computers in gastroenterology

F. R. Vicary (Ed.)

Computers in
Gastroenterology
With 37 Figures

Springer-Verlag Berlin Heidelberg GmbH


ER. Vicary, FRCP
Consultant Physician, Whittington Hospital, Highgate HilI,
London N19, UK

ISBN 978-1-4471-3261-5
DOI 10.1007/978-1-4471-3259-2

ISBN 978-1-4471-3259-2 (eBook)

British Library Cataloguing in Publication Data
Computers in Gastroenterology.
1. Gastroenterology. Applications of Computer systems
I. Vicary, FR. (Frederick Robin), 1946616.3'3'00285
ISBN 978-1-4471-3261-5
This work is subject to copyright. All rights are reserved, whether the whole or part of the
material is concemed, specifically the rights oftranslation, reprinting, re-use of illustrations,
recitation, broadcasting, reproduction on mü;rofilms or in other ways, and storage in data
banks. Duplication of this publication or parts thereof is only permitted under the
provisions ofthe German Copyright Law of September 9,1965, in its version of June 24,
1985, and a copyright fee must always be paid. Violations fall under the prosecution act of

the German Copyright Law.

© Springer-Verlag Berlin Heidelberg 1988
Originally published by Springer-Verlag Berlin Heidelberg New York in 1988
Softcover reprint of the hardcover 1st edition 1988
The use of registered names, trademarks etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the relevant laws and
regulations and therefore free for general use.
Product Liability: The publisher can give no guarantee for information about drug dosage
and application thereof contained in this book. In every individual case the respective user
must check its accuracy by consulting other pharmaceuticalliterature.
Filmset by Tradeset, Welwyn Garden City, Herts AL7IBH
Printed by Henry Ling, The Dorset Press, Dorchester
2128/3916543210


Preface

All over the United Kingdom gastroenterologists have been excited in
the last few years by the microcomputer revolution. All over the United
Kingdom individuals have been beavering away, producing systems for
use in their units. Practising gastroenterologists have wanted endoscopy
systems, surgeons wanted the ability to audit, academics wanted to teach,
and everyone wanted expert diagnostic systems.
In a few centres, mathematicians saw the potential for applying
mathematical techniques to clinical problems. Hardware and software
experts viewed the hospitals and individual units with eagle eyes, desperate to provide complete hospital systems and local networks.
It was clear that all of these people needed to be got together, to discuss
the present and future. This book is the result of this meeting of minds.
The meeting eventually occurred (after one or two hiccoughs!) at the

Whittington Hospital on 12 and 13 May 1987.
This book contains almost all of the papers given at the meeting. The
meeting was divided into five sessions entitled "Existing systems", "Endoscopy systems", "Audit", "Teaching systems", and "Diagnosis and
therapy".
Four of the sessions started with overviews of the areas under discussion, given by invitation by senior figures in the field, to present a
background introduction for those unfamiliar with the field. Each session
concluded with a free-format discussion.
In between sessions, there were "nutrition" breaks, to allow delegates
to visit the two demonstration areas. There was a trade area for commercial companies-both software and hardware retailers and pharmaceutical companies with educational materials.
In a separate room, eight home-developed systems were demonstrated
by individual delegates. Those reading this book are unable to appreciate
the excitement generated by some of these systems-mere black type and
white page being a poor substitute for the communication of the imagination shown in the development of the systems. Undoubtedly the most
excitement was around the stand where Shukri Shami demonstrated his
dynamic, interactive "expert tutorials in gastroenterology". It was the
combination of the purity of the intelligence-written in Prolog-and his
brilliant use of split-screens and colour which made his system so compulsive.


vi

Preface

Other systems enjoyed included Gladys, the grandmother of interactive systems, led along by Robin Knill-Jones. Chris Venables showed
"Ulcers", Kyran Bulger brought his colorectal database, and Peter Finch
brought Gastrofile, his real endoscopy system, based on a commercial
package but adapted and compiled by him for their unit's requirements.
North-east London was well represented; Chris William's polyp
retrieval system was seen, as was the home team's Solubile. Finally one
of the "grand old men" of computing in medicine, John Dickinson from

Barts, brought the MacMaster mainframe educational programme for
gastrointestinal bleeding.
Also in this room were the five posters. These are included in the book
as abstracts. The papers were presented in five-minute sessions each by
their authors, with some discussion on day two.
This book was produced with the help of a number of people, Chris
Venables of Newcastle provided support in the setting-up and design of
the meeting. His experience and knowledge of the area was invaluable to
me. Michael Jackson of Springer provided important technical help and
his relaxed style made him a pleasure to work with. Finally my session
colleague, Eric Beck, gave me great encouragement and provided the
Academic Centre in which to hold the meeting.
London, 1987

Robin Vicary


Contents

PART 1: Existing Systems
1 Current Trends in Microcomputers
A. P. McCann ...................... .......................................
2

3

Computer Networks: A Technology Whose Time Has Come?
D. Bryce and N. W. Carter ........................ .....................

7

3 Experience in Implementing Microcomputer Clinical
Databases in Gastroenterology
E. G. Devas ................................................................

13

4

Metabase: A Microcomputer System in Evolution in a
Gastroenterology Hospital
C. B. Williams and N. Padmanabhan ...............................

21

PART II: Endoscopy Systems

An Overview of Computerised Endoscopy Record Systems
C. W. Venables ...........................................................

29

6 A Computerised System for Recording and Reporting on
Upper Gastrointestinal Endoscopy Data
Ph. Van Hootegem, J. De Troyer, P. Rutgeerts, J. Janssens,
G. Coremans and G. Vantrappen .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

39

5

7

Practical Experience of Soluscope: A New Endoscopy
Records System
F. R. Vicary . . . . . . . . . . . . . .. .... . . . . . . .... . . . . . . . . ... ..... . . . . .... . . . . .. .. .. .

43

A Multiuser Personal Computer Endoscopy Unit Filing System
P. J. Finch and I. T. Gilmore ..........................................

51

9 ULCERS: The BSG Developed Endoscopy Record System
C. W. Venables,A.P.McCann,F. T. DeDombaland
A. P. Manning .......................... .-..................................

57

8

PART III: Audit
10 Audit of Gastrointestinal Surgery in a District General
Hospital Using dBase II
.
R. G. Wilson and J. D. Holdsworth .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .

69

viii

Contents

11 A Computerised Clinical Audit System in Practice
J. R. Coughlan, W. A. Corbett, M. J. Taylor and R. Shields

79

12 A Microcomputer for General Surgical Audit and
Administration
S. G. Pollard, P. J. Friend and D. C. Dunn . .. ........ ........ .. ...

87

13 A Non-application-specific Approach to Medical Software
Design
M.A. Walker, D. Bryce and N. W. Carter ........................

97

14 A Computerised Patient Record System Using a Fourth
Generation Language
A. A. Seifalian and K. E. Hobbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105

PART IV: Teaching Systems
15 Choles: An Expert System for the Interactive Learning of

Gallstone Disease Management
G. Palasciano, A. Circella, P. Portincasa, G. Baldassarre
and 0. Albano .............................................................

109

16 A Clinical Simulation Environment for Medical Education
C. E. Johnson, M. J. Taylor and W. A. Corbett..................

119

17 Surgical Emergencies: A New Learning Experience
P.R. Edwards, W. A. Corbett, M. J. Taylor
and J. R. Coughlan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129

18 Computer-aided Instruction
S. K. Sharni and S. Knight ..............................................

139

19 A Computerised Personal Bibliographic System
K. N. Bulger and T. J. McKenna .. . . . . . . ... . . . . . .......... .. .. .......

145

PART V: Diagnosis and Therapy
20 Computers in Gastroenterology: An Overview of Diagnostic
Applications

R. P. Knill-Jones ..........................................................

149

21 Can A Computer Help in the Management of Small Bowel
Obstruction?
J. A. Pain, D. St J. Collier and R. Hanka ... . . .. .. . . . . . . .. .........

161

22 A Colorectal Cancer Database System
K. Bulger, N. Afdhal and D. O'Donoghue ........................

167

23 Microcomputer-aided Diagnosis of Jaundice (Solubile)
N. Newman, E. W. F. W. Alton and F. R. Vicary ...............

175


Contents

ix

24 Computer-aided Analysis of Dietary Nutrient and Fibre
Intakes in Gastroenterology
A. S. Mcintyre, M. Ibbotson, J. Duthie, S. P. Burnham,
J. O'Brien, S. Day and W. R. Burnham ............................

185

25 Computing and the Histopathology of Intestinal Inflammation
D. Jenkins . .. . . . . . . .......... .. .. . . . . . . .. .. ..... ............. .. .. . . . . . . . . .. .

193

26 Microcomputer Data Acquisition in a Developing Country
with Transfer to a Mainframe Computer for Analysis
T. H. C. Williams, G. F. Sargent, I. R. John, S. P. Ashmore
and M. S. Tanner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

205

27 Gastroenterology, Computing and the Artificially Intelligent
Pathologist
D. Jenkins . .. . . . . . . . . .. ... . . . . . .. ........ .. .. .. .... ............ .. ...........

213

28 Experience with Computer Interviewing of Patients
R. P. Knill-Janes, G. P. Crean, R. J. Holden
and G. Lindberg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

215

29 Pharmaceutical Dosing ofTobramycin
M. Vissing, P. Billesb(lllle, E. Iversen, E. Bruun
and L. K. Hansen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

217

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219


Contributors

N. Afdhal
Gastroenterology and Liver Unit, St. Vincent's Hospital, University
College, Dublin, Ireland

0. Albano
Institute of Clinical Medicine I, 70124 Policlinico, Bari, Italy
E.W.F.W. Alton
Department of Gastroenterology, Whittington Hospital, Highgate Hill,
LondonN19, UK
S.P. Ashmore
Department of Child Health, University of Leicester, Leicester, UK

and

Department of Paediatrics, King Edward Memorial Hospital, Pune, India

G. Baldassarre
Institute of Clinical Medicine I, 70124 Policlinico, Bari, Italy
P. Billesb0lle
Department of Surgical Gastroenterology, Gentofte Hospital,
University of Copenhagen, Denmark
D. Bryce
Department of Medical Computing, Ninewells Hospital and Medical

School, Dundee DDl 9SY, UK
E. Bruun
Department of Surgical Gastroenterology, Gentofte Hospital,
University of Copenhagen, Denmark
K.N. Bulger
763 VFW Parkway, West Roxbury, Boston, Mass. 02132, USA
S.P. Burnham
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE,UK

and

The London Hospital, Whitechapel, London EllBB, UK


xii

Contributors

W.R. Burnham
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE, UK
and
The London Hospital, Whitechapel, London E11BB, UK
N.W. Carter
Medical Computing Unit, Department of Medical Computing,
Ninewells Hospital and Medical School, Dundee DD1 9SY, UK
A. Circella
CSATA Tecnopolis-Valenzano, 70100 Bari, Italy
W.A. Corbett

Department of Surgery, Middlesbrough General Hospital, Ayresome
Green Lane, Middlesbrough, Cleveland TS5 5AZ, UK
G. Coremans
Department of Medicine, Division of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium
J.R. Coughlan
Department of Surgery and Computer Science, University of Liverpool,
PO Box 147, Liverpool L69 3BX, UK
G.P. Crean
Diagnostic Methodology Research Unit, Southern General Hospital,
Glasgow, UK
S.Day
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE, UK
and
The London Hospital, Whitechapel, London E11BB, UK
F.T. DeDombal
Department of Surgery, St. James Hospital, Leeds, UK
J. De Troyer
Department of Medicine, Division of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium
E.G.Devas
Metasa Ltd., Admel House, 24 High Street, Wimbledon, London
SW195DX, UK
D.C. Dunn
NHS General Surgical Unit, Addenbrooke's Hospital, Hills Road,
Cambridge CB2 2QQ, UK


Contributors

xiii

J. Duthie
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE, UK

and

The London Hospital, Whitechapel, London EllBB, UK
P.R. Edwards
Department of Surgery and Computer Science, University of Liverpool,
PO Box 147, Liverpool L69 3BX, UK
P.J. Finch
Gastroenterology Unit, Royal Liverpool Hospital, Liverpool L7 8XP, UK
B.J. Friend
NHS General Surgical Unit, Addenbrooke's Hospital, Hill Road,
Cambridge CB2 200, UK
L.T. Gilmore
Gastroenterology Unit, Royal Liverpool Hospital, Liverpool L7 8XP, UK
R. Hanka
Department of Surgery, The Ipswich Hospital, Ipswich, UK

and

Department of Community Medicine, University of Cambridge,
Cambridge, UK

L.K. Hansen
Department of Surgical Gastroenterology, Gentofte Hospital,

University of Copenhagen, Denmark
K.E. Hobbs
Academic Department of Surgery, The Royal Free Hospital School of
Medicine, London, UK
R.J. Holden
Monklands District General Hospital, Airdrie, Lanarkshire, UK
J.D. Holdsworth
Ashington Hospital, West View, Ashington, Northumberland, UK
J. Hooper
Department of Gastroenterology, Whittington Hospital, Highgate Hill,
London N19, UK
M. Ibbotson
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE,UK

and

The London Hospital, Whitechapel, London EllBB, UK


xiv

Contributors

E. Iversen
Department of Surgical Gastroenterology, Gentofte Hospital,
University of Copenhagen, Denmark
J. Janssens
Department of Medicine, Division of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium

D. Jenkins
Department of Histopathology, Whittington Hospital, Highgate Hill,
LondonN19, UK
I.R. John
Department of Child Health, University of Leicester, Leicester, UK
and
Department of Paediatrics, King Edward Memorial Hospital, Pune, India
C.E. Johnson
Knowledge Based Systems Groups, Computer Science Department,
University of Liverpool, PO Box 147, Liverpool L69 3BX, UK

s. Knight
Wexham Park Hospital, Slough, Bucks, UK
R.P. Knill-Janes
Diagnostic Methodology Research Unit, Southern General Hospital,
Glasgow, UK
G. Lindberg
Karolinska Institut, Huddinge Sjukhus, S-14186 Huddinge, Sweden
A.P. Manning
Department of Gastroenterology, Royal Infirmary, Bradford, UK
A.P. McCann
Department of Computer Studies, University of Leeds, Leeds, UK
A.S. Mcintyre
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE, UK
and
The London Hospital, Whitechapel, London E11BB, UK
T.J. McKenna
Department of Gastroenterology, Mater Misericordiae Hospital,
Dublin, Ireland

N. Newman
Department of Gastroenterology, Whittington Hospital, Highgate Hill,
London N19, UK


Contributors

XV

J. O'Brien
Department of Gastroenterology, Oldchurch Hospital, Romford, Essex
RM70BE, UK
and
The London Hospital, Whitechapel, London E11BB, UK

D. O'Donoghue
Gastroenterology and Liver Unit, St. Vincent's Hospital, University
College, Dublin, Ireland
N. Padmanabhan
St Mark's Hospital for Diseases of the Rectum and Colon, City Road,
LondonEC1V2PS, UK
P.A. Pain
Department of Surgery, The Ipswich Hospital, Ipswich, UK
and
Departments of Community Medicine, University of Cambridge,
Cambridge, UK
G. Palasciano
Institute of Clinical Medicine I, 70124 Policlinico, Bari, Italy
S.G. Pollard
NHS Surgical Unit, Addenbrooke's Hospital, Hills Road, Cambridge

CB22QQ, UK
P. Portincasa
Institute of Clinical Medicine I, 70124 Policlinico, Bari, Italy
P. Rutgeerts
Department of Medicine and Divisions of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium
G.F. Sargent
Department of Child Health, University of Leicester, Leicester,UK
and
Department of Paediatrics, King Edward Memorial Hospital, Pune,
India
A.A. Seifalian
Academic Department of Surgery, The Royal Free Hospital School of
Medicine, London, UK
S.K. Shami
Wexham Park Hospital, Slough, Bucks, UK
R. Shields
Department of Surgery and Computer Science, University of Liverpool,
PO Box 147, Liverpool L69 3BX, UK


xvi

Contributors

B. St J. Collier
Department of Surgery, The Ipswich Hospital, Ipswich, UK
and
Departments of Community Medicine, University of Cambridge,
Cambridge, UK

M.S. Tanner
Department of Child Health, University of Leicester, UK
and
Department of Paediatrics, King Edward Memorial Hospital, Pune,
India
M.J. Taylor
Department of Surgery and Computer Science, University of Liverpool,
PO Box 147, Liverpool L69 3BX, UK
Ph. van Hootegem
Department of Medicine, Division of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium
G. Vantrappen
Department of Medicine, Division of Gastroenterology, University
Hospital Gasthuisberg, Leuven, Belgium
C. W. Venables
Department of Surgery, Freeman Hospital, Newcastle upon Tyne, UK
F.R. Vicary
Department of Gastroenterology, Whittington Hospital, Highgate Hill,
London N19 5NF, UK
M. Vissing
Department of Surgical Gastroenterology,
Gentofte Hospital, University of Copenhagen, Denmark
M.A. Walker
Departments of Surgery and Medical Computing, Ninewells Hospital
and Medical School, Dundee DDl 9SY, UK
C.B. Williams
St Marks Hospital, Diseases of the Rectum and Colon, City Road,
LondonEC1V2PS, UK
T.H.C. Williams
Neville Hall Hospital, Abergavenny, Gwent, UK

R.G. Wilson
The Royal Victoria Infirmary, Newcastle upon Tyne, UK


PART I

Existing Systems


1 Current Trends in Microcomputers
A.P.McCann

Early Days
My first experience of computing was with the Ferranti Pegasus in 1959. This was a
machine with a few hundred bytes of memory held in delay lines and a cathode ray
storage tube, a magnetic drum but no disk and originally not even a magnetic tape
drive. All input and output was by means of 5 hole paper tape. Nevertheless it was a
great advance on log tables or even electric calculators and many complex calculations were done through the long hours of the night.
We had not known that Pegasus was a mainframe but after the emergence of large
machines manufactured by Univac, IBM, ICf, English Electric and many others in
the 1960s, the emergence of DEC minicomputers in the late 1960s caused the coining
of this phrase.

The Microprocessor
In 1970, I recollect the astonishment felt by me and my colleagues as we read of the
announcement by Intel of the first complete microprocessor on a single chip. Admittedly it was a processor only, with a few 4 bit registers, and to make any use of it, it
was essential to design and mount it on a board with such things as a clock, memory
and input/output ports, but it still seemed a technical marvel.
Developments came headlong: 8 bit micros were produced by a variety of manufacturers and soon it was possible to buy complete boards with processors already
mounted and connected to a kilobyte or two of memory, and serial or parallel ports.

Perhaps even more important, rudimentary operating systems appeared in the form


4

Computers in Gastroenterology

of assemblers and dynamic debuggers, DDT being a famous example of the latter.
They were capable of breakpointing, single-stepping instructions and dumping register and memory contents in understandable form onto monitors.

The Microcomputer
The cost of computer memory has been halving approximately every 18 months ever
since 1960, and by the mid-1970s this had a truly dramatic effect on the micro. Faster
CPUs ha\'ing better register structures had been created by then and it was feasible
to equip them with sufficient memory for them to show their paces. Additionally, the
tens of kilobytes built in to these machines were sufficient to develop true operating
systems so that CP/M, UCSD and DOS were produced and marketed by enthusiastic
pioneers. Fortunately disk technology had advanced also, so that cheap means of
storing and distributing both the operating systems and user programs were available
using floppy disks. The net result was the disappearance to a large extent of the single
board microprocessor and its replacement by the boxed microcomputer.
With increasingly sophisticated design tools, the chip manufacturers have run
through two new generations of processor chips. Both 16 bit and 32 bit microprocessors are readily available and at low cost. One problem with the old 8 bit systems was
difficulty in addressing more than 64 kbytes of memory. This is now largely a past
problem, though not entirely, since some of the 16 bit designs in particular carried
forward features from earlier processors, sometimes including addressing restrictions.

The Personal Computer (PC)
The market is currently dominated by the IBM PC which was launched in the early
1980s and has attracted an enormous amount of competition from the so-called

clones. Its operating system is based on MSDOS, a Microsoft operating system first
developed for the 8 bit systems of a decade ago. Manufacturers such as Compaq have
been able to offer machines with greater performance than the PC itself by using
more powerful chips running with a faster clock, and moreover have done this at
lower cost. The various models of the PC and its competitors differ principally in the
amount of main memory provided, the types of disk drives (including Winchester or
"hard" disks offering 10, 20 or more Mbytes) and the provision of special-purpose
boards offering graphics, colour and other features.
Because of the PC predominance, other technically good machines have been
squeezed, so that for example Apricot have dropped almost all but their clone, and
the Apple Macintosh appears to have survived mainly by developing a niche market
in low cost electronic publishing. Right at the bottom of the market, the very cheap
machines such as the Acorn BBC models, Commodore and Sinclair Spectrum have
found it difficult to remain profitable because of falling costs of more powerful
machines. This has been accentuated in this country by the success of Amstrad word


Current Trends in Microcomputers

5

processing and PC type machines costing around £500 to £800 including monitor and
disk drives.

The Future
It is foolhardy to predict too far ahead in the computing business. Nevertheless some
things seem probable. It is said that no manager ever gets sacked for buying IBM, so
the very recent announcement of its PS/2 range will have a major effect over the next
couple of years. The use of a 32 bit Intel chip in this range will finally free IBM's
personal computer from its memory problems and it is to have a new operating system, OS2, which will offer many enhancements. It is interesting to note, however,

that the problems of software development so greatly exceed those of hardware, that
it is anticipated three or four models will be marketed this year before OS2 is ready
for launch in 1988. No manufacturer these days can afford to ignore the enormous
investment of his company, software houses and customers in existing software, so
naturally the new machines will be able to occupy themselves with MSDOS while
awaiting OS2.
The trend to cheaper memory cannot continue for ever, but it has not yet run out
of steam. Chips each containing 256 kbits of memory are now routinely available; 1
Mbit memories are being used in mainframes, and 4 Mbit chips and beyond are at
sampling and development stages. This means that more computing power will be
packed into microcomputers being developed now.
The same is true for disk storage systems. The floppy disk started at a cumbersome
8 inches holding around 100 kbytes. It was largely superseded by 5 inch disks holding
up to 400 kbytes, and these now look likely to be replaced by 3 inch disks with
capacities of 800 kbytes and more. The hard disk has also shrunk physically and
grown in capacity, so it will increasingly be possible to include it in micros of modest
cost. Capacities of 20-80 Mbytes are already common. One problem with such large
disks is that of data integrity Gust imagine backing up 40 Mbytes onto fifty 800 kbyte
floppy disks). Tape streamers look set to provide a cost-effective solution to this difficulty. It seems improbable that the floppy will be superseded for a long time despite
its slow access time, since it has excellent characteristics for exchange of information,
and a few hundred kilobytes of information seem to match naturally onto many
requirements.
High quality printing is in demand and laser printers can now be attached to microcomputers and supplied with the necessary driving software for about £3000. It would
seem likely that this price might halve in real terms over the next 3 or 4 years.

Conclusions
While there will certainly be more, cheaper and better of the items mentioned above
together with sundry others scarely mentioned such as colour, graphics, mice, networks, windows, menus, spreadsheets and vast amounts of applications software,

6

Computers in Gastroenterology

there are likely to be some surprises as well over the next decade. It is a baffling
business trying to keep abreast of developments, and of course it is never a good time
to buy anything, since it will inevitably be obsolescent almost as soon as installed.
After all, even that old friend the Beeb looks like being ousted by something called a
RISC machine, provided the market is prepared to take a chance on it.


2 Computer Networks: A Technology Whose Time

Has Come?

D. Bryce and N.W. Carter

Introduction
Although mainframe computer networks have been in use for many years, the microcomputer network is a fairly recent phenomenon. The introduction of the microcomputer in the mid-1970s saw a move away from the large centralised mainframe and
departmental microcomputers towards the stand-alone microcomputer which gave
the user a truly "personal" computer. As these machines grew in power, and consequently the range of applications that they could handle, the isolation of these systems became a disadvantage.
In the early 1980s, several proprietary products were available which would allow
many, typically about a hundred, microcomputers to communicate with each other
and to share valuable resources. Since these products operated over restricted distances, usually of the order of a few kilometres, a number of microcomputers connected in this way is often called a local area network or LAN. These LANs have
many advantages over a collection of isolated microcomputers. For instance on the
computing side they introduce the concept of distributed processing, where several
processors on the network can contribute to the successful completion of a single
task, such as updating a large database. With shared access to fast high capacity disk
storage, printers, plotters and modems, LANs avoid costly duplication of peripherals
and enhance the performance of existing equipment.

8

Computers in Gastroenterology

Evaluation of LAN
In 1982 the Faculty of Medicine Computing Unit at Ninewells Hospital, Dundee,
decided to investigate this new technology as a possible replacement for an ageing
minicomputer. Our aim was to provide a high speed, relatively low cost computing
resource which could be readily expanded as needed to meet future demands. To this
end we evaluated the LAN products that were available at that time against the
following criteria:
1. Support for all of the popular microcomputers and their operating systems available at that time; in other words, we did not want to be locked into a single computer manufacturer for all of our microcomputer resources.
2. As the potential work areas were scattered throughout the hospital and medical
school, the LAN chosen should be capable of operation over distances of up to
2430m.
3. To maintain a reasonable response time, that is, no more than a second or so,
during periods of fairly heavy use, the LAN operating bandwidth should be at
least 1 Mbit/s-1.
4. Network devices such as magnetic tape back-up units and print servers should be
available.
5. Network interface cards for the microcomputers should be "reasonably" priced.
This was difficult to quantify but it was felt that a network interface card that cost
over £1000 for a microcomputer which only cost £500 was definitely "unreasonable".
6. Support for the UCSD p-System must be available. We had chosen this operating
system as our in-house software development system, as any software developed
under this system would run, without modification, on any computer for which
the p-System was available. All popular microcomputers support this operating
system.

After careful scrutiny of the product literature and several visits to sites running
LANs we chose the Corvus Systems Omninet running Corvus' own Constellation
network software. This is a well-proven product with a large user base worldwide. At
the time Corvus claimed to have the highest number of installed networks in the
world, and had a good reputation for quality and reliability.

Installation
In August 1982 we installed our first LAN equipment consisting of:
2 Apple He with floppy disk drives, Apple Pascal (an early version of UCSD pSystem) and Apple DOS
2 Network interface cards
1 Corvus disk server


Computer Networks: A Technology Whose Time Has Come?

9

1 20 Mbyte Corvus hard disk and
1 Mirror VCR tape backup unit
100m of twisted pair network cable.
Initial setting up was trouble free. The equipment was simply unpacked, connected together and after the various installation procedures carried out, everything
performed flawlessly. A few months later the network cabling was extended to
include three departments who wanted to connect their Apple II microcomputers to
the network and take advantage ofthe rapid access hard disk storage. In 1984 another
two 20 Mbyte hard disk drives were added bringing the total network hard disk storage up to 60 Mbyte and the connection of the first IBM PC running both MS-DOS
and the UCSD p-System. This expansion has continued steadily; the current network
has 105 Mbyte of hard disk storage, over 40 workstations, mainly Olivetti M24s and
IBM PCs, two laser printers, a plotter and a magnetic tape unit for daily backups of
the hard disks. They are all connected by over 2130 m of network cabling. The main
services provided to the 68 users are database facilities using our in-house developed

software for both clinical research and departmentaVunit audit, word processing, and
several research projects for which we have written bespoke software.

Limitations
Overall, the system has proved to be very reliable and is generally liked by the users.
The network hardware is kept running permanently, backups are automatically done
on a daily basis, usually at midnight, to keep disruption of the service to a minimum,
and take about 3 h to run to completion. In the time we have been running the network there have only been three occasions when a component has failed and needed
repair. Each time we have had sufficient redundancy in the system to enable us to
reconfigure the network, keeping disruption to a minimum while the offending item
has been repaired. In almost all respects the LAN has lived up to our expectations.
However, there are some limitations, especially when compared with the more
recently developed networking products that are now available from other suppliers.
Corvus have been slow to develop the Constellation network software which,
although adequate, lacks the sophistication and facilities found in other network
software; for example, the network maintenance software is rudimentary by today's
standards and there is no electronic mail facility. On the hardware side Corvus have
been slow to introduce new network hardware products which would keep them at
the forefront of LAN technology; for example, we have requirement for an X-25
gateway, which would allow any workstation on the network to communicate and
transfer data with the university mainframe situated some 6~ km away, or alternatively to the area health authorities mainframe situated 9~ km away. Until this
long-awaited product is available, workstations have to be connected to these services on an individual basis-a costly exercise.
The original network disk drive is now long past its expected life and is currently
due for replacement. This has given us an opportunity to re-evaluate our networking
strategy against current LAN technology to meet the anticipated demands of the next
5 years. As the number of microcomputers, minicomputers and mainframe installations increase within the health service, the ability of these systems to intercommuni-


10

Computers in Gastroenterology

cate effectively becomes a major requirement when planning a fully integrated data
network. The introduction of patient administration systems (PAS) in the near future
and the increasing demands for a more cost-effective health service seem likely to
accelerate this expansion. With these thoughts in mind, plus the experience of the
past 4 years, we once again surveyed the LAN marketplace.

New Developments
There have been two major developments in the networking field since our last
review in spring of 1982. The first of these has been the gradual emergence of international standards, with which most of the major manufacturers have agreed to comply. The most significant of these is the seven-layer OSI model (which has been
accepted by the International Standards Organisation as a reference model). This
does not attempt to define what the components of a network should be. It simply
sets out the way the various parts of the network should fit together. It divides a network into seven layers. The topmost layer, layer 7, is the layer the user sees-the
application layer-which is connected in tum to the presentation layer, the session
layer, the transport layer, the network layer, the data link layer and finally the physical layer. The concept of this model has been accepted for almost a decade, but so far
only the lower levels have been implemented. It may be some considerable time
before all of the layers have been finally incorporated; however it is a step in the right
direction. More significant, however, was the entry of IBM into the LAN field. In
1985, Microsoft and IBM announced DOS 3.1, and shortly afterwards IBM
announced the Netbios program; both seem set to become the de facto standard for
many years to come. The release of these products prompted many third-party
software developers to produce networking software which ran on existing networking hardware, supporting the new software.

Novell's Netware
Our goal then was to find a product which would support this emerging standard, run
all of our applications software on our existing network hardware and provide the
features and expansion capability that the Corvus Constellation software lacked.
After careful consideration, we decided that Novell's Netware seemed the best
choice. We managed to obtain, on loan, a working Novell Netware system and were

able to confirm, much to our satisfaction, that we did indeed have a simple upgrade
path that would not render our past efforts obsolete and that would also allow us to
meet any reasonable demand for expansion over the next few years. The additional
features of Netware when compared to Constellation software are many, but the
main areas of improvement are better protection against unauthorised access of sensitive information, many more useful utilities like electronic mail incorporated as
standard, the ability to use a workstation itself as a gateway for connecting to other
LANs, mainframes etc., and more significantly the ability to build networks with a
much higher degree of fault detection and error correction than had previously been
possible.


Computer Networks: A Technology Whose Time Has Come?

11

Conclusions
Looking further into the future, networks will play a fundamental role in an integrated computing service. The technology already exists to allow the development of
a wide area network which could link several outlying hospital computer systems,
allowing full interchange of data and services between the various centres. Within a
large hospital, networks can be used to carry telephone circuits and video channels in
addition to the data transmission functions. We may yet see health education videos
being shown in outpatient waiting areas. However, these projects require funding,
and given the current cash crisis within the National Health Service (NHS) may not
be realised for some time.
In the past microcomputer networks have often been thought of as a solution waiting for a problem. The gradual emergence of viable standards and the entry of IBM
into the networking arena has resulted in an explosion of interest in LANs, particularly in the business sector. The NHS has often been criticised for its head-in-the-sand
attitude to innovation; so we hope that on this occasion it will take heed of these new
network products and have the foresight to recognise that LANs do have an important role to play and should be given every consideration when installing new, or
adding to existing, computer installations.