATE CONVERSION OF INTERLACED VIDEO

Processing. Electrical Engineering, Department. IIT Kanpur. NPTEL Online - IIT Kanpur

ZONE PLATE SIGNALS 525 Lines Standard M/NTSC

Release Notes for LAS AF version 1.8.0

Module 8 VIDEO CODING STANDARDS. Version 2 ECE IIT, Kharagpur

A New "Duration-Adapted TR" Waveform Capture Method Eliminates Severe Limitations

Compressed-Sensing-Enabled Video Streaming for Wireless Multimedia Sensor Networks Abstract:

Multi echo Multi slice (MEMS) High Performance fmri at CFMRI... 1

Investigation of Digital Signal Processing of High-speed DACs Signals for Settling Time Testing

MULTI-STATE VIDEO CODING WITH SIDE INFORMATION. Sila Ekmekci Flierl, Thomas Sikora

Perfecting the Package Bare and Overmolded Stacked Dies. Understanding Ultrasonic Technology for Advanced Package Inspection. A Sonix White Paper

NMR. picospin. Maintenance Guide

The trigger for the New Electromagnetic Calorimeter NewCal

PIECEWISE PRODUCTION MACHINES

The Development of a Synthetic Colour Test Image for Subjective and Objective Quality Assessment of Digital Codecs

CHARACTERIZATION OF END-TO-END DELAYS IN HEAD-MOUNTED DISPLAY SYSTEMS

Muscle Sensor KI 2 Instructions

Microincrements XFC. Application Note DK XFC technology microincrements. Technical background CHA CHB. 2fold.

+ Human method is pattern recognition based upon multiple exposure to known samples.

6.111 Project Proposal IMPLEMENTATION. Lyne Petse Szu-Po Wang Wenting Zheng

ESE-10 Multipurpose Color Ultrasound System IMAGING SYSTEMS

INDIAN INSTITUTE OF TECHNOLOGY KHARAGPUR NPTEL ONLINE CERTIFICATION COURSE. On Industrial Automation and Control

1. INTRODUCTION. Index Terms Video Transcoding, Video Streaming, Frame skipping, Interpolation frame, Decoder, Encoder.

Philips Site Yearly Performance Evaluation Philips Achieva - Gibbons 1.5T 1-Jun-08. Table of Contents

Laser Beam Analyser Laser Diagnos c System. If you can measure it, you can control it!

Lecture 2 Video Formation and Representation

EddyCation - the All-Digital Eddy Current Tool for Education and Innovation

CHAPTER 2 SUBCHANNEL POWER CONTROL THROUGH WEIGHTING COEFFICIENT METHOD

Ch. 1: Audio/Image/Video Fundamentals Multimedia Systems. School of Electrical Engineering and Computer Science Oregon State University

AN ARTISTIC TECHNIQUE FOR AUDIO-TO-VIDEO TRANSLATION ON A MUSIC PERCEPTION STUDY

V9A01 Solution Specification V0.1

Interlace and De-interlace Application on Video

Spatio-temporal inaccuracies of video-based ultrasound images of the tongue

Transcription:

Guide to artifacts in MRI: classification, explanation, countermeasures. Poster No.: C-1089 Congress: ECR 2017 Type: Educational Exhibit Authors: A. Cassarà 1, G. Fini 1, T. Bongiovanni 1, F. Cartabbia 1, E. Soligo 1, M. Abruzzese 1, L. Y. Samman 1, A. Stecco 1, A. Carriero 2 ; 1 NOVARA/IT, 2 Novara (NO)/IT Keywords: DOI: Artifacts, Diagnostic procedure, MR, MR physics 10.1594/ecr2017/C-1089 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. www.myesr.org Page 1 of 26

Learning objectives The goal of our work is to classify, explain the genesis and provide countermeasures to artifacts that may be generated in MRI exam. Page 2 of 26

Background An artifact in MRI is configured as a result of distortion / alteration of the image that causes an image representation that does not correspond to reality. It is a finding inside the images not normally anatomically present, but visible due to limitation / malfunction in hardware / software, extrinsic (heat or humidity environment) or intrinsic (pulsation, breathing, implants) factors. The knowledge of the artifacts and the factors that produce noise is important to obtain high quality images. From this the idea to classify, describe and find one or more solutions to artifacts that frequently can jeopardize the attainment of a proper MRI examination. Some artifacts are avoidable, others may also be exploited for diagnostic purposes. The artifacts in this discussion were divided into three categories: tissue, motion and technique related. Tissue related: Chemical-Shift FID Signal Field Inhomogeneity Gibbs Magic angle Magnetic Susceptivity Moirè fringes Partial Volume Gadolinium hyperconcentration Motion related - Ghost - Smear Patient movement Technique related: 1. Scan technique Cross talk Aliasing Page 3 of 26

2. Interpherences RF Noise RF inhomogeneity Eddy current 3. Hardware & Software Central point Crisscross Entry slice Nyquist N/2 Zipper Receiver ignition Hardware errors Data corruption Analogic-digital converter block Artifacts may be very obvious or confined to a few pixels and can mislead the diagnosis, hence the need to apply countermeasures. Page 4 of 26

Findings and procedure details We describe case-by-case from technical reason to countermeasures, evaluating them on both low magnetic field (Esaote, O-Scan) and 1.5 T magnet (Philips D-Stream). Chemical shift (CS) artifact can be divided into type I and II. Type I is found along the direction of frequency encoding, only in the presence of magnetic fields (B 0 ) > 1T. Type II is found in any of the field intensity, especially in Gradient Echo (GE) sequences with particular echo times. The alteration appears as a black band or white at the interface between water and fat (Fig. 1). CS can also be used for detection of certain type of lesion like adrenal gland mass. The artifact increases with the increase in the intensity of B 0, but decreases with the increase of the intensity of magnetic field gradients applied. The solutions can be managed directly from the console: the devices include a parameter that determines how many pixels can be present the CS and in which direction should occur. Possible remedies: increase bandwidth; use TE in phase; use less subject sequences. Free Induction Decay (FID) artifacts can appear when one or more T1 relaxation times corresponding to the tissues in the interest volume are relatively short and, contemporary, low flip angles are used for refocusing. The FID signal artifact in SE sequences appear in the image as straight lines, while in 3D FSE follows the underlying anatomy and looks like intensity undulations of the subcutaneous fat (Fig. 2). This artifact is similar to the receiver turned on artifact (hardware and software) that also occurs in GE sequences, where the signal from the receiver (not real) extends in the free signal region, while the FID (real signal) is confined to the sample projection along the reading axis. Possible solutions: increase the slope of the spoiler gradient applied to the sequence of pulses; add an average with relative increase of the acquisition time; contact support. Page 5 of 26

Field inhomogeneity artifact is characteristic of the scans at the edge of the field: when images are obtained through a progression from the center to the edge of the coil, the homogeneity of B changes with increasing distance. The same problem occurs if you scan at a certain distance from isocenter volume in the left-right direction, or when using a too large Field Of View (FOV). Similar artifacts are related to surface coils and magnetic field gradients. The images can be noisy, distorted (geometric distortion in EPI), with incomplete magnetization signal and broke in SE sequences or without fat suppression (incorrect setting of the currents). Possible remedies: use less susceptible sequences (SE and TSE); use greater FOV; oversample. Gibbs artifact, also known as "by truncation" or "ring" artifact, occurs when in an image there is a sudden change of contrast, as the interface fat / muscle or spinal cord / liquor. In this artifact "wave fronts" at regular intervals are formed (Fig. 4). This phenomenon takes the name of "ringing" and is derived from a sub-sampling along the direction of frequency encoding. In the images on the left we have two different matrices: at the top 128x128 pixels, at the bottom 256x256: note how the artifact is reduced increasing the size. Possible remedies to reduce it is: to increase the number of stages or samples; increase the size of the matrix; perform high-resolution scans; use fat suppression techniques in T1w sequencies. Magic angle artifact occurs in the structures rich in fibers and collagen (such as tendon and ligament structures) with lapse of 55 with respect to B 0. The water protons in these structures generally have a very low signal because of the dipolar interactions which occur in the fibers, and that lead to a rapid dephasing of the signal. These interactions, however, progressively decrease if the orientation of the fibers is close to 55 with respect to the magnet, with an increase of the signal which could simulate a pathological area. A typical example is the light signal present at the level of the posterior cruciate ligament, the supraspinatus tendon and, sometimes, even at the level of the patella (Fig. 5). Usually it occurs in sequences with short TE (less than 32ms, T1, DP, GE) and tends to decrease if the magnetic field is 3T. Possible remedies: increase the TE over 40 ms which leads to a deletion of the magic angle, but also to the loss of mild inflammatory diseases. Magnetic susceptibility artifact is caused by the presence of objects in the FOV with very high or very low magnetic susceptibility. In the interfaces between regions with Page 6 of 26

different magnetic susceptibility, the induction field is slightly different and field gradient (inhomogeneities) are created. The artifact is visible especially with the GE sequences and is dependent on some acquisition parameters (more evident with long TE). The most frequently are represented by metal and cosmetics artifacts, which do not only cause the distortion of neighboring structures but can also cause signal losses on the entire image (empty signal surrounded by hyper intense lines and spots), depending on the technique used (Fig. 6). Possible remedies: avoid GE sequencies; use short TE (reducing spin dephasing); employ more intense gradients; use smaller FOV, thin layers and volumetric acquisitions. Moiré Fringes artifact is caused by the lack of perfect homogeneity of B 0 from one side of the body to another, with reversal of the scanned volume and consequent overlapping of signals of different phases that automatically are added up and erase. It can, also, be given by the interference of different echoes that fall in the same reading window. This results in "Zebra Stripes" inhomogeneity of magnetic field, visible at the edges of a large FOV, especially when the patient's boundaries (eg elbows) are very close to the transmitter coil (Fig. 7). It typically occurs when GE sequences are made with the coil body. Possible remedies: use SE sequences; use surface coils; pay attention to shimming and gradients susceptibility. Partial volume artifact occurs when the size of image voxel is larger than the size of the detail that it is necessary to analyze. This results in a loss of spatial resolution. Another manifestation of this type of artifact is the loss of contrast resolution between two adjacent tissues in an image due to poor spatial resolution. If fat and water spins occupy the same voxel, their signals interact and delete. Possible remedies: reduce layer thickness; increase spatial resolution. Gadolinium (Gd) hyperconcentration consist in a loss of signal intensity in T1w images instead of the hyperintensity that occurs when the concentration is normal. In the blood, in the joints and in the bladder, often the concentration of the "contrast" could exceed a critical level in which the shortening of T2 exceeds T1, leading to a loss of signal. The consequence of this effect is an "artificial layering" of images, which are unreadable. Possible remedies: the only way to avoid this artifact is to perform the earliest possible acquisitions in order to avoid over-concentration. Page 7 of 26

Movement related artifacts represent all those artifacts that result from interactions between the equipment and the patient's body due to coding errors, both in the phase and in the frequency (negligible) of direction. The data in the direction of phase encoding are collected during the entire scan duration: consequently the coding error causes the data to be placed in a wrong pixels in the final image. The movements may be voluntary or involuntary, those represented by blood flow in the vessels or CSF flow in the ventricular cavity and the subarachnoid space, by cardiac respiratory and peristaltic movements. The artifacts are presented in various ways: images called "ghost" (depend on whether the raw data the "K" space are acquired alternately during inspiratory and expiratory / systolic and diastolic, they are represented in the form of side dishes or as an image "fuzzy" homogeneously) or of type "smear" (for non-periodic movements, characterized by continuous bands). Possible remedies: immobilize the affected area; use the cardiac or respiratory gating; capture scans in apnea; increase the number of acquisitions; use the presaturation; reverse the encoding axes of phase and frequency; use gradients of the flow compensation; sequences preferred with fat saturation; use of surface coils; increase the number of average signal; Antiperistaltic medications. Crosstalk artifact (overlap, cross excitation) is always present when placing packets of layers that intersect and are detected in the same acquisition. The signals of the two different layers for the same anatomical point create interference that generate empty or very low signal, making unreadable images in those voxels. Possible remedies: avoid overlapping of layers of packages, or to do so in areas that are not important from a diagnostic point of view; acquire interspersed mode (mandatory in 2 batches), not sequential; perform an acquisition (a batch) for each packet of parallel layers (expensive solution in terms of time). Aliasing artifact (Wrap Around, Foldover, by tilting, Backfolding-only parallel imaging-) is the presence in the image of a part of the examined anatomical region that is outside the FOV. It occurs in the K-space region, when the FOV is smaller than the affected anatomical region. The parts of the object outside the FOV still produce a signal, but the intensity of the applied gradients causes them to gain an upper phase and, therefore, the reconstruction algorithm considers them overlapped at the opposite side of the image (Fig. 11). On modern tomographs it only occurs along the encoding phase direction, while in those of old generation is also reported along the direction of frequency encoding. Page 8 of 26

Possible remedies: on older systems the only way to overcome the problem of aliasing was oversample the signal, eliminating the data collected from the anatomical regions outside the FOV; in newer equipment it's possible to use a digital filter system, which eliminate upstream frequencies outside the range of interest or place the object of study in the center of the FOV; increase the FOV in the encoding phase direction; use of phase oversampling techniques; swap the encoding phase axis with that of the frequencies; using saturation bands only where oversampling is not possible; using algorithms capable of recognizing the overlays which exploit the different intensity received by the receiving coils placed at different points of the space (parallel imaging). Noise artifacts from radio frequency (RF) are caused by a malfunction of the RF screen whose purpose is to avoid that the external noise can reach the detector They are due to RF coming from the environment, resulting not only from the radio or television signals, but also by electric motors, by neon lamps placed in the examination room and from the monitoring equipment of the vital functions of the patient, if not MRIcompatible. In general they are not present since the penetration of RF into the MRI room is prevented by the Faraday cage. However, some equipment dedicated to the study of the joints are constructed so that the Faraday cage is internal to the magnet. The way in which this artifact noise occurs in the image depends on the noise source and in which part of the signal has had effect. A receiving coil with non-uniform sensitivity causes an undesired signal change in the image known as RF inhomogeneity artifact. Some RF coils, such as surface coils, have inherent sensitivity variations and almost always show this artifact, while the presence of this artifact in other coils is synonymous with rupture of an element of the coil itself or the presence of a metal in the surveyed area. Metal implants tend to absorb the energy of the RF and, in this way, nearby tissues can't run the right flip angle: the signal will then be reduced with artifact similar to the one from magnetic susceptibility. Both GE and SE sequences suffer the effect of RF inhomogeneity and there is no way to avoid it. Moreover, a distortion of RF field can cause a non-uniform fat suppression. Possible remedies: modify automatic pre-scanning parameters in order to reduce the signal amplification factor (gain); perform a more accurate control of the transmitting elements of the RF coils (RF-shimming); post-processing methods (take too long times). Eddy current artifact is due to eddy currents generated by magnetic field gradients and it is a potential source of B 0 spatial and temporal distortion. It is particularly frequent in the Page 9 of 26

sequences using fast gradient changes, such as EPI (Fig. 12). Three types of distorsions are possible: contraction, displacement and folding. Possible remedies: eddy current compensation; gradient coils screens; contact support. There are a group of artifacts due to hardware and software problems: the most frequent are listed below. Central Point (DC offset) artifact appears as a bright spot in the center of the image and is given by an increase or decrease in intensity and is caused by a constant deviation of the receiver voltage. This disorder is very common since it depends on the fluctuation of the temperature, which cause amplification of the signal. Possible remedies: normally is compensated by special software; make sure the temperature in the room keeps constant. Crisscross looks like an oblique structure (herringbone) that runs through the image in any direction, in one single or more slices. The causes are multiple (electromagnetic surges created by gradients, electronic equipment within the room, AC power fluctuations). Possible remedies: change the bulb; contact support. N/2: the echoes belonging to the rows in even and odd position in the K-space are acquired with read echoes, having opposite directions, and the gradient reversal takes place between a line and the next. Small timing errors in the sampling in relation to gradient variations can cause a modulation in the data location of alternate lines in the K-space (especially in EPI sequences). Possible remedies: change the geometry of the sequence (increase or decrease the angle of the slice); re-shimming; reduce the TE; reduce the phase encoding resolution; Use EPI sequences multi-shot or parallel imaging. Entry slice artifact occurs when rotations without saturation enter for the first time in one or more slices, causing an increase in signal; it is characterized by a light signal coming from a blood vessel (artery or vein) in the first not saturated slice and is often mistaken for thrombosis in SE sequences. Possible remedies: spatial saturation bands positioned before the first slice and after the last one; use of GE techniques are fundamental in the DD between artifact and occlusion. Page 10 of 26

Star (Feed-through) artifact, due to RF coming from the device itself. It is visible in T2w images or with fat signal suppression: it propagates along the direction of encoding phase. Similar to the RF artifact, with the difference that the intensity band of variable signal is unique, of the size of a single pixel and passing exactly through the center of the image. Possible remedies: Make sure that the room door is closed properly; move any mechanical-electrical equipment in the room. Rare and unpredictable artifacts are those from hardware errors, data damage and analogic-to-digital converter block: they can give different results, as they are caused by calculation errors or problems in coil connection. Possible remedies: store data to a new disk; contact support. Page 11 of 26

Images for this section: Fig. 1: The use of chemical shift may be useful to discriminate between malignant or benigne adrenal-gland lesion. The adrenal gland shows a mass-lesion (T1 in-phase; upper image) that loses signal intensity in out of phase T1 (lower image) confirming the benigne nature. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 12 of 26

Fig. 2: In this picture the FID signal artifact follows the underlying anatomy and looks like intensity undulations of the subcutaneous fat. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 13 of 26

Fig. 3: On this brain SE T1 image, Field Inhomogeneity artifact on fronto-temporal right region is shown. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 14 of 26

Fig. 4: "Wave fronts" at regular intervals are formed in Gibbs artifact. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 15 of 26

Fig. 5: A typical example is the light signal present at the level of the tibial inserction of patellar tendon, as shown in this figure. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 16 of 26

Fig. 6: In the images we compare sensibility to distortions from metal implants between T2-SPAIR (on the left), T1-TSE (in the middle). CT scout of the right elbow (on the right). Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 17 of 26

Fig. 7: "Zebra Stripes" due to inhomogeneity of magnetic field are visible at the edges of a large FOV, especially when the patient's boundaries (eg elbows) are very close to the transmitter coil. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 18 of 26

Fig. 8: We have the representation of two tissue (ligament and fluid) in the same voxel in the axial image with partial volume effect, while, in the coronal image, tissues occupy two different voxel and are distinguished as separate entities. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 19 of 26

Fig. 9: Ghost artifact due to cardiac movement along the encoding direction of the leftright phase. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 20 of 26

Fig. 10: Smear artifact due to respiratory motion along the direction of the anteroposterior phase encoding sequence. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 21 of 26

Fig. 11: As shown in the figure aliasing artifact occurs when parts of the object outside the FOV still produce a signal, but the intensity of the applied gradients causes them to gain an upper phase and, therefore, the reconstruction algorithm considers them overlapped at the opposite side of the image. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 22 of 26

Fig. 12: In the upper figure on the left is shown an example of eddy current artifact in EPI image: this blurred image of the brain is not real as is correlated in the sagittal view on the right side. Radiologia, Università del Piemonte Orientale A. Avogadro,, AOU Maggiore della Carità NOVARA - NOVARA/IT Page 23 of 26

Conclusion It's necessary to understand the source of the artifact: some artifacts can be fixed only by adjusting the parameters directly into the MRI Client workstation (typical solution for the physical-chemical artifacts); others are solved using specific equipment (eg. cardiac or respiratory gating, bolus tracking); for others (hardware and software) the only possible solution is to request technical assistance. Page 24 of 26

Personal information A. Cassarà, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY G. Fini, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY T. Bongiovanni, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY F. Cartabbia, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY E. Soligo, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY M. Abruzzese, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY L. Y. Samman, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY A. Stecco, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY A. Carriero, Department of Diagnostic Imaging, AOU Maggiore della Carità, Novara, ITALY Page 25 of 26

References 1. Elster AD, Burdette JH. Questions & Answers in Magnetic Resonance Imaging. 2nd ed. St. Louis: Mosby, 2001; 2. Nguyen TD, Ding G, Watts R, Wang Yi. Optimization of view ordering for motion artifact suppression. Magn Reson Imaging 2001; 19: 951-957; 3. Chang SD, Lee MJ, Munk PL, Janzen DL, MacKay A, Xiang QS. MRI of spinal hardware comparison of conventional T1- weighted sequence with a new metal artifact reduction sequence. Skeletal Radiol 2001; 30: 213-218; 4. Smith RC, Lange RC,McCarthy SM. Chemical shift artifact: dependence on shape and orientation of the lipid-water interface. Radiology 1991; 181: 225-229; 5. Elgavish RA, Twieg DB. Improved depiction of small anatomic structures in MR images using Gaussian-weighted spirals and zero-filled interpolation. Magn Reson Imaging 2003; 21: 103-112; 6. Archibald R, Gelb A. A method to reduce the Gibbs ringing artifact in MRI scans while keeping tissue boundary integrity. IEEE Trans Med Imaging 2002; 21: 305-319; 7. Pusey E, Lufkin RB, Brown RK, et al. Magnetic resonance imaging artifacts: mechanism and clinical significance. Radiographics 1986; 6: 891-911; 8. Jones RW and Witte RJ, "Signal intensity artifacts in clinical MR imaging," Radiographics 2000 May-Jun;20(3):893-901; 9. Mirowitz SA, "MR imaging artifacts. Challenges and solutions," Magn Reson Imaging Clin N Am 1999 Nov;7(4):717-32 10.Heiland, S. Clin Neuroradiol (2008) 18: 25. doi:10.1007/s00062-008-8003-y 11. http://mriquestions.com/hellipmr-artifacts.html 12. http://www.mr-tip.com/serv1.php?type=art Page 26 of 26