Grainger Allison S Diagnostic Radiology 5th Edition Andy Adam Mbbs (Hons) FRCP Frcs FRCR Ffrrcsi (Hon) Ready To Read
Grainger Allison S Diagnostic Radiology 5th Edition Andy Adam Mbbs (Hons) FRCP Frcs FRCR Ffrrcsi (Hon) Ready To Read
https://ebookfinal.com/download/grainger-allison-s-diagnostic-
radiology-5th-edition-andy-adam-mbbshons-frcp-frcs-frcr-ffrrcsihon/
★★★★★
4.8 out of 5.0 (50 reviews )
ebookfinal.com
Grainger Allison s Diagnostic Radiology 5th Edition Andy
Adam Mbbs(Hons) Frcp Frcs Frcr Ffrrcsi(Hon) Pdf Download
EBOOK
Available Formats
https://ebookfinal.com/download/grainger-allison-s-diagnostic-
radiology-2-volume-set-6th-edition-andreas-adam/
https://ebookfinal.com/download/vagabond-vol-29-29-inoue/
https://ebookfinal.com/download/lister-s-the-hand-diagnosis-and-
indications-4th-edition-paul-smith-mbbs-frcsglas-frcseng/
https://ebookfinal.com/download/geographers-biobibliographical-
studies-volume-29-1st-edition-hayden-lorimer/
Physics for Diagnostic Radiology 3rd Edition Dendy
https://ebookfinal.com/download/physics-for-diagnostic-radiology-3rd-
edition-dendy/
https://ebookfinal.com/download/advances-in-photochemistry-
volume-29-1st-edition-douglas-c-neckers/
https://ebookfinal.com/download/batik-gems-29-dazzling-quilt-
projects-1st-edition-laurie-j-shifrin/
https://ebookfinal.com/download/medical-pharmacology-and-
therapeutics-4e-4th-edition-derek-g-waller-bsc-dm-mbbs-frcp/
https://ebookfinal.com/download/studies-in-symbolic-interaction-
vol-29-1st-edition-norman-k-denzin-ed/
Grainger Allison s Diagnostic Radiology 5th Edition Andy
Adam Mbbs(Hons) Frcp Frcs Frcr Ffrrcsi(Hon) Digital
Instant Download
Author(s): Andy Adam MBBS(Hons) FRCP FRCS FRCR FFRRCSI(Hon),
Adrian K. Dixon MD FRCR FRCP FRCS FMedSci FFRRCSI(Hon)
FRANZCR(Hon)
ISBN(s): 9780443101632, 0443101639
Edition: 5th Edition
File Details: PDF, 129.18 MB
Year: 2007
Language: english
Contributors
†
: deceased
Andy Adam MBBS (Hons), FRCP, FRCS, Clive I. Bartram FRCS, FRCP, FRCR Gisele Brasil Caseiras PhD
FRCR, FFRRCSI (Hon) Emeritus Consultant, St Mark’s Hospital Research Fellow
Professor of Interventional Radiology and Honorary Professor of Insitute of Neurology
Department of Radiology Gastrointestinal Radiology University College London
St Thomas’ Hospital Faculty of Medicine London, UK
King’s College London Imperial College
London, UK London, UK Jackie E. Brown BDS, MSc, FDSRCP, DDRRCR
Consultant Oral and Maxillofacial
E. Jane Adam MBBS, MRCP, FRCR Philip P. W. Bearcroft FRCP, FRCR Radiologist
Consultant Radiologist Consultant Radiologist Kings College London Dental Institute
Department of Radiology Department of Radiology Guy’s Dental Hospital
St George’s Hospital Cambridge University Hospitals NHS London, UK
London, UK Foundation Trust
Addenbrooke’s Hospital Dina F. Caroline MD, PhD
Judith E. Adams MBBS, FRCR, FRCP Cambridge, UK Professor Emerita
Chair, Diagnostic Radiology Department of Radiology
Imaging Science and Biomedical Anna-Maria Belli DMRD, FRCR Temple University Hospital
Engineering Consultant Vascular Radiologist and Philadelphia
University of Manchester Reader in Radiology Pennsylvania, USA
Honorary Consultant Radiologist Department of Radiology
Royal Infirmary St George’s Hospital Silvia D. Chang MD, FRCP(C)
Manchester, UK London, UK Assistant Professor
University of British Columbia
David J. Allison BSc, MD, MRCS, LRCP, MBBS, Anthony R. Berendt BM, BCh, FRCP Department of Radiology
DMRD, FRCR, FRCP Consultant Physician-in-Charge Vancouver General Hospital
Emeritus Professor of Imaging Bone Infection Unit Vancouver, Canada
Imperial College Nuffield Orthopaedic Centre
London, UK Oxford, UK W. K. ‘Kling’ Chong BMedSci, MD, MRCP, FRCR
Consultant Neuroradiologist
Sandra Allison MD Lol Berman FRCP, FRCR Department of Radiology
Assistant Professor of Radiology University Department of Radiology Great Ormond Street Hospital for
Director, Radiology Residency Program Addenbrooke’s Hospital Children NHS Trust
Director, Ultrasound Cambridge, UK London, UK
Georgetown University Hospital
Washington DC, USA Martin J. K. Blomley† Bairbre Connolly MBBCh, BAO, FRCSI, MCH,
FFRRCSI, FRCP(C)
Philip Anslow FRCR Carol A. Boles MD Medical Director and Division Head of
Consultant Neuroradiologist Associate Professor of Radiology Image Guided Therapy
Department of Radiology Associate, Surgical Sciences Pediatric Interventional Radiologist
Radcliffe Infirmary Orthopedic Surgery Assistant Professor, University of Toronto
Oxford, UK Wake Forest University Department of Diagnostic Imaging
North Carolina, USA The Hospital for Sick Children
Susan M. Ascher MD Toronto, Canada
Georgetown University Medical Center Jamshed B. Bomanji MBBS, MSc, PhD
Washington DC, USA Consultant in Nuclear Medicine Susan J. Copley MBBS, MD, MRCP, FRCR
UCLH Trust Consultant Radiologist and Honorary
Zelena A. Aziz MD, MRCP, FRCR Middlesex Hospital Senior Lecturer
Consultant Radiologist London, UK Radiology Department
Department of Radiology Hammersmith Hospital
London Chest Hospital London, UK
London, UK
x CONTRIBUTORS
David O. Cosgrove MA, MSc, FRCP, FRCR Robert J. Eckersley PhD Philip C. Goodman MD
Emeritus Professor Research Associate Professor of Radiology
Imaging Sciences Department Imaging Sciences Department Chief, Thoracic Imaging Division
Faculty of Medicine Faculty of Medicine Department of Radiology
Imperial College Imperial College Duke University Medical Center
Hammersmith Hospital Hammersmith Hospital Durham
London, UK London, UK North Carolina, USA
Nigel Cowan PGDipLATHE, FRCP Andrew J. Evans MRCP, FRCR Isky Gordon FRCR, FRCP
Oxford, UK Consultant Radiologist Professor of Paediatric Imaging
Nottingham Breast Institute Institute of Child Health
Justin J. Cross MRCP, FRCR Nottingham City Hospital London, UK
Consultant Neuroradiologist Nottingham, UK
Department of Radiology Nicholas Gourtsoyiannis FRCR (Hon)
Addenbrooke’s Hospital Jane Evanson BSc, MBBS, MRCP, FRCR Professor of Radiology
Cambridge, UK Consultant Neuroradiologist, Barts & The University of Crete
London Hospital NHS Trust Faculty of Medicine
Paras Dalal BSc, MRCP, FRCR The Royal London Hospital Heraklion, Crete
Research Fellow in Thoracic Imaging London, UK Greece
Department of Radiology
Royal Brompton Hospital Laura Fender BMedSci, BMBS, MRCP, FRCR Andrew J. Grainger BM, BS, MRCP, FRCR
London, UK Consultant Radiologist Consultant Radiologist
Nottingham University Hospital Chapel Allerton Orthopaedic Centre
Maria Daskalogiannaki MD Nottingham, UK Leeds Teaching Hospitals
Registrar in Radiology Leeds, UK
Department of Radiology Alan H. Freeman MBBS, FRCR
University Hospital of Heraklion Consultant Radiologist Ronald G. Grainger MB ChB(Hons), MD, FRCP,
Crete, Greece Department of Radiology DMRD, FRCR, FACR(Hon), FRACR(Hon)
Addenbrooke’s Hospital Professor of Diagnostic Radiology
A. Mark Davies FRCR Cambridge, UK (Emeritus)
Consultant Radiologist University of Sheffield
Royal Orthopaedic Hospital Julia Gates MD Honorary Consultant Radiologist
Birmingham, UK Assistant Professor of Radiology Royal Hallamshire Hospital and Northern
Department of Radiology General Hospital
Adrian K. Dixon MD, FRCR, FRCP, FRCS, Tufts University School of Medicine Sheffield, UK
FMedSci, FFRRCSI (Hon), FRANZCR (Hon) Springfield
Professor of Radiology Massachusetts, USA Philippe Grenier FRCR (Hon)
Department of Radiology Professor of Radiology
Addenbrooke’s Hospital Robert N. Gibson MBBS, MD, FRANZCR, DDU Service de Radiologie Polyvalente
University of Cambridge Professor of Radiology Diagnostique et Interventionnelle
Cambridge, UK Department of Radiology Hôpital Pitié-Salpêtrière
University of Melbourne Paris, France
Rose de Bruyn DMRD, FRCR Royal Melbourne Hospital
Consultant Radiologist Victoria, Australia Roxana S. Gunny BS, BSc, MRCP, FRCR
Department of Radiology Consulant Neuroradiologist
Great Ormond Street Hospital for Sick Raymond J. Godwin MA, MB, Bchir, FRCP, FRCR Department of Radiology
Children NHS Trust Consultant Radiologist Great Ormond Street Hospital for
London, UK Department of Radiology Children NHS Trust
West Suffolk Hospital London, UK
Claudio Defilippi MD Suffolk, UK
Consultant Radiologist Christine M. Hall MBBS, DMRD, FRCR MD
Department of Radiology Karen Goldstone BSc, MSc, Csci, FIPEM Professor of Paediatric Radiology
OIRM - S. Anna Hospital Radiation Protection Advisor Great Ormond Street Hospital for
Turin, Italy Acting Head of Department of Medical Children NHS Trust
Physics and Clinical Engineering London, UK
Sujal R. Desai MD, MRCP, FRCR East Anglian Regional Radiation
Consultant Radiologist Protection Service (EARRPS) David M. Hansell MD, FRCP, FRCR, LRSM
Department of Radiology Addenbrooke’s Hospital Professor of Thoracic Imaging
King’s College Hospital Cambridge, UK Department of Radiology
London, UK Royal Brompton Hospital
London, UK
CONTRIBUTORS xi
George G. Hartnell FRCR, FRCP Adrian K. P. Lim MD, FRCR Stuart E. Mirvis MD, FACR
Director of Cardiovascular and Interven- Consultant Radiologist and Professor of Radiology
tional Radiology Senior Lecturer Department of Radiology
Department of Radiology Imaging Sciences Department University of
Baystate Medical Center Faculty of Medicine Imperial College Maryland School of Medicine
Springfield Hammersmith Hospital Baltimore
Professor of Radiology London, UK Maryland, USA
Tufts University Medical School
Boston David J. Lomas MA, MB, BChir, FRCR, FRCP Sameh K. Morcos FRCS, FFRRCSI, FRCR
Massachusetts, USA Professor of Clinical Magnetic Resonance Professor of Diagnostic Imaging
Imaging University of Sheffield
Hedvig Hricak MD, Dr. Med, SC, Dr. h.c, FRCR (Hon) Department of Radiology Consultant Radiologist
Chairman, Department of Radiology Addenbrooke’s Hospital Department of Diagnostic Imaging
Carroll and Milton Petrie Chair Cambridge, UK Northern General Hospital
Professor of Radiology, Weill Medical Sheffield, UK
College of Cornell University Sharyn L. S. MacDonald MBChB, FRANZCR
Memorial Sloan-Kettering Cancer Center Consultant Radiologist Robert A. Morgan MBChB, MRCP, FRCR
New York, USA Department of Radiology Consultant Radiologist
Christchurch Hospital Department of Radiology
James E. Jackson MRCP, FRCR Christchurch, New Zealand St George’s Hospital
Consultant Radiologist London, UK
Department of Imaging David MacVicar MA, MRCP, FRCP, FRCR
Hammersmith Hospital Consultant Radiologist Iain Morrison MBBS, MRCP, FRCR
London, UK Department of Diagnostic Radiology Consultant Radiologist
Royal Marsden Hospital Radiology Department
H. Rolf Jäger FRCR, MD Sutton, UK Kent and Canterbury Hospital
Reader in Neuroradiology Canterbury, UK
Institute of Neurology Adrian Manhire BSc, MBBS, FRCP, FRCR
University College London Consultant Radiologist Nestor L. Müller MD, PhD, FRCPC
Honorary Consultant Neuroradiologist Nottingham City Hospital Professor and Chairman
The National Hospital for Neurology Nottingham, UK Department of Radiology
and Neurosurgery and University College University of British Columbia
Hospital Tarik F. Massoud MA, MD, PhD, FRCR Head and Medical Director
London, UK University Lecturer and Honorary Department of Radiology
Consultant in Neuroradiology Vancouver General Hospital
Jonathan J. James BMBS, FRCR University Department of Radiology Vancouver, Canada
Consultant Radiologist University of Cambridge School of
Nottingham Breast Institute Clinical Medicine Graham Munneke MRCP, FRCR
Nottingham City Hospital Addenbrooke’s Hospital Consultant in Interventional Radiology
Nottingham, UK Cambridge, UK Department of Radiology
St. George’s Hospital
Renee M. Kendzierski DO Kieran McHugh FRCPI, DCH, FRCR London, UK
Assistant Professor of Radiology Department of Radiology
Department of Radiology Great Ormond Street Hospital for Sick Alison D. Murray MB ChB (Hons), FRCR, FRCP
Temple University Hospital Children NHS Trust Senior Lecturer in Radiology
Philadelphia London, UK Department of Radiology
Pennsylvania, USA University of Aberdeen
James Meaney FRCR, FFRRCSI Aberdeen, UK
Dow-Mu Koh MRCP, FRCP Director of MRI
Senior Lecturer and Honorary Consultant St James’s Hospital Richard A. Nakielny FRCR
Department of Radiology Dublin, Ireland Honorary Clinical Lecturer
Royal Marsden Hospital Directorate of Medical Imaging &
Sutton, UK Hylton B. Meire FRCR, DRCOG, DMRD Medical Physics
Consultant Radiologist (Retired) Royal Hallamshire Hospital
Isla Lang MBChB, MRCP, FRCR King’s College Hospital Sheffield, UK
Consultant Paediatric Radiologist London, UK
Sheffield Children’s Hospital Hrudaya Nath MD
Sheffield, UK Kenneth A. Miles MD, FRCR, MSc, FRCP Professor of Radiology
Clinical Imaging Sciences Centre Department of Radiology
Brighton and Sussex Medical School University of Alabama Hospitals
University of Sussex Birmingham
Falmar, Brighton, UK Alabama, USA
xii CONTRIBUTORS
Tony Nicholson MSc, FRCR Sheila Rankin FRCR John Rout BDS, FDSRCS, MDentSc, DDRRCR, FRCR
Consultant Vascular Radiologist Consultant Radiologist Consultant Oral and Maxillofacial
Department of Clinical Radiology Department of Radiology Radiologist
Leeds Teaching Hospitals Guy’s and St. Thomas’ Foundation Trust Birmingham Dental Hospital
Leeds, UK London, UK Birmingham, UK
Amaka C. Offiah BSc, MBBS, MRCP, FRCR, PhD Padma Rao MBBS, BSc, MRCP, FRCR, Michael B. Rubens MB, DMRD, FRCR
Consultant Radiologist (Academic) FRANZCR Consultant Radiologist
Great Ormond Street Hospital for Consultant Paediatric Radiologist Department of Radiology
Children NHS Trust Royal Children’s Hospital Royal Brompton Hospital
London, UK Parkville London, UK
Melbourne
Simon Padley BSc, MBBS, FRCP, FRCR Victoria, Australia Asif Saifuddin BSc (Hons), MBChB, MRCP, FRCR
Consultant Radiologist Consultant Radiologist
Department of Radiology Christine Reek BSc, FRCR Department of Radiology
Chelsea and Westminster Hospital Consultant Radiologist Royal National Orthopaedic Hospital
London, UK Department of Radiology NHS Trust
Greenfield Hospital Stanmore, UK
Martyn N. J. Paley PhD, FInstP Leicester, UK
Professor of MR Physics Evis Sala MD, PhD, FRCR
Academic Radiology John H. Reynolds DMRD, FRCR, MMedSci University Lecturer in Oncology Imaging
University of Sheffield Consultant Radiologist University Department of Radiology
Sheffield, UK Birmingham Heartlands Hospital Addenbrooke’s Hospital
Birmingham, UK Cambridge, UK
Nickolas Papanikolaou PhD
Biomedical Engineer Rodney H. Reznek FRANZCR (Hon), FRCP, FRCR Caron Sandhu FRCR
Department of Radiology Professor of Diagnostic Imaging Consultant Radiologist
University Hospital of Heraklion The Centre for Cancer Imaging Department of Radiology
Crete, Greece St Bartholomew’s Hospital and The Guy’s and St. Thomas’ Hospital
London Queen Mary’s School of London, UK
Jai Patel MBChB, MRCP, FRCR Medicine and Dentistry
Consultant Vascular Radiologist London, UK Dawn Saunders MD, MRCP, FRCR
Department of Clinical Radiology Consultant Neuroradiologist
St James’s University Hospital Philip M. Rich BSc, FRCS, FRCR Department of Radiology
Leeds, UK Consultant Neuroradiologist Great Ormond Street Hospital for
Department of Neuroradiology Children NHS Trust
Anne Paterson MBBS, MRCP, FRCR, FFR RCSI Atkinson Morley Wing London, UK
Consultant Paediatric Radiologist St George’s Hospital
Radiology Department London, UK Daniel J. Scoffings MRCP, FRCR
Royal Belfast Hospital for Sick Children Specialist Registrar in Neuroradiology
Belfast, UK Andrea Rockall MD, BS, BSc, MRCP, FRCP Addenbrooke’s Hospital
Department of Radiology Cambridge, UK
Praveen Peddu MRCS, FRCR St Bartholomew’s Hospital
Specialist Registrar in Radiology London, UK Djilda Segerman MA, MSc, MIPEM
Department of Radiology Head of Nuclear Medicine Physics
King’s College Hospital Giles Roditi FRCP, FRCR Department of Medical Physics
London, UK Consultant Radiologist Brighton and Sussex University Hospitals
Department of Radiology NHS Trust
A. Michael Peters BSc, MD, MSc, MRCP, Glasgow Royal Infirmary Brighton, UK
MRCPath, FRCR Glasgow, UK
Professor of Nuclear Medicine Kathirkama Shanmuganathan MD
Brighton and Sussex Medical School Lee F. Rogers MD Associate Professor of Radiology
University of Sussex Clinical Professor of Radiology Department of Radiology
Brighton, UK Department of Radiology University of Maryland School of
University of Arizona Health Services Medicine
William H. Ramsden BM, FRCR Tucson Baltimore
Consultant Paediatric Radiologist Arizona, USA Maryland, USA
Department of Clinical Radiology
St James’s University Hospital Giles Rottenberg FRCR Ashley S. Shaw MRCP, FRCR
Leeds, UK Consultant Radiologist Consultant Radiologist
Department of Radiology Department of Radiology
Guy’s and St. Thomas’ Foundation Trust Addenbrooke’s Hospital
London, UK Cambridge, UK
CONTRIBUTORS xiii
Satinder P. Singh MD, FCCP Mihra S. Taljanovic MD, MA Sarah J. Vinnicombe BSc, MRCP, FRCR
Associate Professor of Radiology Associate Professor of Clinical Radiology Consultant Radiologist
Director Cardiac CT and Clinical Orthopedic Surgery Department of Radiology
Director, Combined Cardiopulmonary Head - Musculoskeletal Imaging Section St Bartholomew’s Hospital
and Abdominal Fellowship Department of Radiology London, UK
Chief of Cardiopulmonary Radiology Tucson
Department of Radiology Arizona, USA Gustav K. von Schulthess MD, PhD
University of Alabama Hospitals Professor and Director
Birmingham Andrew M. Taylor BA (Hons), BM BCh, MRCP Department of Radiology
Alabama, USA FRCR University Hospital
Consultant Radiologist Zurich, Switzerland
S. Aslam A. Sohaib MRCP, FRCR Department of Clinical Radiology
Radiology Department Great Ormond Street Hospital for Iain D. Wilkinson BSc, MSc, PhD, CSci,
Royal Marsden Hospital Children NHS Trust ARCP, FIPEM
London, UK London, UK Reader in Magnetic Resonance
& Consultant Clinical Scientist
Alan Sprigg MBChB, DCH, DRCOG, DMRD, FRCR, Stuart Taylor BSc, MD, MRCP, FRCR Academic Radiology
FRCPCH Consultant Radiologist University of Sheffield and
Consultant Paediatric Radiologist Department of Intestinal Imaging Sheffield Teaching Hospitals
Sheffield Children’s Hospital St Marks Hospital NHS Foundation Trust
Sheffield, UK Northwick Park Sheffield, UK
Harrow, UK
John M. Stevens MBBS, DRACR, FRCR A. Robin M. Wilson FRCR, FRCP(E)
Lyshom Department of Neuroradiology Henrik S. Thomsen MD Consultant Radiologist
Radiology Department Professor and Chairman King’s College Hospital and
The National Hospital for Neurology and Department of Diagnostic Radiology Guy’s and St Thomas’ Foundation Trusts
Neurosurgery Copenhagen University Hospital London, UK
London, UK Herlev, Denmark
David J. Wilson MBBS, BSc, FRCP, FRCR
Dennis J. Stoker MB, FRCP, FRCS, FRCR Paolo Toma MD Consultant Musculoskeletal Radiologist
Emeritus Consultant Radiologist Radiologist-in-Chief Nuffield Orthopaedic Centre and
Henley-on-Thames, UK Radiology Department University of Oxford
G. Gaslini Institute Oxford, UK
Nicola H. Strickland BM BCh, MA (Hons), Genoa, Italy
(Oxon), FRCP, FRCR Stuart J. Yates MSci, MSc, CSci, MIPEM
Consultant Radiologist Peter Twining FRCR, BSc, BS MB Principal Physicist
Imaging Department Consultant Radiologist Department of Medical Physics & Clinical
Hammersmith Hospitals NHS Trust Nottingham University Hospital Engineering
Honorary Senior Lecturer Nottingham, UK Cambridge University Hospitals NHS
Imperial College Foundation Trust
London, UK John A. Verschakelen MD, PhD Cambridge, UK
Professor of Chest Radiology
Louise E. Sweeney MBBCH, BAO, DCH, DMRD, Department of Radiology
FRCR, FFR, RCSI University Hospitals Gasthuisberg
Consultant Paediatric Radiologist Leuven, Belgium
Radiology Department
Royal Belfast Hospital for Sick Children
Belfast, UK
Preface
We hope that this 5th edition of Diagnostic Radiology will year basic course followed by two years of training in selected
continue to build on the original vision of Professors Grainger subspecialties, the factual examination in the UK has moved to
and Allison who, back in the early 1980s, saw the need for an earlier stage in training with less emphasis on some of the
a ‘bible’ for doctors studying for postgraduate examinations diagnostic rarities so beloved by examiners of old. The cur-
in radiology, and to provide a bench book for reporting riculum is now somewhat less comprehensive and the reduc-
and reference. The success of the first four editions, which tion in size of this 5th Edition reflects that – down from three
were extremely well received by an increasingly interna- volumes to two. In this electronic age there are many data-
tional readership, speaks for the realisation of their dream. bases of images available on the internet, with accompanying
Few could have predicted at that stage the extraordinary text. Nevertheless, we believe that well structured textbooks
growth of radiology, or its increasing importance within all remain an essential part of medical education and practice as
aspects of modern medicine. The unprecedented expansion in they present information in a format that facilitates learning,
the imaging repertoire, together with the trend for increasing guiding the reader through an unfamiliar field. We are con-
subspecialisation, have led to changes in training and in the vinced that Diagnostic Radiology will remain a valuable resource
methods used for teaching and learning. This book has had for many years to come.
to evolve to reflect these changes, adapting to the perceived We are again extremely grateful to the distinguished inter-
needs of those facing postgraduate examinations and also to national cast of authors who have all worked hard to deliver
all radiologists who wish to have an up-to-date basic general fresh and up-to-date material. We are also most grateful to
textbook for ready reference and illustration. Michael Houston, Gavin Smith and Nora Naughton for their
An attempt to cover every subject in detail would have professional skills and publishing expertise and to Jeremy
resulted in a huge book that would have been very difficult to Rabouhans for invaluable help with proof reading. We could
use.We have chosen to concentrate on those subjects that most never have done it without them!
radiologists need to know well, and to pay special attention to
the needs of trainee radiologists preparing for examinations. Andy Adam
Because training throughout Europe is moving towards a three Adrian Dixon
Acknowledgements
This edition could not have occured without the large amount leagues remained remarkably cheerful throughout and kept
of work done by all the contributing authors and their col- strong heads even when chapters were late, images missing,
leagues. However, the vision and overall planning of Michael and all the other hiccups that can hinder progress in a project
Houston at Elsevier have been fundamental in bringing the of this kind. At a local level, all the Editors would like to thank
book to fruition. So, too, has the meticulous gathering and their various secretaries, technicians and colleagues who have
editing of material by Gavin Smith. Finally, the skilful copy- helped proofread, collect material and made various other
editing, and other tasks provided by Nora Naughton, and invaluable contributions.
her remarkable team, must not be forgotten; without them
the Editors simply could not have managed. All of these col- The Editors
1
CHAPTER
Nicola H. Strickland
Were these the only roles of a PACS, it would be an extremely ing upon a potential time scale in which the hospital-wide
complex and expensive means of replacing traditional film. A PACS is to be achieved, and deciding upon a scaleable PACS
PACS must improve upon a film-based system, preferably in architecture. Ideally the whole hospital infrastructure should
a cost-neutral manner. The major added value of a PACS is be adapted at the outset so that a hospital-wide PACS can
efficiency of data management. be accommodated at a later stage. This includes providing
True efficiency benefits can only be realized once a PACS an uninterruptable power supply (UPS), allowing sufficient
is at least hospital-wide, since any more limited installation cabling space in floors and ceilings, and adapting the air con-
means running two systems in parallel, i.e. it entails continu- ditioning system for PACS. The hospital information tech-
ing to produce conventional film and moving it around the nology (IT) network is likely to need upgrading to enable
hospital, as well as the cost of installing and maintaining a large amounts of image data to be transported, and it is advis-
PACS. Thus, even if funds are limited initially, it is advisable able to install multiple PACS ‘drops’ (workstation outlets) so
at least to aim and plan for growing the PACS installation that more PACS workstations can easily be added at a later
into a hospital-wide system ultimately. This means decid- date.
has been installed, before the film store can be dispensed DISADVANTAGES OF PACS
with altogether. (In the UK it is the radiological report,
not the images that are deemed to be the legal document, The advantages of PACS described earlier must be set against
with certain restrictions for paediatric and educationally its potential disadvantages:
challenged patients.) • Even though the costs of hardware and storage media
• The installation of a PACS infrastructure in an institution (a continue to reduce in price, PACS remains an expensive
local area network [LAN]) sets the stage for the introduc- technology. Most estimates suggest that a PACS installation
tion of teleradiology over a wide area network (WAN) if should aim at becoming cost-neutral in less than 5 years.
desired. Teleradiology offers the potential for improvements Some would argue that PACS should be viewed in the same
in efficiency, for example in geographically remote areas by way as any new imaging technique, and as such it represents
centralizing a reporting service, or increasing the referrals to an advance in health care management and should not be
a particular institution. assessed merely in terms of cost–benefit.
• The technological complexity of a PACS and the absolute
A number of perceived potential benefits of PACS have not dependency of a hospital on the PACS once it becomes
been substantiated, or not consistently demonstrated, by filmless require a dedicated maintenance programme for the
audit studies. These include the possibility of a reduced hos- PACS and a carefully devised plan detailing how to supply
pital inpatient stay or a greater throughput of outpatients6. a minimal clinical service should the whole PACS fail for a
It is hardly surprising that such benefits cannot be attrib- significant period of time. This inevitably means that there
uted to PACS since there are so many other variable factors will be a requirement for new or retrained hospital personnel
that influence these issues. Also the type of study required specializing in computer engineering/information technol-
to prove any given putative benefit of PACS is fraught with ogy, as well as a vendor-provided maintenance service, and
practical difficulties. The classical ‘before’ and ‘after’ study these costs must be set against the savings made in respect of
comparing the pre-PACS with the post-PACS era is inevi- less highly paid clerical staff and darkroom technicians.
tably complicated by the numerous other concurrent envi- • Once a hospital-wide PACS is in operation and film has
ronmental, technological (and often political) changes that been withdrawn, there is no ‘fall-back position’. The hospi-
have taken place in the interim. These include, for example, tal is no longer equipped to run a film-based service. This is
changes in the medical personnel and in the type of clinical a daunting prospect that may act as an initial psychological
practice pursued in the hospital under study. Studies com- deterrent to embarking upon a large-scale PACS project.
paring a PACS institution with a nonPACS institution may • Changing from a hard-copy to a soft-copy imaging environ-
be similarly flawed by the difficulty in adequately correct- ment will raise many issues necessitating a change of work
ing for other inherent differences between the two institu- patterns involving: the training of the users, maintenance of the
tions, which may or may not be related to the presence of a system, action to be taken in the event of a PACS failure and
hospital-wide PACS7. the institution of specific quality assurance protocols (see later).
It is important to define what is meant by ‘the life of the a private company (with a maintenance contract). Nothing
system’. This is often taken as being 8–10 years after the com- is owned by the hospital, i.e. the private company retains all
pletion date.The majority of PACS installations are now based the assets. The advantage of a leasing arrangement is that it
on some form of leasing agreement, rather than an outright allows hospitals with no hope of ever having the large capital
capital purchase of the hardware and software. The date at sum necessary to purchase a hospital-wide PACS to make a
which the maintenance contract will commence, its cost and quantum leap in technology to move to a PACS solution, and
its terms, all need careful definition. Responsibility must be it transfers the risk to the company. It has to be appreciated,
defined for migration of the PACS archive data at the end of as with renting a house, that the hospital would be left with
the PACS contract, or in the event of premature termination nothing if it were to stop paying the lease. At the end of the
of the contract. contractual term, the hospital does not own anything. How-
It is advisable to define exactly what is meant by an ‘update’ ever, in these circumstances the hospital is usually given an
to the system (generally a minor software release that merely option to buy the installed assets (the PACS hardware and soft-
corrects bugs in the system), and what is meant by an ‘upgrade’ ware) at ‘a fair market value’, as negotiated with the provider.
(which constitutes a new software installation comprising In a leased service the PACS company provides the PACS
major new features giving enhanced functionality), whether hardware and software necessary to meet the workload and
these are to be included in the purchase price/maintenance performance requirements stipulated by the hospital, both at
agreement, and at what frequency these will occur. An impor- the time of leasing and in the future. For example, such require-
tant consideration in negotiation with a vendor is the terms ments might include the need to perform 400 000 imaging
under which the delivered PACS hardware will be replaced if examinations per year, with a short-term storage of a year, and
upgrades are released that it cannot support. a display time for these examinations (first image to screen) of
It is vital to define which party is ultimately responsible 3 s or less.The PACS company is responsible for maintaining a
for the functioning of the interfaces to the various pieces of PACS with an agreed level of technology throughout the con-
imaging equipment, and to other hospital computer systems tractual term. The technological risk rests with the provider.
including the hospital and radiological information systems The risk covers the following three areas:
(HIS and RIS), electronic patient record (EPR), speech rec- • Utilization of the system
ognition dictation system, and electronic remote requesting • Availability of the system to users
(order communications) system, if these are already extant in • Future planning and implementation of new technology
the institution. into the hospital’s system.
for example, of acquiring a new 64-slice (or greater) multidetec- responsibility to list the make and model of all the equipment
tor CT machine, the intention to start performing high image possessed by the hospital/health care facility, and the PACS
acquisition studies in MR such as cardiac or breast imaging vendor’s responsibility to look up and interpret the DICOM
and so forth. The predicted image storage requirement should conformance statements and to make it clear which pieces
always be an overestimate to allow for unexpected demand. It is of equipment will need upgrading/replacing, and to explain
also advisable to know which outpatient clinics are performed the connectivity limitations if such DICOM upgrades are
when, and how many film packets are pulled for each, to give not undertaken. One of the major causes of interoperability
some estimate of the network traffic to be expected. It is also problems when linking equipment from different vendors to a
important to define the role of the imaging department in other PACS is the use of ‘private DICOM attributes’ by many ven-
imaging-related activities, such as the radiological steps involved dors, which, in simple terms, means that information stored
in conducting multidisciplinary team meetings (MDTMs) and in these particular private DICOM fields is not available to
radiological presentations at staff rounds, and undergraduate be shared with other apparatus, manufactured by a different
and postgraduate teaching sessions. PACS must be able to fulfill vendor, which may be linked to it on a network. This leads to
all these functions. a loss of functionality on the recipient apparatus, for example
The Imaging Directorate would be wise to know before not being able to post-process scanner images received on a
PACS is installed, exactly how many imaging studies are never workstation from a different vendor.
reported (for various reasons such as: the film packets are never
returned to the department for reporting), the time between ACRONYMS: DICOM, HL7 AND IHE
image examination acquisition and dictation of a report by a
radiologist for each investigation and the time delay (if any) DICOM stands for digital communication in medicine and
between dictation of the report and the availability of the ver- refers to a worldwide multipart standard to which all modern
ified report to other clinicians. These workflow deficiencies imaging equipment and PACS must adhere, and has now been
need to be addressed and corrected prior to the installation extended to other disciplines, including cardiology, endoscopy,
of a PACS, since the PACS will not ameliorate the situation and ophthalmology. The DICOM conformance statement
but instead will expose these issues by making such informa- of every piece of modern imaging equipment is obligatorily
tion available in computerized form throughout the PACS available on the Internet, and the description of the various
institution. DICOM attributes possessed by each appliance will predict its
Seamless integration with other IT systems, and with connectivity with another piece of equipment. All apparatus
imaging acquisition devices, is absolutely vital for a PACS to is described as being a ‘user’ or a ‘provider’ of services such as
function successfully, and every modern PACS depends upon storage.The DICOM standard is continuously under develop-
adherence to DICOM and HL7 standards and IHE (see next ment, but each new part of the DICOM standard has to be
paragraph for an explanation of each) workflow processes backwardly compatible with the current DICOM standard.
to achieve this integration13. Before installing PACS, old HL7 stands for health level 7 and refers to a worldwide
equipment and old data information systems will need to be standard for data information systems such as HIS and RIS. It
upgraded to a minimum level of DICOM and HL7 compli- is a less rigorous standard than DICOM. HL7 messages from
ance, respectively. Often it is cheaper to replace such products data information systems are conveyed to DICOM apparatus
with modern versions, rather than to pay to upgrade them. A (including PACS) by a ‘broker’, which acts as an integrating
full inventory must be made of the equipment and IT systems and translation device.
to be connected to the PACS with a precise description of IHE stands for integrating the health care enterprise and
the level (if any) of DICOM or HL7 compliance supported, is not a standard, but a comprehensive workflow descriptor
before a PACS project can be properly planned and costed. of how processes, such as reporting, for example, are achieved
The full DICOM conformance statement of every piece of in imaging. Its use of integration profiles eliminates the need
DICOM-compliant equipment (e.g. a computed tomography to reconcile the details of HL7 messages and DICOM con-
[CT] scanner, an ultrasound [US] machine, a workstation formance statements among multiple vendors. It is now being
etc.) is available on the Internet. It is generally the users’ expanded outside the discipline of imaging.
A realistic definition of the expected dates of completion were underexposed on the original film remain underexposed
of each milestone (including installation dates), with penalty when digitized; areas that were overexposed on the original
clauses written into the contract to come into effect if these film would normally yield more information when viewed
dates are overrun, protects the institution from major delays over a bright light, but tend just to look black when digitized.
since penalties can be extracted from the vendor if such delays Digitization on a large scale is not a practical option and com-
occur. bines the expense of a film budget with the cost of digitization
and maintenance of a soft-copy archive. Digitization may be
IMPLEMENTATION OF DIGITAL IMAGE a viable solution in a very-low-throughput teleradiology ser-
ACQUISITION PRIOR TO PACS vice, and is necessary in a PACS environment to allow hard-
copy films acquired at a nonPACS institution to be entered
Plain radiography (chest, abdominal and skeletal plain images) onto the PACS for comparison with subsequent digital images
still constitute the majority (usually 60 per cent) of imaging acquired at the PACS institution.
examinations in most general radiology departments. These Since digitization is a costly and labour-intensive process,
examinations therefore need to be acquired in a digital format yielding digital images of suboptimal quality, a sensible mini-
in order to be transferred to PACS. This represents a consider- malist digitization policy should be adopted when PACS is
able challenge since conventional plain film work is still such a installed. A reasonable compromise is only to digitize one
significant part of the total departmental workload. The other previous ‘key’ image in patients for whom it is regarded as
imaging investigations (computed tomography [CT], magnetic clinically highly desirable to have a comparison image avail-
resonance [MR], nuclear medicine [NM], positron emission able long term on the PACS archive (for example, oncology
tomography [PET], US, digital subtraction angiography, and patients on chemotherapy, or patients with slow-growing
fluoroscopy) are either already digital in nature at acquisition or tumours or chronic disease). Some institutions choose not to
can be rendered so by screen capture/frame grabbing from the digitize any previous films, but undertake to make past hard-
acquisition device, and can thus be easily transferred to PACS. copy imaging available on request (from off-site storage) for
There are three basic means of rendering plain radiographic over a year. It is clear that unselective ‘blind’ digitization of
images digital: batches of previous films is a waste of time and money consid-
1 Digitizing conventional analogue film ering how few previous imaging studies are ever subsequently
2 Photostimulable phosphor plate technology, commonly known reviewed4.
as computed radiography (CR).
3 Direct digital radiography. Photostimulable phosphor plate technology/
computed radiography
There is considerable new technology associated with acquir- This is the commonest means of acquiring plain radiographic
ing plain radiographic images digitally, whichever method is images in a digital format. Photostimulable phosphor plates
chosen. In situations other than the opening of a new hospi- replace the conventional film/screen combination in a cassette
tal/health care facility, it is circumspect not to introduce this and can be used with standard X-ray equipment (although
new technology concurrently with a PACS to avoid the risk not for fluoroscopy). X-ray photons falling on the phosphor
of ‘technology overload’ for the users. Once the bulk of the plate produce a latent image by promoting electrons to meta-
imaging studies (i.e. the plain radiography) is being acquired stable traps of higher energy level. When the plate is ‘read out’
and stored digitally, the introduction of PACS and the complete in raster fashion by a scanning laser beam in the plate reader,
withdrawal of film is a less daunting task. The overlap period, these electrons gain sufficient energy to fall back to their basal
in which a digital archive is being built up whilst continu- energy level, and in so doing, emit photons of visible light that
ing to distribute film outside the imaging department, should is converted to a final digital signal (Fig. 1.1).
be kept to a minimum for financial reasons: it is obviously Phosphor plates have a flatter, more linear, optical den-
expensive to run hard- and soft-copy systems concurrently. A sity versus radiation exposure response curve than the classic
period of <3 months is a reasonable timescale for the phasing sigmoid-shaped curve of the conventional film/screen combi-
in of PACS after digital plain radiography has been introduced. nation (Fig. 1.2). This technology thus gives a greater latitude
The digital archive acquired during this time is sufficient for of exposure, which is useful under challenging radiographic
the majority of comparisons with previous plain images4, and circumstances, such as in the intensive care unit, and allows
obviates the need for a great deal of retrospective digitization structures of widely differing radiographic density to be well
of conventional film. visualized on the same image, such as the area behind the car-
diac silhouette and the lung vessels, or the soft tissues of the
Digitization of analogue film neck and the cervico-thoracic junction. This feature allows a
This is a time-consuming process that requires relatively skilled reduction, compared with film, in the number of repeat images
operators who are able to differentiate the various imaging taken due to errors in selecting the correct exposure. The spa-
modalities and body parts and orientate the films correctly tial resolution of CR is lower (approximately 2½ line-pairs per
for digitization and storage on PACS. Digitization must take mm) than that of film (approximately 5 line-pairs per mm),
place in a nondusty environment. A digitized image retains but its superior contrast resolution more than compensates for
any intrinsic limitations present in the analogue film: areas that its limited spatial resolution.
CHAPTER 1 • PICTURE ARCHIVING AND COMMUNICATION SYSTEMS (PACS) AND DIGITAL RADIOLOGY 9
Luminescence
Read out Read out electronics Read out electronics
HIS
HIS
RIS
RIS PACS
PACS
Figure 1.5 Bi-directional integration between PACS, HIS and RIS.
Scanner
is aware of the patients who have been scheduled to attend
particular outpatient clinics, and this information can be used RIS
Single data entry
directly by the PACS to prefetch the imaging studies for these
Figure 1.6 DICOM modality worklist.
patients from a long-term archive prior to their attendance at
the outpatient clinic.
There are three main reasons in favour of having bi-
directional integration between these three systems (Fig. 1.5): whereby the demographic data of a scheduled list of patients
• There is input of demographic data only once, minimizing is downloaded directly to the imaging apparatus (e.g. CT sys-
human error. tem) due to perform the studies (provided the imaging modal-
• Any update to a patient’s demographic data on HIS, RIS or ity also supports DMWL). This obviates the need to retype
PACS is propagated to all systems automatically, ensuring the demographic details of patients on the imaging equip-
that all systems contain accurate information. ment or the PACS, and thereby eliminates work and human
• Any scheduling or status data is distributed to all systems typographical errors, which would otherwise cause the images
automatically, providing all systems with advanced notice not to be recognized when later sent to a mismatched entry
of events and allowing them to prepare to play their roles for that patient on PACS. A reference number (an ‘accession
in these events. A good example of this is the concept of number’) is assigned by the RIS to each study and commu-
pre-fetching, whereby pertinent historical examinations are nicated to the imaging room as part of the DMWL to obtain
fetched to the PACS short-term storage unit in advance of a one-to-one correspondence between the DICOM data set
a patient’s attendance at an outpatient clinic, hospital admis- and the RIS-related information.
sion or imaging examination; these data are known to the
HIS and/or RIS and passed to the PACS prior to the event Hospital information system or electronic patient/
medical record
concerned.
The HIS (or in some health care institutions, the EPR) is the
Radiology information system most important computer system in a hospital. It constitutes
The RIS may be a stand-alone computer platform, or may be a the patient master index that stores demographic (textural) data
module of the HIS12. It stores information specific to the radiol- on all patients known to the hospital, whether or not they have
ogy department, including the registration of patients attending ever had any imaging procedures performed. In addition to
only the imaging department for an imaging investigation, but this demographic data, e.g. patient name, identification number
not otherwise under the care of the hospital (e.g. general prac- and date of birth, the HIS also records admission and discharge
titioner referrals); schedules examinations, including date, time dates, outpatient appointments, clinician(s) responsible for the
and room allocation; may track the patient’s progress through patient’s care and so forth. Other modules of the HIS document
the imaging department; and stores radiological reports and laboratory data, usually including haematology, biochemistry
produces statistics concerning imaging workflow. and histopathological test results. Ultimately a true EPR should
A modern RIS will incorporate some DICOM features contain all the case notes for a patient in digital format, but most
such as DICOM modality worklist (DMWL), (Fig. 1.6). systems are a long way from achieving this goal.
‘VALUE-ADDED’ PACS
Once hospitals have PACS installed, it will gradually become Paper request forms can be integrated into the PACS envi-
clear to them that having a PACS, whereby imaging studies ronment using a work-around solution, such as digitizing the
and their reports are electronically stored and transmitted, is paper request forms prior to acquisition of the imaging study, or
only the first step in achieving a fully integrated, seamless, using a bar code hand-held laser reader on the patient label on
electronic system of health care workflow. Becoming film- the request form at the time of imaging and reporting to locate
less is not enough: all health care systems are a long way off the electronic entry on RIS, PACS and the speech recognition
the aim of also becoming paperless. In order fully to achieve system. Such processes have a number of drawbacks, do not
the latter, there can be no paper request forms, and no paper produce a streamlined workflow and, most importantly, they
copies of reports. do not eliminate the existence of a paper request. It is only by
CHAPTER 1 • PICTURE ARCHIVING AND COMMUNICATION SYSTEMS (PACS) AND DIGITAL RADIOLOGY 11
ceasing to be dependent upon the physical existence of a paper and signed by the radiologist it is instantaneously available to
request that health care institutions can function as virtual hos- any authorized user accessing the study on the PACS.
pitals, and become location independent with respect to the The ‘imaging loop’ should be closed by installing elec-
sites of generation of the imaging study request, the acquisition tronic feedback of results, whereby the relevant clinician(s)
of its imaging study, and the generation of its report17. are electronically shown the reports on imaging studies they
To eliminate paper requests, imaging studies must be have requested on their patients. Returned results therefore
booked using a remote electronic requesting system (some- have to be grouped by referring clinician, or preferably, by
times known as ‘order communications [order comms]’). The referring clinical firm (team). There needs to be a means
image study request details must be entered electronically at whereby the reporting radiologist can, at the time of dictat-
the time of booking (either by directly typing them into the ing the report, electronically add an ‘alert’ flag to a report
system or by using speech recognition to enter them). The revealing an unexpected or urgent clinical finding to draw it
booking of imaging studies is rather more demanding of such urgently to the attention of the referring clinician and his/her
electronic requesting systems than the booking of other inves- team. The imaging loop is closed by the clinician (or his/her
tigative studies, such as haematology or biochemistry profiles team) responsible for the health care of the patient reading
or histopathology examinations. Radiologists in the UK are the report on the imaging study and taking responsibility for
legally bound by IRMER (ionizing regulations in medicine it. There needs to be a robust audit trail to record who read
regulations), whereby studies using ionizing radiation must the report, and, more importantly, who took responsibility for
be clinically justified and therefore the clinicians’ requests for understanding and acting upon it to avoid imaging reports
such imaging studies must be vetted by a radiologist to ensure being overlooked, and also for medico-legal purposes should
they are indeed clinically indicated. An electronic requesting an important imaging result be missed. It is not adequate to
system for radiological studies must allow this vetting step to rely upon communication of urgent results by telephone in
take place electronically, with an audit trail of the person who case (as often happens nowadays) the relevant clinician is no
changed an imaging request and the reason for the change; the longer on duty or is unable to be contacted for some reason.
details of the new procedure scheduled must be made known Paper reports sent by conventional mail or fax can also be
to the referrer and his/her team, and to the patient. unreliable and there is no confirmation that they have reached
The generation of the radiology reports should ideally be their destination and been read by the relevant clinician. Thus
via an electronic speech recognition system fully integrated the electronic communication of results system must incor-
with both the RIS and the PACS. In this way when an imag- porate a two-step receipt procedure whereby the reader must
ing study is displayed on the PACS monitors for reporting, the actively click twice (e.g. in two different check boxes), once
integration means that both the RIS and the speech recogni- to acknowledge reading and again to indicate understand-
tion system will automatically ‘know’ which imaging study is ing of (i.e. taking responsibility for) the result of the imag-
being reported. The speech recognition system immediately ing procedure. These steps must be traceable by the system’s
converts the dictated report into text, which is corrected there audit trail. If this form of verification is not followed then an
and then by the radiologist author, and is electronically signed. electronic result may appear to have been read (for example,
This report is permanently stored in the RIS (the location as with an email programme, by changing from bold to nor-
where radiological reports are conventionally stored long mal type font) whereas in fact it has merely been opened, or
term) and is visible on the PACS and in the HIS/EPR. Speech opened by a medical student or junior doctor who will not
recognition adds considerable value to the PACS since imag- take responsibility for dealing with the consequences of the
ing workflow efficiency is markedly improved, by eliminating report findings. Unfortunately these electronic requirements
any delays between dictation and typing, and between typing for closing the imaging loop have yet to be implemented on
and verification by the radiologist. Once a report is dictated current commercial system software.
It is the quality and speed (bandwidth) of the enterprise- Historically, the costs of image data storage were the main
wide network that determines the rapidity with which determinant of the cost of a PACS, but nowadays, they should
images are displayed on the PACS workstation and web only be a minor component, as disk costs have fallen so dra-
browser computer monitors. The quality of the PACS hard- matically, and PACS software costs are much higher than its
ware and software is immaterial if image delivery is rate hardware costs (although vendor mark-up for hardware can
limited by an inadequate portion of the network. Radio- skew this ratio).
logical images contain very large amounts of data: a single
plain chest radiograph contains from 8–10 megabytes of Storage volumes and types
data, and there are considerably more data in the very large In the early days of PACS, storage volumes were mainly deter-
(around 1000–2000 individual images) thin-section CT and mined by the size and number of CR images, but multi-slice
MR cross-sectional image datasets often now acquired for CT and large MR studies are now overtaking CR, and required
a single study. data volumes are still increasing, such that even after applying
Networks nowadays run standard ethernet protocol, with lossless compression, an average district general hospital will
TCP/IP (Transmission Control Protocol/Internet Proto- need about 2 Tbyte and a major teaching hospital about 5
col) of the Internet. At least 100 megabit ethernet is needed Tbyte per year18.
to transmit radiological images and some institutions are There have been many types of storage used over the years
now running 1 gigabit ethernet throughout the health care for PACS to handle such large volumes, including optical disks
enterprise. The core of many PACSs still run propriety net- (magneto-optical disks, MOD or compact disk/digital versatile
work protocols to achieve greater speed of transmission. disk, CD/DVD), tape-drives and jukeboxes containing both of
Modern hospitals have a switched network rather than a these. All these mechanical devices, with their attendant reli-
hub network because a switched network allows the whole ability issues, are now obsolete, and the only sensible method
bandwidth (e.g. 100 Mbps ethernet) to be available to each of storing images for normal use is Redundant Array of Inex-
switch in the network (whereas in a hub network the entire pensive Disks (RAID) – a system that spreads data across many
hub accommodates the total bandwidth (e.g. the 100 Mbps) disks, providing the following advantages:
and all the attached items have to share a portion of that • The data are still accessible if one (or, in some configura-
bandwidth). tions, more than one) drive fails
• The speed of the system as a whole is faster than the
individual disks, because they work in parallel.
STORAGE REQUIREMENTS AND SOLUTIONS
FOR PACS RAID itself can be used in multiple different configurations
and ways of connecting the PACS server, with widely different
The PACS requires two basic different types of storage for: pricing, but the main categories are:
• Image data 1 Directly attached to the server. This is cheap and simple, but
• Indexing data. expansion then requires the whole system to be stopped
and restarted, and expansion is also limited.
PACS image data storage 2 Attached to a network using file servers (network attached
The PACS requires a very large amount of storage for the storage [NAS]).
images (which includes associated objects such as presenta-
tion states), but the requirements are fundamentally different This is mature standard technology that offers an excellent
from the types of storage used by other large data users such as choice of products as fully supported ‘off-the-shelf ’ hardware,
banks. The peculiar features of PACS image storage are: to which any competent server should be able to connect.
• There is an ever-increasing need for more storage space
• The data once written are rarely (if ever) changed Storage area networks
• Images tend only to be accessed a few times during their SANs are basically RAID, but connected to their servers
entire life and normally by only one user at a time using special connections (normally optical fibre). They
• Required disk access speed is not actually very high, as end- provide extensive facilities for management, backup, flex-
user access is more limited by network speed than by disk ible use of space and sharing of disk space across multi-
limitations. ple servers, but this flexibility comes at a price, as they are
much more expensive than equivalent space bought in the
PACS indexing data storage form of file servers, normally being 5–10 times as expensive
The PACS has more ‘conventional’ requirements for its index- per Gbyte.
ing data storage where:
• The total volume of data is tiny by modern standards (a few Proprietary storage systems
gigabytes at most) It is possible to replace the above ‘open’ and standard connec-
• Data are constantly being updated and changed tion methods by proprietary equivalents, which are claimed
• Concurrent access by multiple users (for queries) is to have unique special features, but costs are high and there is
required. then ‘lock-in’ to the particular vendor concerned.
Random documents with unrelated
content Scribd suggests to you:
Geology - Research Paper
Spring 2024 - Laboratory
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookfinal.com