3D2: Three Decades of Three-Dimensional Dosimetry

The development of three-dimensional (3D) dosimetry was motivated by its promise as an effective methodology for the validation of the complex dose distributions achieved by modern techniques such as Intensity Modulated and Volumetric Arc Radiation Therapy. 3D techniques were first proposed in the 1980s when clinics were just starting to move from two-dimensional contour plan based delivery to more conformal techniques. Advances in dosimeter materials, readout systems, and workflow and software systems for the registration and analysis of the volumetric dose data have made 3D dosimetry more attainable, yet to date it has not made major inroads into the clinic. This keynote address will highlight some 3D dosimetry developments over the years, many of which were first revealed through the past twelve International Conferences on 3D and Advanced Radiation Dosimetry (IC3Ddose). These conferences resulted in the publication of more than 130 didactic review articles and over 650 proffered research papers, the majority openly available on the internet (years before the current drive to publishing in open access journals). In this joint keynote address to the IC3Ddose community (at the end of its conference) and to attendees of the Canadian Organization of Medical Physicists annual scientific meeting (at its start), I will briefly review the radiation sensitive materials used for 3D dosimetry, the imaging systems required to read out the volumetric dose information, the workflows and systems required for efficient analysis, and the protocols required for reproducible dosimetry, and how the dosimetry has come into the clinic. The address includes some personal reflections of the motivational and practical changes in 3D dosimetry over time. And as we are all meeting in person for the first time in over two years, the address will end with some observations on the importance of conferences for the exchange of ideas and associated debate necessary for scientific advancement.


Introduction
This year's meeting of the IC3Ddose community is the 12 th in a series of conferences on three dimensional and advanced clinical dosimetry [1][2][3][4][5][6][7][8][9][10][11].Therefore, it is fitting in this keynote address to review three decades of three-dimensional (3D) radiation dosimetry through the lens of these conferences.The field effectively started with Gore's seminal paper in 1984 [12] and then developed into an active discipline as the medical physics community addressed the challenges in dose measurement associated with the implementation of conformal radiation delivery techniques.These techniques which better limit the irradiated volume to the tumour were foreseen as early as the 1950s [13] but came into wide clinical practice in the 1980s and 1990s with advances in imaging, treatment planning, and external beam techniques such as Intensity Modulated Radiation Therapy (IMRT).The move to conformal dose delivery and planning was challenging and initial clinical implementation of IMRT took considerable effort.The careful investigations of using multileaf collimators (MLCs) rather than poured blocks for field delineation [14,15], of the impact of MLC leaf edges on treatment planning [16,17], and the introduction and validation of dose calculation algorithms able to predict dose from dynamic photon fluence patterns, etc., are mainly long forgotten.Within the next decade the use of conformal techniques increased in the clinic through Volumetric Modulated Arc Therapy (VMAT), Stereotactic Ablative and Fractionated Stereotactic Radiation Therapy (SABR and FSRT), and through improved procedures for high-dose rate brachytherapy.Of course, radiation therapy using these advanced delivery techniques would only give the intended benefit if resulting dose distributions correctly registered to the patient's anatomy.Thus, there was a heightened requirement for quality assurance (QA) both for technical components of the delivery (including the treatment planning systems, the treatment unit and its associated on-line imaging systems), and for the validation of the dose delivery planned for individual patients.The requirements for delivery validation motivated much of the initial interest in 3D dosimetry especially during the early stages of IMRT development when the main dosimetry was predominately with point detector ion chambers (see figure 1).The possibility of measuring actual IMRT delivered doses in 3D in homogeneous phantoms under standard irradiation conditions and in anthropomorphic phantom tests mimicking patient conditions was particularly attractive for IMRT implementation [18,19].With time other tools were developed to provide dose delivery measurement and validation.Radiochromic film [20], 2D and 3D arrays of point detectors [21][22][23], electronic portal imaging detector (EPID) systems [23][24][25][26], scintillator dosimeters [27] and Cherenkov based dosimetry [28] became clinically very useful.But these dosimetries typically provide sparse 3D data and only surrogate validation of 3D dose delivery and are not considered full 3D dosimeters.

3D Dosimetry
The features of ideal "true" 3D dosimeters were set by Oldham through his RTAP criteria [29].Under RTAP a 3D dosimetry system (dosimeter and associated readout) would provide dose measurements throughout a 3D volume with 1 mm isotropic spatial resolution in less than one hour with 3% accuracy and a precision of 1%.Occasionally the criteria for the resolution, accuracy and precision have been relaxed in clinic practice, depending on the specific validation being performed (for example, in external dose delivery audits under IROC [30,31]).Nevertheless, the criterion for high resolution isotropic measurement limits full 3D radiation dosimetry to chemical radiation dosimetry based on quantifying the effects of radiation-induced chemical changes occurring within some volume of material [29,32,33].The clinical applicability of 3D dosimetry has advanced considerably in the last decades through the development of improved dosimeters [33] (e.g., radiochromic plastics [34,35], radiochromic gel dosimeters [36,37] and normoxic and less toxic polymer gel systems [38,39]) and by improved readout protocols using optical computed tomography (CT) [40][41][42][43][44][45], x-ray CT [46,47], or magnetic resonance imaging [48,49].These advances were often first reported in the proceedings of the DosGel and IC3Ddose International Conferences on 3D Radiation Dosimetry [1-11].Moreover, the 135 invited review papers and 651 submitted proffered papers therein (including the material in these proceedings, see table 1) provide a detailed account of the evolution of three-dimensional dose measurement [50].
Each conference has included a strong didactic component with invited papers that provide the fundamental science and basic mechanisms that guided the development and response of various 3D dosimetry systems.The reviews extend also to various 3D readout techniques, describing the properties of the irradiated materials enabling dose measurement with the various imaging modalities, and providing details that inform readers of the imaging protocols they must adopt to achieve reproducible and accurate dose readout.Review and proffered papers also discuss the data analysis and data workflow integral in 3D dosimetry.Initially the focus was on data analysis concentrating on the scientific fundamentals of converting readout to dose, with emphasis on uncertainty and error analysis.Eventually the reports evolved more to description of the analytic tools required for fast efficient data analysis and for the comparison of 3D measurements to the dose distributions from treatment planning [51][52][53].These reviews, and the proffered papers throughout the proceedings, provide multiple reports on practical issues in 3D dosimetry and offer a compendium of best practices for dosimeter preparation (including setting timescales for various steps in the preparation of the dosimeters, periods between sample preparation and irradiation, and then time to readout [33]) and include clear caveats and practices to avoid.The proceedings offer the medical physics community a valuable resource highlighting the importance of developing and using careful dosimetry protocols to ensure consistent Table 1.A historical review of the twelve International Conferences on 3D and Advanced Radiation Dosimetry originally instigated as the DosGel conferences to provide a forum for researchers developing 3D dosimetry.The meetings were renamed in 2010 to acknowledge that a full complement of radiation dosimetry technologies is required in clinical dosimetry.Ten of the proceedings have been published in the refereed IOP open access Journal of Physics: Conference Series, the first two proceedings are available by request from the author.These proceedings offer a comprehensive summary of developments in the field, each proceedings offering didactic review articles on select topics proffered papers presenting initial reports of many novel advances in 3D and other advanced dosimetry over the conferences' three decades.and reproducible validation of clinical dose delivery.These advances enabled multiple groups to present examples of the use of 3D dosimetry in the technical and clinical implementation of new techniques and treatments within their clinics [50].

3D dosimetry and more: lessons for COMP from DosGel and IC3Ddose
While the proceedings have from the onset [54] presented a strong motivation for the clinical use of true 3D dosimetry, they have signalled that the needs of the clinic [19,[55][56][57][58] extend the required dosimetry  [59,60].Other dosimetry techniques are more efficient and effective in certain roles ( [61] and Table 1 in ref. [50]).The meetings were renamed from 'DosGel' to 'IC3Ddose' to explicitly recognize that different dosimetry techniques are needed in the clinic.This extended the conference focus and encouraged attendance to a wider community with expertise in film, EPID based [23][24][25][26], scintillation [62,63], Cherenkov [28] and point array measurements [21,64].This evolution enabled a more critical assessment of the role of the different systems and how they complement each other [64,65].
If the legacy of IC3Ddose was only in its multiple proceedings, then one might condense their impact to a simple technical description of the various 3D and other clinical tools available for dose delivery validation.A truer analysis would have to acknowledge that proceedings are only part of what a conference provides to a community.For example, multiple IC3Ddose presentations during the meetings have reported on the successful use of a particular dosimetry technique in some particular QA test or treatment verification, but questions and discussion after the presentations often deliberated whether the tests were appropriate, or if other techniques would have been better suited for the task.Hence, there was a scientific exchange during the meeting itself that benefited attendees and influenced further progress.This suggests some important closing points to consider in this keynote address as the current IC3Ddose meeting concludes and the Canadian Organization of Medical Physicists annual scientific meeting starts.
The first point is the technical challenge for the medical physicist that presents itself after nearly three decades of IC3Ddose experience: which particular dosimetry system should be used for a particular test in the clinic.The answer must take into consideration many components: including determining whether it is suitable for the specific radiation delivery; whether another dosimeter is simpler to use with measurements more easily analysed and interpreted; whether or not the software for a particular device (especially when commercially supplied) is well documented so that the output can be validated independently; whether the dosimetry system is readily commissioned and can be validated against other dosimetry systems already well-established in your clinic, etc.This leads to a first point (which has been called Schreiner's First Commandment [58]): 'know and understand your dosimetry system completely, including its limitations, before applying it to a particular validation task' [61].
The second point is communal, recognizing the importance of regular interaction for scientific progress [66] and has been represented as Schreiner's Second Commandment [58]: 'engage in the clinical and scientific exchange of ideas and knowledge through publication in scientific journals, and, perhaps more importantly, through regular communication, meetings and workshops with colleagues locally, nationally and internationally' [61].This commandment is perhaps especially relevant at our meetings in Quebec City as we have come from two years of scientific separation during COVID-19.

Conclusions
This keynote address to attendees of the IC3Ddose and COMP conferences is a personal review of the evolution of 3D dosimetry over nearly three decades.The review highlights the influence of the IC3Ddose meetings in this evolution.The work reported over four days at IC3Ddose 2022 in Quebec City and for four more days at the COMP AGM is important as it has considerable influence on the clinical success of radiation therapy [67,68].

Acknowledgments
The ideas presented in this address have been formed through research funded through Canadian CIHR project MOP-115101, CHRP projects CIHR CPG 151964 and NSERC CHRP 508528-17, and through many years of interaction with many colleagues in the IC3Ddose and COMP communities.I especially thank Kim McAuley, Tim Olding, Kevin Alexander, John Miller and Clive Baldock who could not make the meeting this year.

Figure 1 .
Figure 1.An illustration of the advancement of radiotherapy dose delivery over the last four decades to better treat a complex shaped target (upper left inset).The treatments have progressed from simple uniform doses delivered to the target and surrounding normal tissues as with a 4-field box, through treatments with individual ports conformally blocked to reduce normal tissue irradiation, to fully conformal dose delivery achievable with dynamic IMRT and VMAT deliveries.These dose delivery advances were anticipated from the 1980s onwards and it was thought that dose delivery validation would require commensurate advances in radiation dosimetry techniques which motivated the development of 3D dosimetry (as indicated in the text at the bottom of the figure).

Figure 2 .
Figure 2. a) Three dimensional dosimeters are chemical systems with some specific molecular constituents that change under irradiation.The degree of change is dose dependent.Three different hydrogel dosimetry systems are shown in b) to d). b) Four different ferrous xylenol orange gel (FXG) dosimeters; the three dosimeters on the right have been irradiated by electron beams of increasing energy.The spatial information in these Fricke gels will disappear within a few hours because of diffusion of the indicator ferrous and ferric ions.c) Two views of a poly-acrylamide (PAG) polymer gel dosimeter after a Cobalt-60 beam irradiation.The photograph on the left was taken soon after the dose delivery in 1998, the image on the right was taken over two decades later in 2021.d) A radiochromic leucodye micelle gel dosimeter irradiated with a 6MV external beam delivery mimicking a prostate treatment.The photograph was taken a few weeks after the irradiation.