Patient views on the implementation of artificial intelligence in radiotherapy.

PURPOSE/OBJECTIVE: To date there has been limited research looking at patient views on the implementation of artificial intelligence (AI) in radiotherapy. The aim of this study is to adapt and utilise a validated patient questionnaire to develop an understanding of current patient views on the use of AI in radiotherapy. MATERIALS/METHODS: An existing questionnaire, developed to assess understanding of patients' views on the implementation of AI in radiology, was adapted to the field of radiotherapy. The questionnaire was distributed to cancer patients receiving radiotherapy treatment between November 2021 and March 2022. Completed questionnaires were analysed to assess patient levels of positivity or negativity towards AI. Results were grouped into five factors, representing underlying patient perspectives, and correlation of factors with demographic variables was assessed. RESULTS: In total, 95 patients participated. Overall, there was a moderately negative patient view towards the use of AI in radiotherapy. Certain factors drew a more negative response than others, for example patients desire significant personal interaction with healthcare professionals during the course of their treatment. No significant correlation was found between the demographics of age and gender and the strength of views towards the use of AI in radiotherapy. CONCLUSION: This study has found that there are clear patient concerns around the use of AI in radiotherapy. As the use of AI in this field increases in future years, it will therefore be extremely important to educate and involve patients in the future direction of this technology.

An isotoxic planning comparison study for stage II-III non-small cell lung cancer: is intensity-modulated radiotherapy the answer?

AIMS: Recent clinical series suggest that treating patients with isotoxic twice-daily radiotherapy may be beneficial. This dosimetric planning study compared the use of intensity-modulated radiotherapy (IMRT) and three-dimensional conformal radiotherapy (3DRT) to deliver isotoxic treatment for non-small cell lung cancer (NSCLC) patients. MATERIALS AND METHODS: Twenty patients with stage II/III NSCLC were selected. A dose-escalated plan was produced retrospectively for each using three different methods: (i) three to five beams 3DRT; (ii) seven beams inverse-planned conformal radiotherapy; (iii) seven beams IMRT. The starting point for dose escalation was 55.8 Gy in 1.8 Gy per fraction twice-daily. The number of fractions was then increased until one or more organ at risk tolerance dose was exceeded or a maximum dose of 79.2 Gy was reached. RESULTS: The median escalated doses were 70.2, 66.6 and 64.8 Gy for IMRT, 3DRT and inverse-planned conformal radiotherapy, respectively. IMRT allowed a significant dose increase in comparison with the other two methods (P < 0.05), whereas no significant difference was found between 3DRT and inverse-planned conformal radiotherapy. IMRT was more successful at escalating dose in patients where the brachial plexus and spinal canal were close to the planning target volume. IMRT did not allow the escalation of dose beyond 70.2 Gy (82.8 Gy BED10, 69 Gy EQD2) due to the proximity of disease to the great vessels and the proximal bronchial tree. CONCLUSIONS: IMRT allows increased dose escalation compared with conformal radiotherapy. However, there is limited opportunity to escalate the prescription dose beyond 70.2 Gy twice-daily in disease close to the central mediastinal structures.

Marvin: an anatomical phantom for dosimetric evaluation of complex radiotherapy of the head and neck.

We report on the design of Marvin, a Model Anatomy for Radiotherapy Verification and audit In the head and Neck and present results demonstrating its use in the development of the Elekta volumetric modulated arc therapy (VMAT) technique at the Christie, and in the audit of TomoTherapy and Varian RapidArc at other institutions. The geometry of Marvin was generated from CT datasets of eight male and female patients lying in the treatment position, with removable inhomogeneities modelling the sinuses and mandible. A modular system allows the phantom to be used with a range of detectors, with the locations of the modules being based on an analysis of a range of typical treatment plans (27 in total) which were mapped onto the phantom geometry. Results demonstrate the use of Gafchromic EBT2/EBT3 film for measurement of relative dose in a plane through the target and organs-at-risk, and the use of a small-volume ionization chamber for measurement of absolute dose in the target and spinal cord. Measurements made during the development of the head and neck VMAT protocol at the Christie quantified the improvement in plan delivery resulting from the installation of the Elekta Integrity upgrade (which permits an effectively continuously variable dose rate), with plans delivered before and after the upgrade having 88.5 ± 9.4% and 98.0 ± 2.2% respectively of points passing a gamma analysis (at 4%, 4 mm, global). Audits of TomoTherapy and Varian RapidArc neck techniques at other institutions showed a similar quality of plan delivery as for post-Integrity Elekta VMAT: film measurements for both techniques had >99% of points passing a gamma analysis at the clinical criteria of 4%, 4 mm, global, and >95% of points passing at tighter criteria of 3%, 3 mm, global; and absolute dose measurements in the PTV and spinal cord were within 1.5% and 3.5% of the planned doses respectively for both techniques. The results demonstrate that Marvin is an efficient and effective means of assessing the quality of delivery of complex radiotherapy in the head and neck, and is a useful tool to assist development and audit of these techniques.

Modelling the throughput capacity of a single-accelerator multitreatment room proton therapy centre.

OBJECTIVE: We describe a model for evaluating the throughput capacity of a single-accelerator multitreatment room proton therapy centre with the aims of (1) providing quantitative estimates of the throughput and waiting times and (2) providing insight into the sensitivity of the system to various physical parameters. METHODS: A Monte Carlo approach was used to compute various statistics about the modelled centre, including the throughput capacity, fraction times for different groups of patients and beam waiting times. A method of quantifying the saturation level is also demonstrated. RESULTS: Benchmarking against the MD Anderson Cancer Center showed good agreement between the modelled (140 ± 4 fractions per day) and reported (133 ± 35 fractions per day) throughputs. A sensitivity analysis of that system studied the impact of beam switch time, the number of treatment rooms, patient set-up times and the potential benefit of having a second accelerator. Finally, scenarios relevant to a potential UK facility were studied, finding that a centre with the same four-room, single-accelerator configuration as the MD Anderson Cancer Center but handling a more complex UK-type caseload would have a throughput reduced by approximately 19%, but still be capable of treating in excess of 100 fractions per 16-h treatment day. CONCLUSIONS: The model provides a useful tool to aid in understanding the operating dynamics of a proton therapy facility, and for investigating potential scenarios for prospective centres. ADVANCES IN KNOWLEDGE: The model helps to identify which technical specifications should be targeted for future improvements.

A megavoltage scatter correction technique for cone-beam CT images acquired during VMAT delivery.

Kilovoltage cone-beam CT (kV CBCT) can be acquired during the delivery of volumetric modulated arc therapy (VMAT), in order to obtain an image of the patient during treatment. However, the quality of such CBCTs is degraded by megavoltage (MV) scatter from the treatment beam onto the imaging panel. The objective of this paper is to introduce a novel MV scatter correction method for simultaneous CBCT during VMAT, and to investigate its effectiveness when compared to other techniques. The correction requires the acquisition of a separate set of images taken during VMAT delivery, while the kV beam is off. These images--which contain only the MV scatter contribution on the imaging panel--are then used to correct the corresponding kV/MV projections. To test this method, CBCTs were taken of an image quality phantom during VMAT delivery and measurements of contrast to noise ratio were made. Additionally, the correction was applied to the datasets of three VMAT prostate patients, who also received simultaneous CBCTs. The clinical image quality was assessed using a validated scoring system, comparing standard CBCTs to the uncorrected simultaneous CBCTs and a variety of correction methods. Results show that the correction is able to recover some of the low and high-contrast signal to noise ratio lost due to MV scatter. From the patient study, the corrected CBCT scored significantly higher than the uncorrected images in terms of the ability to identify the boundary between the prostate and surrounding soft tissue. In summary, a simple MV scatter correction method has been developed and, using both phantom and patient data, is shown to improve the image quality of simultaneous CBCTs taken during VMAT delivery.

The use of a realistic VMAT delivery emulator to optimize dynamic machine parameters for improved treatment efficiency.

The delivery of volumetric modulated arc therapy (VMAT) requires the simultaneous movement of the linear accelerator gantry, multi-leaf collimators and jaws while the dose rate is varied. In this study, a VMAT delivery emulator was developed to accurately predict the characteristics of a given treatment plan, incorporating realistic parameters for gantry inertia and the variation in leaf speed with respect to gravity. The emulator was used to assess the impact of dynamic machine parameters on the delivery efficiency, using a set of prostate and head and neck VMAT plans. Initially, assuming a VMAT system with fixed dose rate bins, the allowable leaf and jaw speeds were increased and a significant improvement in treatment time and average dose rate was observed. The software was then adapted to simulate a VMAT system with continuously varying dose rate, and the increase in delivery efficiency was quantified, along with the impact of an increased leaf and jaw speed. Finally, a set of optimal dynamic machine parameters was derived assuming an idealized scenario in which the treatment is delivered in a single arc at constant maximum gantry speed.

A VMAT planning solution for prostate patients using a commercial treatment planning system.

Volumetric modulated arc therapy (VMAT) is a rotational delivery technique which offers the potential of improved dose distributions and shorter treatment times when compared to fixed-beam intensity-modulated radiation therapy (IMRT). This note describes the use of an existing treatment planning system (Philips Pinnacle(3) v.8.0), supplemented by in-house software, to produce a single-arc VMAT prostate plan. While a number of planning systems for the Elekta VMAT platform are commercially available, the use of an in-house solution has allowed more detailed investigations of VMAT planning, as well as greater control over the optimization process. The solution presented here begins with a static step-and-shoot IMRT approach to provide initial segment shapes, which are then modified and sequenced into 60 equally spaced control points in a 360 degrees arc. Dose-volume histogram comparisons demonstrate that this VMAT planning method offers multiple dose level target coverage comparable to that from a standard IMRT approach. The VMAT plans also show superior sparing of critical structures such as the rectum and bladder. Delivery times are reduced with the VMAT method, and the results of dosimetric verification, resilience and repeatability tests indicate that the solution is robust.

A novel imaging technique for fusion of high-quality immobilised MR images of the head and neck with CT scans for radiotherapy target delineation.

Uncertainty and inconsistency are observed in target volume delineation in the head and neck for radiotherapy treatment planning based only on CT imaging. Alternative modalities such as MRI have previously been incorporated into the delineation process to provide additional anatomical information. This work aims to improve on previous studies by combining good image quality with precise patient immobilisation in order to maintain patient position between scans. MR images were acquired using quadrature coils placed over the head and neck while the patient was immobilised in the treatment position using a five-point thermoplastic shell. The MR image and CT images were automatically fused in the Pinnacle treatment planning system using Syntegra software. Image quality, distortion and accuracy of the image registration using patient anatomy were evaluated. Image quality was found to be superior to that acquired using the body coil, while distortion was < 1.0 mm to a radius of 8.7 cm from the scan centre. Image registration accuracy was found to be 2.2 mm (+/- 0.9 mm) and < 3.0 degrees (n = 6). A novel MRI technique that combines good image quality with patient immobilization has been developed and is now in clinical use. The scan duration of approximately 15 min has been well tolerated by all patients.

Shading correction algorithm for improvement of cone-beam CT images in radiotherapy.

Cone-beam CT (CBCT) images have recently become an established modality for treatment verification in radiotherapy. However, identification of soft-tissue structures and the calculation of dose distributions based on CBCT images is often obstructed by image artefacts and poor consistency of density calibration. A robust method for voxel-by-voxel enhancement of CBCT images using a priori knowledge from the planning CT scan has been developed and implemented. CBCT scans were enhanced using a low spatial frequency grey scale shading function generated with the aid of a planning CT scan from the same patient. This circumvents the need for exact correspondence between CBCT and CT and the process is robust to the appearance of unshared features such as gas pockets. Enhancement was validated using patient CBCT images. CT numbers in regions of fat and muscle tissue in the processed CBCT were both within 1% of the values in the planning CT, as opposed to 10-20% different for the original CBCT. Visual assessment of processed CBCT images showed improvement in soft-tissue visibility, although some cases of artefact introduction were observed.

Intensity-modulated radiotherapy planning from limited anatomical information: is Sim-CT sufficient for planning women with breast cancer receiving intensity-modulated radiotherapy?

AIMS: To investigate intensity-modulated radiotherapy (IMRT) plans for women with carcinoma of the breast, using a small number of Sim-CT slices, thus avoiding changing the patient's position and potential problems with CT capacity. MATERIALS AND METHODS: Ten CT scans of women with breast cancer were obtained for use in the study. IMRT plans based on an open tangent pair and additional top-up segment fields were created using the full CT scan, and represented the gold standard treatment plan for comparison purposes. Five-slice CT simulator scans were artificially created by omitting intermediate slices from the full CT scans. Additionally, the intermediate CT slices were recreated via interpolation of the five slices using a standard interpolation algorithm. IMRT plans were created in the same way as for the full CT scans. To allow a suitable plan comparison to be made, the beam segments and monitor units were transferred to the full CT scans, and the dose distribution calculated. RESULTS: The interpolated five-slice plans showed no significant difference in the volume of tissue receiving dose outside the range 95-105%, compared with the IMRT plans created using the full CT data set (1.3 +/- 2.2%, P = 0.092). In contrast, the discrete slice CT simulator plans increased by 6.3 +/- 5.4%, P = 0.0054, showing a statistically significant difference in the dose distribution produced and a clinically inferior plan. CONCLUSIONS: Plans created using five discrete slice CT scans were inferior to full CT-derived IMRT treatment plans, and are therefore not acceptable for IMRT. However, interpolating five CT simulator slices provides adequate anatomical information to produce comparable IMRT plans to those created by full CT scans of the patient. This allows the introduction of IMRT for this patient group without the need to change treatment position to accommodate CT scanning.

Generating compensation designs for tangential breast irradiation with artificial neural networks.

In this paper we discuss a study comparing an algorithm implemented clinically to design intensity-modulated fields with two artificial neural networks (ANNs) trained to design the same fields. The purpose of the algorithm is to produce compensation for tangential breast radiotherapy in order to improve dose homogeneity. This was achieved by creating intensity-modulated fields to supplement standard wedged fields. Portal image data were used to create thickness maps of the medial and lateral fields, which in turn were used to design the wedged and intensity-modulated fields. The ANNs were developed to design the intensity-modulated fields from the portal image data and corresponding fluence map alone. One used localized groups of portal image pixels related to the fluence map (method 2), and the other used a one-to-one mapping between spatially corresponding pixels (method 3). A dosimetric comparison of the methods was performed by calculating the overall dose distribution. The volume of tissue outside the dose range 95-105% was used to assess dose homogeneity. The average volume outside 95-105%, averaged over 80 cases, was shown to be 2.3% for the algorithm, whilst average values of 9.9% and 13.5% were obtained for methods 2 and 3, respectively. The results of this study demonstrate the ability of an ANN to learn the general shape of compensation required and explore the use of image-based ANNs in the design of intensity-modulated fields.

Simultaneous optimization of beam orientations and beam weights in conformal radiotherapy.

A methodology for the concurrent optimization of beam orientations and beam weights in conformal radiotherapy treatment planning has been developed and tested on a cohort of five patients. The algorithm is based on a beam-weight optimization scheme with a downhill simplex optimization engine. The use of random voxels in the dose calculation provides much of the required speed up in the optimization process, and allows the simultaneous optimization of beam orientations and beam weights in a reasonable time. In the implementation of the beam-weight optimization algorithm just 10% of the original patient voxels are used for the dose calculation and cost function evaluation. A fast simulated annealing algorithm controls the optimization of the beam arrangement. The optimization algorithm was able to produce clinically acceptable plans for the five patients in the cohort study. The algorithm equalized the dose to the optic nerves compared to the standard plans and reduced the mean dose to the brain stem by an average of 4.4% (+/- 1.9, 1 SD), p value = 0.007. The dose distribution to the PTV was not compromised by developing beam arrangements via the optimization algorithm. In conclusion, the simultaneous optimization of beam orientations and beam weights has been developed to be routinely used in a realistic time. The results of optimization in a small cohort study show that the optimization can reliably produce clinically acceptable dose distributions and may be able to improve dose distributions compared to those from a human planner.

Optimisation of radiotherapy for carcinoma of the parotid gland: a comparison of conventional, three-dimensional conformal, and intensity-modulated techniques.

BACKGROUND AND PURPOSE: To compare external beam radiotherapy techniques for parotid gland tumours using conventional radiotherapy (RT), three-dimensional conformal radiotherapy (3DCRT), and intensity-modulated radiotherapy (IMRT). To optimise the IMRT techniques, and to produce an IMRT class solution. MATERIALS AND METHODS: The planning target volume (PTV), contra-lateral parotid gland, oral cavity, brain-stem, brain and cochlea were outlined on CT planning scans of six patients with parotid gland tumours. Optimised conventional RT and 3DCRT plans were created and compared with inverse-planned IMRT dose distributions using dose-volume histograms. The aim was to reduce the radiation dose to organs at risk and improve the PTV dose distribution. A beam-direction optimisation algorithm was used to improve the dose distribution of the IMRT plans, and a class solution for parotid gland IMRT was investigated. RESULTS: 3DCRT plans produced an equivalent PTV irradiation and reduced the dose to the cochlea, oral cavity, brain, and other normal tissues compared with conventional RT. IMRT further reduced the radiation dose to the cochlea and oral cavity compared with 3DCRT. For nine- and seven-field IMRT techniques, there was an increase in low-dose radiation to non-target tissue and the contra-lateral parotid gland. IMRT plans produced using three to five optimised intensity-modulated beam directions maintained the advantages of the more complex IMRT plans, and reduced the contra-lateral parotid gland dose to acceptable levels. Three- and four-field non-coplanar beam arrangements increased the volume of brain irradiated, and increased PTV dose inhomogeneity. A four-field class solution consisting of paired ipsilateral coplanar anterior and posterior oblique beams (15, 45, 145 and 170 degrees from the anterior plane) was developed which maintained the benefits without the complexity of individual patient optimisation. CONCLUSIONS: For patients with parotid gland tumours, reduction in the radiation dose to critical normal tissues was demonstrated with 3DCRT compared with conventional RT. IMRT produced a further reduction in the dose to the cochlea and oral cavity. With nine and seven fields, the dose to the contra-lateral parotid gland was increased, but this was avoided by optimisation of the beam directions. The benefits of IMRT were maintained with three or four fields when the beam angles were optimised, but were also achieved using a four-field class solution. Clinical trials are required to confirm the clinical benefits of these improved dose distributions.

Improvements in target coverage and reduced spinal cord irradiation using intensity-modulated radiotherapy (IMRT) in patients with carcinoma of the thyroid gland.

BACKGROUND AND PURPOSE: External beam radiotherapy for thyroid carcinoma poses a significant technical challenge as the target volume lies close to or surrounds the spinal cord. The potential of intensity-modulated radiotherapy (IMRT) to improve the dose distributions was investigated. MATERIALS AND METHODS: A planning study was performed on patients with thyroid carcinoma. Plans were generated to irradiate the thyroid bed alone or to treat the thyroid bed and the loco-regional lymph nodes in two phases. Conventional plans with minimal beam shaping were compared to three-dimensional conformal radiotherapy (3DCRT) and inverse-planned IMRT plans to assess target coverage and normal tissue sparing. IMRT techniques were optimized to find the minimum number of equispaced beams required to achieve the clinical benefit and a concomitant boost technique was explored. RESULTS: For the thyroid bed alone and the thyroid bed plus loco-regional lymph nodes, conventional and conformal techniques produced low minimum doses to the planning target volume (PTV) if spinal cord tolerance was respected. 3DCRT reduced the irradiated volume of normal tissue (P=0.01). IMRT plans achieved the goal dose to the PTV (P<0.01) and also reduced the spinal cord maximum dose (P<0.01). IMRT, using a concomitant boost technique, produced better target coverage than a two-phase technique. For both the two-phase and concomitant boost techniques, IMRT plans with seven and five equispaced fields produced similar dose distributions to nine fields, but three fields were significantly worse. CONCLUSIONS: 3DCRT reduced normal tissue irradiation compared to conventional techniques, but did not improve PTV or spinal cord doses. IMRT improved the PTV coverage and reduced the spinal cord dose. A simultaneous integrated boost technique with five equispaced fields produced the best dose distribution. IMRT should reduce the risk of myelopathy or may allow dose escalation in patients with thyroid cancer.

Mutagenicity of electrophilic N-acyloxy-N-alkoxyamides.

N-acyloxy-N-alkoxybenzamides are mutagenic in TA100 without the need for metabolic activation with S9. Electronic effects of substituents on both the benzamide ring in N-acetoxy-N-butoxybenzamides or the benzyloxy ring in N-acetoxy-N-benzyloxybenzamides do not influence mutagenicity levels. For N-benzoyloxy-N-benzyloxybenzamides, mutagenicity levels are inversely related to the electron-withdrawing effect of substituents on the benzoyloxy leaving group. Since reactivities increase with increasing electron-withdrawing effects, mutagenicity correlates with stability rather than reactivity of these mutagens. Hydrophobicity is the dominant factor controlling mutagenicity levels and data for all mutagens correlate with computed logP values with a lower dependence (h=0.22) than that recorded for indirect mutagens (h=1.0), except where a sterically demanding p-tert-butyl substituent or a naphthyl group is present. N-acetoxy-N-butoxynaphthamide exhibits a much higher level of mutagenicity than predicted by its logP value and activity may be ascribed to an intercalative binding process with DNA rather than straightforward hydrophobic binding in the major or minor groove. Since these are direct-acting mutagens, structural factors influence binding and reactivity towards DNA.

Beam-orientation optimization of intensity-modulated radiotherapy: clinical application to parotid gland tumours.

BACKGROUND AND PURPOSE: An optimization algorithm has been developed to determine the best beam-arrangement for a small number of intensity-modulated radiotherapy (IMRT) fields. The algorithm is designed to avoid, if possible, beam-orientations that pass through organs-at-risk (OARs) with low radiation tolerance. MATERIALS AND METHODS: An independent, fast IMRT algorithm based on the Bortfeld algorithm was developed to determine the profile of the intensity-modulated beams (IMBs) for each beam-arrangement and a fast-simulated-annealing algorithm finds the 'optimal' beam-arrangement. The final beam-arrangement was transferred to the CORVUS (NOMOS Corporation) treatment planning system, and the IMBs were re-optimized for comparison with a standard nine-field, equi-spaced arrangement. The algorithm has been initially tested on a single example patient, with a parotid gland carcinoma. RESULTS: The nine-field, IMRT plan for an example patient with a parotid gland tumour significantly reduced the dose to the cochlea compared with the conformal radiotherapy plan. In addition, the planning-target-volume (PTV) homogeneity was improved, but the plan produced a higher dose to the contralateral parotid (73% of the OAR received more than 6 Gy). The beam-orientation optimization algorithm produced a three-field plan that greatly reduced the dose to the contralateral parotid (maximum dose of 2 Gy), whilst maintaining the PTV dose homogeneity and the reduced cochlear dose of the nine-field plan. Some changes in the dose to the other OARs, namely the brain and the oral cavity, were seen, but were deemed not to be clinically significant. CONCLUSIONS: In conclusion, IMB-orientation optimization for head and neck treatment sites can produce improvements in treatment plans with only a few fields.

Reduction of small and large bowel irradiation using an optimized intensity-modulated pelvic radiotherapy technique in patients with prostate cancer.

PURPOSE: To investigate the role of intensity-modulated radiation therapy (IMRT) to irradiate the prostate gland and pelvic lymph nodes while sparing critical pelvic organs, and to optimize the number of beams required. METHODS AND MATERIALS: Target, small bowel, colon, rectum, and bladder were outlined on CT planning scans of 10 men with prostate cancer. Optimized conventional (RT) and 3-dimensional conformal radiotherapy (3D-CRT) plans were created and compared to inverse-planned IMRT dose distributions using dose-volume histograms. Optimization of beam number was undertaken for the IMRT plans. RESULTS: With RT the mean percentage volume of small bowel and colon receiving >45 Gy was 21.4 +/- 5.4%. For 3D-CRT it was 18.3 +/- 7.7% (p = 0.0043) and for 9-field IMRT it was 5.3 +/- 1.8% (p < 0.001 compared to 3D-CRT). For 7, 5, and 3 IMRT fields, it was 6.4 +/- 2.9%, 7.2 +/- 2.8%, and 8.4 +/- 3.8% (all p < 0.001 compared to 3D-CRT). The rectal volume irradiated >45 Gy was reduced from 50.5 +/- 16.3% (3D-CRT) to 5.8 +/- 2.1% by 9-field IMRT (p < 0. 001) and bladder from 52.2 +/- 12.8% to 7 +/- 2.8% (p < 0.001). Similar benefits were maintained for 7, 5, and 3 IMRT fields. CONCLUSIONS: The reduction in critical pelvic organ irradiation seen with IMRT may reduce side effects in patients, and allow modest dose escalation within acceptable complication rates. These reductions were maintained with 3-5 IMRT field plans which potentially allow less complex delivery techniques and shorter delivery times.

Beam-orientation customization using an artificial neural network.

A methodology for the constrained customization of coplanar beam orientations in radiotherapy treatment planning using an artificial neural network (ANN) has been developed. The geometry of the patients, with cancer of the prostate, was modelled by reducing the external contour, planning target volume (PTV) and organs at risk (OARs) to a set of cuboids. The coordinates and size of the cuboids were given to the ANN as inputs. A previously developed beam-orientation constrained-customization (BOCC) scheme employing a conventional computer algorithm was used to determine the customized beam orientations in a training set containing 45 patient datasets. Twelve patient datasets not involved in the training of the artificial neural network were used to test whether the ANN was able to map the inputs to customized beam orientations. Improvements from the customized beam orientations were compared with standard treatment plans with fixed gantry angles and plans produced from the BOCC scheme. The ANN produced customized beam orientations within 5 degrees of the BOCC scheme in 62.5% of cases. The average difference in the beam orientations produced by the ANN and the BOCC scheme was 7.7 degrees (+/-1.7, 1 SD). Compared with the standard treatment plans, the BOCC scheme produced plans with an increase in the average tumour control probability (TCP) of 5.7% (+/-1.4, 1 SD) whilst the ANN generated plans increased the average TCP by 3.9% (+/-1.3, 1 SD). Both figures refer to the TCP at a fixed rectal normal tissue complication probability (NTCP) of 1%. In conclusion, even using a very simple model for the geometry of the patient, an ANN was able to produce beam orientations that were similar to those produced by a conventional computer algorithm.

Constrained customization of non-coplanar beam orientations in radiotherapy of brain tumours.

A methodology for the constrained customization of non-coplanar beam orientations in radiotherapy treatment planning has been developed and tested on a cohort of five patients with tumours of the brain. The methodology employed a combination of single and multibeam cost functions to produce customized beam orientations. The single-beam cost function was used to reduce the search space for the multibeam cost function, which was minimized using a fast simulated annealing algorithm. The scheme aims to produce well-spaced, customized beam orientations for each patient that produce low dose to organs at risk (OARs). The customized plans were compared with standard plans containing the number and orientation of beams chosen by a human planner. The beam orientation constraint-customized plans employed the same number of treatment beams as the standard plan but with beam orientations chosen by the constrained-customization scheme. Improvements from beam orientation constraint-customization were studied in isolation by customizing the beam weights of both plans using a dose-based downhill simplex algorithm. The results show that beam orientation constraint-customization reduced the maximum dose to the orbits by an average of 18.8 (+/-3.8, ISD)% and to the optic nerves by 11.4 (+/-4.8, ISD)% with no degradation of the planning target volume (PTV) dose distribution. The mean doses, averaged over the patient cohort, were reduced by 4.2 (+/-1.1, ISD)% and 12.4 (+/-3.1, ISD)% for the orbits and optic nerves respectively. In conclusion, the beam orientation constraint-customization can reduce the dose to OARs, for few-beam treatment plans, when compared with standard treatment plans developed by a human planner.