Physica Medica (2008) 24, 42e48

ava i lab le a t www.sc iencedi rec t .com

journa l homepage : h t tp : / / in t l . e l sev ierhea l th .com/ j ourna l s /e jmp

INVITED PAPER

Combatting cancer in the third millennium eThe contribution of medical physics*

Steve Webb*

Joint Department of Physics, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Downs Road,Sutton, Surrey SM2 5PT, UK

Received 15 October 2007; accepted 4 January 2008Available online 4 March 2008

KEYWORDSFuture prophesy;Radiotherapy physics;Cancer goals;Cancer treatment

* Invited Opening Plenary Lecture, 1* Tel.: þ44 208 661 3350; fax: þ44 2

E-mail address: steve.webb@icr.ac

1120-1797/$ - see front matter ª 200doi:10.1016/j.ejmp.2008.01.002

Abstract In this invited opening Plenary Lecture at the 1st European Conference on MedicalPhysics, I indicated some of the roles of medical physics, and specifically medical radiationtherapy physics, in the development of improved cancer care for the third millennium. It issaid that you are only asked to predict the future if you are seriously old and/or will not bearound long enough to know if you were correct. Hopefully, however, I will be able to contrib-ute to this field for many years to come and many of my suggestions will be enacted. ‘‘Crystalball gazing’’ is, however, a very unscientific process. Scientists are trained to study and analysesituations, report their findings and stop at that. ‘‘Future gazing’’ is not predicting the short-term developments; it is about being bold, radical and stating what today is impossible or al-most unthinkable. So-called ‘‘scientific prophets’’ can be entertaining (which I also hope I wasin this lecture) but at worst look egocentric and possibly ridiculous. I seem to have survivedthree previous requests to make scientific predictions [Webb S. The future of photon exter-nal-beam radiotherapy: the dream and the reality. Physica Medica 2001;17(4):207e15; WebbS. Radiotherapy physics: the next ten years of technical development. Imaging and Oncology2005;1:43e50; Webb S, Evans PM. Innovative techniques in radiation therapy editorial, over-view and crystal ball gaze to the future. In: Webb S, Evans PM, editors. Innovative techniquesin radiation therapy. Seminars in Radiation Oncology 2006;16(4):193e8].

I proposed that important progress usually comes from two quite distinct directions. Firstly,there is ‘‘big hit science’’, that is discoveries or inventions so important that the medical worldchanges forever because of them. These are what people remember, what reaches the mediaand what make some people household names. These are rare. Secondly, there is ‘‘incremen-tal development’’ which is how the vast majority of scientists work. Small parts of a big prob-lem are dissected out, solved and contribute to the progress of a bigger field. Sometimes thesecond way leads to the first, often unplanned to be so.

st European Conference on Medical Physics, Il Ciocco, Tuscany, Italy.08 643 3812..uk

8 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica.

Combatting cancer in the third millennium 43

To set the scene I briefly told the story of a few famous ‘‘big hit science’’ stories in whichmedical physicists have played a leading role e the invention of X-ray computed tomography(CT), the development of intensity-modulated radiation therapy (IMRT) and the invention ofthe emission tomography imaging modalities: single-photon emission computed tomography(SPECT) and positron emission tomography (PET).

I suggested some of the areas I consider important for development. Some are fairly easy toidentify and others are more speculative and unusual.

I suggested that the goal of medicine and supporting science is to ensure that people livelong and die quickly and I contrasted this with the past scenarios.

Digressing to philosophy I suggested that there may be a difficulty in that the highly devel-oped world works mainly to make itself even more developed and that many developed gov-ernments and aspiring medical physicists may not be as interested as they should be inassisting developing countries. There is therefore, sadly, likely to be an ongoing imbalanceof resources. Scientific publishing is also at cross roads where the need to act independently,openly and with wide availability clashes somewhat with the need to generate revenue andsupport learned societies.

Turning to detailed observations, I described how I think the following fields can beadvanced: (1) improving the diagnosis of disease, (2) improving the planning of radiotherapy,(3) improving the delivery of radiation treatment and (4) improving the assessment of responseto treatment.

I ended on a highly philosophical note, which is somewhat critical of how much practicalmedical physics is currently organised in universities and hospitals and I suggested what shouldbe the real agenda for scientific progress.ª 2008 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica.

The perils of prophets

A prophet stands in front of four listeners. He says‘‘I predict that all my predictions will be wrong’’. Oneonlooker responds ‘‘Gosh! He said that last year and he wasright!’’ [4]. The moral is obvious. If one makes empty pre-dictions it takes us nowhere. In his seminal book for chil-dren ‘‘Alice’s Adventures in Wonderland’’ Lewis Carroll,alias the Professor of Mathematics at Oxford University,Charles Dodgson, has Alice meeting the Cheshire Cat ina wood. Alice asks ‘‘Would you tell me please which way Iought to go from here?’’ The smiling Cheshire Cat replies(before disappearing, leaving only its smile) ‘‘That dependsa good deal on where you want to get to’’. This is a wise an-swer. One cannot form scientific plans without clear goals.

I was flattered to be asked by the Conference President,Professor Alberto del Guerra of Pisa, to speak to thisoutrageously broad and bold title. It is said that they onlyask you to predict the future if you are very old and/or willnot be living long enough to know if you were correct.Hopefully I will live long enough to see if some of my pre-dictions come true and also to contribute to some ofthem. However, ‘‘crystal ball gazing’’ is a very unscientificprocess because physicists are trained to study and analysesituations, report their findings, draw conclusions and stopat that. Predicting the future is almost the antithesis of thescientific process. So-called prophets can perhaps be enter-taining, but at worst they look egocentric and possiblyridiculous. I seem to be a popular target for role playingas prophet having been asked three times to do this beforewith accompanying publications [1e3]. I seem to have sofar survived in this role. ‘‘Future gazing’’ is not, of course,predicting the short-term developments, which oftenamount to no more than ‘‘more of the same’’. It is about

being bold, radical, thinking laterally and stating whattoday is impossible and unthinkable.

As light relief, and to underline this theme, we mightquote some famous prophetic remarks in the past, whichproved somewhat wide of the mark [5]. Thomas Watson,chairman of IBM said in 1943 ‘‘I think there is a world mar-ket for maybe five computers’’. Ken Olsen, president,chairman and founder of DEC said in 1977 ‘‘There is no rea-son anyone would want a computer in their home’’. Theeditor in charge of business books for Prentice Hall said in1957 ‘‘I have travelled the length and breadth of this coun-try and talked with the best people and I can assure youthat data processing is a fad that won’t last out theyear’’. Bill Gates said in 1981 ‘‘640K ought to be enoughfor anybody’’. How wrong they all were.

It should be said, upfront, that this paper is unusual. Ithas no data, no graphs, no materials, methods, results anddiscussion. It is the printed form of a lecture in which I wasinvited to give opinion. Opinion is not normally the stuff ofpapers. I hope what I write is valuable but, were thisa regular paper in a regular journal, it should probably bedestroyed by the referees. This paper is effectively anInvited Commentary.

The lessons of history

Whence comes important progress? I propose that there aretwo major routes. It is important to distinguish between‘‘big hit science’’ and ‘‘incremental development’’. ‘‘Bighit science’’ is the invention or discovery of something soimportant that the medical world changes forever becauseof it. It is what people remember, what reaches the mediaand what makes some people household names. This israre. ‘‘Incremental development’’ is how the vast majority

Figure 1 The EMI prototype X-ray CT scanner, unveiled in1972 and now displayed in the London Science Museum along-side the Apollo 10 capsule. The CT scanner is an example of‘‘big hit science’’ which built on earlier incremental research.

44 S. Webb

of scientists work. Small parts of a problem are dissectedout and solved. Sometimes the scientific method of in-cremental development leads to a big-hit outcome and Ican illustrate this notion with three different histories. It iswith this in mind that we, as ‘‘jobbing scientists’’, can beassured that the day-to-day travail, careful publication andreview might, with some added lateral thinking, lead toa surprising and important outcome. It is otherwise highlyunlikely we can ‘‘set out to discover something major’’. Weshould, instead, identify a problem, study it carefully and,of the large number of us doing this, someone will makesignificant history. The others would have played their part.

The invention of X-ray computed tomography is anexample of ‘‘big hit science’’. In April 1972 the EMICompany announced to a largely unsuspecting world theinvention of the world’s first X-ray CT scanner. This machinewas capable of scanning a human head, relatively slowlyand differentiating white and grey matter. The Presscarried lengthy text and pictures drawing this immediatelyto public attention. The event was described as the great-est radiological advance since the discovery of the X-ray onNovember 8th 1895. It replaced erstwhile dangerousinvestigations such as the injection of intra-cranial air andthe Somersaults of Zeidses des Plantes. Lives, otherwiselost, were saved. CT also transformed radiotherapy plan-ning. In reality the invention had taken place during theprevious 5 years or so, led by Sir Godfrey Hounsfield(28.8.1919e12.8.2004), who later shared the 1979 NobelPrize for Medicine with Alan Cormack (23.2.1924e7.5.1998) who, in South Africa, had made pioneering labo-ratory measurements and reconstructions. The world’s firstCT scanner can today be seen in the London Science Mu-seum alongside the Apollo 10 lunar orbiter e what juxtapo-sition (Fig. 1)! CT scans appeared on a British postage stampin 1994. Hounsfield is honoured as one of the ‘‘1000 Makersof the 20th Century’’ [6] (Fig. 2). This book was controver-sial. It collated scientists, humanists, religious leaders, pol-iticians alongside dictators and pop stars. Whether for goodor for bad, these people changed the world as we know itand Hounsfield was in the former category.

However, 1972 was not the year in which CT was born.The principles of CT had been known for years and manyother scientists ‘‘almost invented’’ the CT scanner. Apatent by Gabriel Frank in Budapest in 1940 showeda technique and apparatus for back projected tomography[7]. Dozens of other precursors are known. The full story istold in one of my books [8]. How much the EMI inventorsknew of these earlier developments is highly controversial,but it is believed much prior art was understood. Theachievement of EMI and Hounsfield was to make CT scan-ning a practical reality. Others had fallen short of thispracticality.

A second example of ‘‘big hit science’’ is the inventionof intensity-modulated radiation therapy (IMRT) which hasrevolutionised cancer treatment by radiotherapy. Mostworkers would agree that a seminal paper [9] by AndersBrahme et al., in which a special inverse problem in radio-therapy was solved, was the beginning of IMRT, although itwas not called by that name at that time. Brahme in 1988produced one of the first papers on what then becamecalled inverse planning. He visited key international cen-tres such as DKFZ Heidelberg, Memorial Sloan Kettering

Cancer Center, New York and the Royal Marsden Hospital,London and work in those centres (DKFZ: Schlegel and Bort-feld; MSKCC: Mohan and Mageras; RMH London: Webb) wasamong the very few early research activities on IMRT. In1992 Mark Carol announced the NOMOS MIMiC, the world’sfirst purpose-built equipment for actually delivering IMRT.(Previous work had largely concentrated on planningIMRT.) The first treatments were made at Baylor in April1994 and this delivery method dominated the world deliv-ery of IMRT for at least 3 years thereafter before multileafcollimator (MLC)-based methods were used. The first ever‘‘IMRT School’’ was held in the Strater Hotel, Durango, Col-orado, May 17the18th 1996. Thomas Bortfeld and Art Boyerperformed manually the world’s first MLC-based IMRT in1993, an experiment now much copied for IMRT qualitycontrol.

But this invention of IMRT also did not really come fromnowhere. In fact, it was the result of decades of in-cremental science [10]. In 1940 the American mathemati-cian George Birkhoff had published the theory of IMRTwithout realising it. He showed that any drawing could beconstructed from an array of straight lines from differentdirection, provided the lines could have any blackness in-cluding negative blackness [11]. The blackness of a drawingat any point was the sum of the individual blackness of allthe lines passing through that point. Basil Proimos in 1960

Figure 2 The cover of Ref. [6].

Combatting cancer in the third millennium 45

showed that concave isodose distributions could be createdby crude forms of modulated beams [12]. In the mid 1980sYair Censor wrote some seminal papers. These stories areamplified in my four books on the physics of radiation ther-apy [13e16].

A third example of ‘‘big hit science’’ is the invention ofmodern equipment for SPECT and PET, the pre-history ofwhich goes back 50 years [8].

These three stories serve to underline that ‘‘big hitscience’’ almost never emerges from a background ofnothing. It generally emphasises the words of Isaac Newton:‘‘If I have seen further than most men it is through standingon the shoulders of giants’’. So it might be extrapolatedthat the route to achieve something very important is tothink laterally from ‘‘ordinary research’’ and build on thework of others.

Looking to the future

It is of course much easier to look back (than to lookforward) and see from where we have come and todocument the big achievements of the past. It is a properscientific process. Facts are known. Data are at hand.Looking to the future is an unscientific process. We are inthe realms of hope, expectation and promise. Nothing is forsure. If it were for sure what would happen next in researchwe would not be research scientists. However, we can havegoals and can identify some key fields of endeavour andplan the process. What goals and what fields?

Research in SPECT, PET, MRI, MRS, ultrasound and opticsshould continue to be fruitful. High speed digital X-ray andCT imaging will advance. Focused ultrasound therapy is inits clinical infancy and likely to be better used in the futureand more clearly understood. The human genome projectis complete and understood but capitalising on this

knowledge and delivering gene-based therapies is in itsinfancy. There is a need to gain a better understanding oftissue motion on many timescales and to use this in-formation to improve the physical basis of radiotherapyfor the moving patient (rather than the presumed staticpatient or phantom). Nanotechnology is in its infancy andthe use of medical nanoparticles should expand. Forexample they can be used as contrast agents, becomevehicles for drug delivery, act as microsensors and measurelocally pH, temperature, drug concentration, dose, DNAdamage and may be cell survival. Nanocrystals of semi-conductor material coated with a shell, quantum dots,might develop further. Already super paramagnetic ironoxide (SPIO) is used to label cells and gadolinium is used asa paramagnetic MRI contrast agent. Ultrasound micro-bubbles are already in use and could become moresignificant deliverers of drugs to local tissues.

More unusually it is claimed that high-intensity pulsedelectric fields applied to cancer cells can significantlyenhance the uptake of cytotoxic drugs. It is possible thatphotochemical internalisation can treat cancer byreleasing drugs internally due to the light shining onselected areas. It would be interesting to explore if thereis a genetic link to late radiation toxicity. As well asadvanced molecular imaging, and biomedical optics mayfurther aid cancer screening. On a technical, and maybeless profound, level might holographic display methodsbetter improve the display of medical images and radio-therapy plans?

I was stimulated by an excellent overview of the futuretrends in radiotherapy physics by Jerry Battista at the 15thICCR in Toronto [17] in which he shared some of the almostincredible trends in computing and medical computing thatcan be expected to continue. Most technologies haveaccelerated experiential growth. So the time to research50 million users is variably 38 years for radio, 16 years fortelevision, 13 years for the PC, 4 years for the Internetand 1 year for wireless Internet. Such is our embrace oftechnology. The computer growth factors, measured in2005, were a doubling time of 2 years for transistors permicroprocessor, 1.8 years for million instructions per sec-ond (MIPS), 2 years for MIPS per $1K, 2 years for RAM Mbytesper $, 1 year for storage Mbytes per $ and 1.2 years for In-ternet Mbytes per second. The speed of computers has gonefrom 10�6 MIPS per $1000 in 1930 to 10þ3 MIPS per $1000 in2005, a speed only about three orders of magnitude slowerthan that of the human brain. Moore’s law has it that just asone technology ‘‘runs out of steam’’ so another takes itsplace and so we have seen the sequential use of electrome-chanical punch cards, the relay, the vacuum tube, the tran-sistor and the integrated circuit. A fascinating lecture onDNA computing by Lila Kari [18] showed the power of allthe world’s current computers could be had in a test tubeof DNA and the solution of problems, many orders of mag-nitude faster than today, will be possible. When that hap-pens just dream of what medical computing advances willbe possible! Finally, the Internet backbone has leapt from104 bits per second in 1970 to 108 in 1995 (Internet proper)to 1013 in 2005. We may expect further growth and themedical possibilities will grow. Much progress in cancerresearch will come simply through the acceleratedcomputer power and storage at lower cost.

46 S. Webb

Life and death

Let us return to the goals for cancer therapy. Firstly wemust recognise that over the years, decades, and centuriesthese have changed (Fig. 3). Primitive man died young ofaccidents, malnutrition, poor housing and probably hadno cancer at all. He simply did not live long enough. Maybea wild animal consumed him in his prime? A 1950 man, as weshall call him, probably wore himself out with normal work,had fought in a world war, may have had poor diet andsmoked too much tobacco. He died early after retirementand had little cancer. A 2007 man in most developed coun-tries has better housing, better nutrition, perhaps a lessmanual working life and lives many years into retirement(or may not even retire at all). But as he ages with chronicill health and cancer, quality of life may seriously deterio-rate. Life literally may not be worth living. So a good cancertreatment is not necessarily to cure cancer but to improvethe quality of life for cancer survivors. Indeed, given thatmany primary tumours can now be cured, sparing organsat risk becomes of surfacing importance. As more people‘‘survive cancer’’ they need to live well and, if cancerrecurs years later, may well need re-treatment. The earliersparing of organs at risk will be of paramount importance.I like to quote from John Grimley Evans, Professor of Geron-tology at the University of Oxford, that the goal is to ‘‘livelong and die quickly’’ [19]. Amusingly, he tells a story that,given a choice between guaranteed life to age 80 the imme-diate death and, alternatively, taking the chance of un-known years of survival, most people will choose thelatter despite the statistical advantage of the former. Wemight also observe that the human body is not designedto survive indefinitely and, even if cancer is eradicated,some other chronic disease must surely escalate to causeeventual death. I am not being unnecessarily morbid butlogical, unless some method is found to rebuild organs.Would this even be desirable?

Uncomfortable truths

The developed world tends to concentrate on making itselfeven more developed. High technology is at the heart of

health

20 40 60 80 A 100

Age (years)

100% (aliveand welland happy)

0% (dead!)

Killed by wild animal

Primitiveman

1950s man

2007 man withchronic old agedisease

Idealman

But what is thisideal value A?

Figure 3 The goal of cancer research e to live long andhealthily, placed in context with other less desirable scenarios.

this. Yet many developing countries have little hope ofaccess to this. We would probably do more, ‘‘world good’’by translating and adopting current techniques to the needsof developing countries. However, this is not the highestpriority for many developed governments or aspiring youngmedical physicists who need to forge their careers. So ittends to not happen. Perhaps we should feel a bit guilty andchange that?

A second somewhat philosophical, but also practical,difficulty is that, in an ideal world, scientific informationshould not be ‘‘owned’’ but be freely available. Indeed,many scientific funding bodies want ‘‘open-access publish-ing’’. However, someone has to pay publishing costs.Journal owners (usually learned societies) need to makemoney to keep their membership subscriptions low. Journalpublishers also need to make some financial gain to at leastcover their costs. There have been many learned articles onthis fundamental dichotomy.

Wish lists for the third millennium

Turning to the future of radiotherapy of cancer let me makesome lists of what areas should develop. Some statementsmight be construed to be little more than extensions ofcurrent foci of research. Others are verging on theoutrageously unthinkable to those working in today’sconstrained climate.

Improving the diagnosis of disease

More thorough screening programmes should be developedto lead to a shift in stage profile and aim towards a greaterfocus on cancer survival. 3D multimodality imaging should beroutinely requested for symptomatic patients alongsideblood tests and genetic tests. The prevailing diagnosticinformation will shift to functional rather than anatomicalinformation. A holistic diagnostic picture should be con-structed. Telemedicine could become the norm. Doctorsmight not need to be local to the patient and could be calledfrom anywhere in the world by Internet. Provided suffi-ciently carefully selected diagnostic information were col-lected, actually seeing the patient might not be needed.Robotic palpation could even be developed. This wouldwiden the availability of high quality diagnostic care. As longago as 1998 it was suggested [20] that the normal populationcould be monitored by implanted biosensors and that thesecould be read by home-room-based sensors which couldeven automatically summon a doctor when a problem arose.The patient might even be asymptomatic at this stage. Sadlythis prediction has not materialised in the ensuing 9 years,perhaps serving to emphasise the long time scale associatedwith futuristic (almost science fiction) predictions.

Improving the planning of radiotherapy

Treatment planning should be patient-individualised basedon measured radiosensitivity and response to tissue assaysand 3D functional imaging. Functional and anatomical 3Dinformation should be merged and fuzzy logic may need tobe expanded to assign structures to the imaged data. Thequest, so far largely fruitless, to create tissue contours

Combatting cancer in the third millennium 47

automatically with minimal human intervention shouldcontinue. If all volumes of tissue could be contoured anddose data recorded, then a database of response versusdelivered dose can be prepared. The treatment itself couldbecome multimodality with combined photon, proton andion therapy merged with brachytherapy, radionuclide ther-apy and even targeted drug therapy and focused ultra-sound. The use of single-modality treatment and worse stillclass solutions (largely introduced for ergonometric rea-sons) could become a thing of the past. Multiple plansshould be prepared per patient and these should becompared to select the best. All planning should be 4Dwith multiple imaging throughout the course of treatmentand adaptive readjustment of plans to accommodatechanged functional status and geometry with knowledgeof the already-delivered dose. Dose calculations will beable to be made entirely by Monte Carlo cumulative historytechniques and there will be no further need for all thosequaint, artificially invented, 20th century parameters suchas TAR, TMR, OF, depthedose, etc. Such dose calculationswill also routinely predict biological outcome.

Improving the delivery of treatment

All treatments should include truly integrated feedback intothe delivery process. Before each treatment fraction 3D on-linac imaging will inform to reposition the patient accom-modating inter-fraction changes. If accumulated dose hasbeen calculated then closed-loop adaptive feedback can beused to modify the delivery of the subsequent fractions. 4Dimaging should be used to monitor intra-fraction motion andfeed this information back to the delivery equipment tocorrect for it. To this end predict-ahead methods will bedeveloped to overcome the latency between acquiring suchinformation and correcting for it. It could be envisaged thatmultiple robots be used in the same room to treat thepatient, just as multiple robots (not one) are used toconstruct automobiles. Every country should have at leastone hadron therapy centre with a particle accelerator forcancer therapy. It should act as a National Referral Centreand this will allow to gain experience to join the debate as tothe relative usefulness of hadron therapy. In 2007 a compactproton accelerator is already under construction with thehope that such equipment could become as common asphoton therapy [21]. Micro-beam radiotherapy is just start-ing in the laboratory and should be progressed. Finally, isthere the possibility that gene therapy and radiotherapycould synergistically combine?

Improving the assessment of response to treatment

All patients should have tissue samples taken to enabletreatment outcome to be related to biological mechanisms.All patient dose, treatment outcome and tissue assay datashould be stored centrally for future recovery and analysis.Patients should fill in patient-specific symptom-and-response data. Symptom, outcome and dose data will bebrought together to refine the biological models of out-come. This post-treatment data collection will becomestandard routine practice rather than the (current) excep-tion. Data will be co-ordinated through international trials.

Planning the future: conditions for progress

These developments and outcomes will not happen unlesssuitable conditions are engineered. The role of medicalphysicists must be central. Firstly, the future supply ofmedical physicists must be ensured by educating schoolchildren to find science (which is difficult to do!) rewardingas a specialty to continue. Post-school, they should be ledto understand that, while specialising in medical physics ispossible and desirable for many people, many medicalphysics breakthroughs have come from applying the physicsfrom mainstream physics research areas which also need tobe studied.

Whilst training they should be supported by National andInternational Organisations with suitable training schemesand education programmes. However, these schemesshould not become stifling and over-long or enthusiasmwill wane if the practicalities of living on training or studentsalaries become unmanageable. There should not be toomany professional ladders to climb. The Bologna processshould be implemented with border-free movement ofinternational physicists.

As for the conditions of performing research, physicistsmust be allowed enough ‘‘big me time’’ to do researchproperly. Research cannot be fitted into small time gapsbetween competing clinical duties and service. The latterare important too for the here-and-now but only throughresearch can medicine improve.

At present there is, in many countries, a tension be-tween pursuing a university research career in medicalphysics and a service career. The goals from the patientviewpoint are the same but, from the viewpoint of theuniversity and service hospital, they are very different. Atpresent many young medical physicists have to make anagonising choice. It should be the duty of those who set thecareer agendas to address and ease this problem. Researchphysicists should be also allowed to follow blind alleys andsometimes fail. Too much grant-driven and goal-specifiedresearch with over-precise ‘‘deliverables’’ will kill theprocess. A role for senior physicists is to encourageavoidance of repetitive re-invention. Too much peer-reviewed publication and conference presentation are, inmy view, not sufficiently novel. The need to publish forcareer and research assessment processes must not over-take the need to be truly inventive. Finally e and this maybe a little inflammatory e medical physicists need to beactually doing medical physics, not spending too much timeon administration, business, grant writing, scientific politicsand unnecessary committee and professional society work.All these are, of course, necessary, but medical physicistsshould remember that they partake in these mattersbecause once they were full-time active workers. Theyshould at the very least, stay active part-time researchworkers and/or clinical service scientists.

Conclusions

Predicting the long-term future is not a scientific process,but it is possible to make wish lists and to define goals.Reviewing the past is easier. In this Commentary key ‘‘bighit science’’ achievements have been identified and linked

48 S. Webb

to ‘‘incremental science’’. Areas in which to work for thefuture have been identified. The goal of cancer researchhas been re-defined in terms of increasing the quality of lifefor cancer survivors. The influence of computer progresshas been highlighted. The atmosphere in which medicalphysics research must grow has been elucidated.

References

[1] Webb S. The future of photon external-beam radiotherapy:the dream and the reality. Physica Medica 2001;17(4):207e15.

[2] Webb S. Radiotherapy physics: the next ten years of technicaldevelopment. Imaging and Oncology 2005;1:43e50.

[3] Webb S, Evans PM. Innovative techniques in radiationtherapy editorial, overview and crystal ball gaze to thefuture. In: Webb S, Evans PM, editors. Innovative techniquesin radiation therapy. Seminars in Radiation Oncology 2006;16(4):193e8.

[4] From a (not remembered which) British newspaper cartoon.[5] Extracted from Internet sites: <http://ifaq.wap.org/

computers/famousquotes.html>; <http://www.pcsndreams.com/Pages/TechQuotes1.htm>.

[6] ‘‘1000 makers of the 20th century’’ Sunday Times book; 1990.Section on Godfrey Hounsfield written by S. Webb. Book orga-nized alphabetically by person featured.

[7] Frank G. Verfahren zur Herstellung von Korperschnittbildernmittels Rontgenstrahlen. Deutches Reich patent number693374; 1940.

[8] Webb S. From the watching of shadows e the origins ofradiological tomography. Bristol: IOPP, ISBN 0-85274-305-X;1990.

[9] Brahme A, Roos JE, Lax I. Solution of an integral equationencountered in rotation therapy. Physics in Medicine andBiology 1982;27:1221e9.

[10] Webb S. Historical perspective on IMRT. In: Palta JR,Mackie TR, editors. Intensity-modulated radiation therapy ethe state of the art; 2003. p. 1e23.

[11] Birkhoff GD. On drawings composed of uniform straightlines. Journal des Mathematiques Pures et Appliques 1940;19:221e36.

[12] Proimos BS. Synchronous field shaping in rotational megavolttherapy. Radiology 1960;74:753e7.

[13] Webb S. The physics of three-dimensional radiation therapy econformal radiotherapy, radiosurgery and treatment planning.Bristol: IOPP; 1993. ISBN 0-7503-0247-X; 0-7503-0254-2 Pbk.

[14] Webb S. The physics of conformal radiotherapy e advances intechnology. Bristol: IOPP; 1997. ISBN 0 7503 0396 4; 0 75030397 2.

[15] Webb S. Intensity modulated radiation therapy. Bristol: IOPP;2000. ISBN 0 7503 0699 8.

[16] Webb S. Contemporary IMRT e developing physics and clinicalimplementation. Bristol: IOPP; 2004. ISBN 0 7503 1004 9.

[17] Battista J. The use of computers in radiation therapy: dynamicfuture. In: Bissonnette J-P, editor. Proceedings of the 15thinternational conference on computers in radiation therapy;2007. p. 7.

[18] Kari L. see website: <http://www.csd.uwo.ca/wlila/>.[19] Grimley Evans J. Ageing. In: Marinker M, Peckham M, editors.

Clinical futures. London: BMJ Books; 1998. p. 170e83.[20] Delpy D. Engineering. In: Marinker M, Peckham M, editors.

Clinical futures. London: BMJ Books; 1998. p. 43e73.[21] Caporaso GJ, Mackie TR, Sampayan S, Chen Y-J, Blackfield D,

Harris J, et al. A compact linac for proton therapy based on a di-electric wall accelerator. Physica Medica 2008;24(2):98e101.