Improving the efficiency of breast radiotherapy treatment planning using a semi-automated approach

Clinical implementation and evaluation of deep learning-assisted automatic radiotherapy treatment planning for lung cancer

PURPOSE

The purpose of the study is to investigate the clinical application of deep learning (DL)-assisted automatic radiotherapy planning for lung cancer.

METHODS

A DL model was developed for predicting patient-specific doses, trained and validated on a dataset of 235 patients with diverse target volumes and prescriptions. The model was integrated into clinical workflow with DL-predicted objective functions. The automatic plans were retrospectively designed for additional 50 treated manual volumetric modulated arc therapy (VMAT) plans. A comparison was made between automatic and manual plans in terms of dosimetric indexes, monitor units (MUs) and planning time. Plan quality metric (PQM) encompassing these indexes was evaluated, with higher PQM values indicating superior plan quality. Qualitative evaluations of two plans were conducted by four reviewers.

RESULTS

The PQM score was 40.7 ± 13.1 for manual plans and 40.8 ± 13.5 for automatic plans (P = 0.75). Compared to manual plans, the targets coverage and homogeneity of automatic plans demonstrated no significant difference. Manual plans exhibited better sparing for lung in V5 (difference: 1.8 ± 4.2 %, P = 0.02), whereas automatic plans showed enhanced sparing for heart in V30 (difference: 1.4 ± 4.7 %, P = 0.02) and for spinal cord in Dmax (difference: 0.7 ± 4.7 Gy, P = 0.04). The planning time and MUs of automatic plans were significantly reduced by 70.5 ± 20.0 min and 97.4 ± 82.1. Automatic plans were deemed acceptable in 88 % of the reviews (176/200).

CONCLUSIONS

The DL-assisted approach for lung cancer notably decreased planning time and MUs, while demonstrating comparable or superior quality relative to manual plans. It has the potential to provide benefit to lung cancer patients.

Automated treatment planning for whole breast irradiation with individualized tangential IMRT fields

PURPOSES

This study aimed to develop and validate algorithms for automating intensity modulated radiation therapy (IMRT) planning in breast cancer patients, with a focus on patient anatomical characteristics.

MATERIAL AND METHODS

We retrospectively selected 400 breast cancer patients without lymph node involvement for automated treatment planning. Automation was achieved using the Eclipse Scripting Application Programming Interface (ESAPI) integrated into the Eclipse Treatment Planning System. We employed three beam insertion geometries and three optimization strategies, resulting in 3600 plans, each delivering a 40.05 Gy dose in 15 fractions. Gantry angles in the tangent fields were selected based on a criterion involving the minimum intersection area between the Planning Target Volume (PTV) and the ipsilateral lung in the Beam's Eye View projection. ESAPI was also used to gather patient anatomical data, serving as input for Random Forest models to select the optimal plan. The Random Forest classification considered both beam insertion geometry and optimization strategy. Dosimetric data were evaluated in accordance with the Radiation Therapy Oncology Group (RTOG) 1005 protocol.

RESULTS

Overall, all approaches generated high-quality plans, with approximately 94% meeting the acceptable dose criteria for organs at risk and/or target coverage as defined by RTOG guidelines. Average automated plan generation time ranged from 6 min and 37 s to 9 min and 22 s, with the mean time increasing with additional fields. The Random Forest approach did not successfully enable automatic planning strategy selection. Instead, our automated planning system allows users to choose from the tested geometry and strategy options.

CONCLUSIONS

Although our attempt to correlate patient anatomical features with planning strategy using machine learning tools was unsuccessful, the resulting dosimetric outcomes proved satisfactory. Our algorithm consistently produced high-quality plans, offering significant time and efficiency advantages.

Barriers and facilitators to clinical implementation of radiotherapy treatment planning automation: A survey study of medical dosimetrists

PURPOSE

Little is known about the scale of clinical implementation of automated treatment planning techniques in the United States. In this work, we examine the barriers and facilitators to adoption of commercially available automated planning tools into the clinical workflow using a survey of medical dosimetrists.

METHODS/MATERIALS

Survey questions were developed based on a literature review of automation research and cognitive interviews of medical dosimetrists at our institution. Treatment planning automation was defined to include auto-contouring and automated treatment planning. Survey questions probed frequency of use, positive and negative perceptions, potential implementation changes, and demographic and institutional descriptive statistics. The survey sample was identified using both a LinkedIn search and referral requests sent to physics directors and senior physicists at 34 radiotherapy clinics in our state. The survey was active from August 2020 to April 2021.

RESULTS

Thirty-four responses were collected out of 59 surveys sent. Three categories of barriers to use of automation were identified. The first related to perceptions of limited accuracy and usability of the algorithms. Eighty-eight percent of respondents reported that auto-contouring inaccuracy limited its use, and 62% thought it was difficult to modify an automated plan, thus limiting its usefulness. The second barrier relates to the perception that automation increases the probability of an error reaching the patient. Third, respondents were concerned that automation will make their jobs less satisfying and less secure. Large majorities reported that they enjoyed plan optimization, would not want to lose that part of their job, and expressed explicit job security fears.

CONCLUSION

To our knowledge this is the first systematic investigation into the views of automation by medical dosimetrists. Potential barriers and facilitators to use were explicitly identified. This investigation highlights several concrete approaches that could potentially increase the translation of automation into the clinic, along with areas of needed research.

Automated hybrid volumetric modulated arc therapy (HVMAT) for whole-breast irradiation with simultaneous integrated boost to lumpectomy area : A treatment planning study

PURPOSE

To develop an automated treatment planning approach for whole breast irradiation with simultaneous integrated boost using an automated hybrid VMAT class solution (HVMAT).

MATERIALS AND METHODS

Twenty-five consecutive patients with left breast cancer received 50 Gy (2 Gy/fraction) to the whole breast and an additional simultaneous 10 Gy (2.4 Gy/fraction) to the tumor cavity. Ipsilateral lung, heart, and contralateral breast were contoured as main organs-at-risk. HVMAT plans were inversely optimized by combining two open fields with a VMAT semi-arc beam. Open fields were setup to include the whole breast with a 2 cm flash region and to carry 80% of beams weight. HVMAT plans were compared with three tangential techniques: conventional wedged-field tangential plans (SWF), field-in-field forward planned tangential plans (FiF), and hybrid-IMRT plans (HMRT). Dosimetric differences among the plans were evaluated using Kruskal-Wallis one-way analysis of variance. Dose accuracy was validated using the PTW Octavius-4D phantom together with the 1500 2D-array.

RESULTS

No significant differences were found among the four techniques for both targets coverage. HVMAT plans showed consistently better PTVs dose contrast, conformity, and homogeneity (p < 0.001 for all metrics) and statistically significant reduction of high-dose breast irradiation. V55 and V60 decreased by 30.4, 26.1, and 20.8% (p < 0.05) and 12.3, 9.9, and 6.0% (p < 0.05) for SWF, FIF, and HMRT, respectively. Pretreatment dose verification reported a gamma pass-rate greater than the acceptance threshold of 95% for all HVMAT plans. In addition, HVMAT reduced the time for full planning optimization to about 20 min.

CONCLUSIONS

HVMAT plans resulted in superior target dose conformity and homogeneity compared to other tangential techniques. Due to fast planning time HVMAT can be applied for all patients, minimizing the impact on human or departmental resources.

Assessing the performance of an automated breast treatment planning software

Purpose

To assess the dosimetric performance of an automated breast planning software.

Methods

We retrospectively reviewed 15 breast cancer patients treated with tangent fields according to the RTOG 1005 protocol and 30 patients treated off‐protocol. Planning with electronic compensators (eComps) via manual, iterative fluence editing was compared to an automated planning program called EZFluence (EZF) (Radformation, Inc.). We compared the minimum dose received by 95% of the volume (D95%), D90%, the volume receiving at least 105% of prescription (V105%), V95%, the conformity index of the V95% and PTV volumes (CI95%), and total monitor units (MUs). The PTV_Eval structure generated by EZF was compared to the RTOG 1005 breast PTV_Eval structure.

Results

The average D95% was significantly greater for the EZF plans, 95.0%, vs. the original plans 93.2% (P = 0.022). CI95% was less for the EZF plans, 1.18, than the original plans, 1.48 (P = 0.09). D90% was only slightly greater for EZF, averaging at 98.3% for EZF plans and 97.3% for the original plans (P = 0.0483). V105% (cc) was, on average, 27.8cc less in the EZF breast plans, which was significantly less than for those manually planned. The average number of MUs for the EZF plans, 453, was significantly less than original protocol plans, 500 (P = 8 × 10⁻⁶). The average difference between the protocol PTV volume and the EZF PTV volume was 196 cc, with all but two cases having a larger EZF PTV volume (P = 0.020).

Conclusion

EZF improved dose homogeneity, coverage, and MU efficiency vs. manually produced eComp plans. The EZF‐generated PTV eval is based on the volume encompassed by the tangents, and is not appropriate for dosimetric comparison to constraints for RTOG 1005 PTV eval. EZF produced dosimetrically similar or superior plans to manual, iteratively derived plans and may also offer time and efficiency benefits.

Beam complexity and monitor unit efficiency comparison in two different volumetric modulated arc therapy delivery systems using automated planning

BACKGROUND

To investigate the beam complexity and monitor unit (MU) efficiency issues for two different volumetric modulated arc therapy (VMAT) delivery technologies for patients with left-sided breast cancer (BC) and nasopharyngeal carcinoma (NPC).

METHODS

Twelve left-sided BC and seven NPC cases were enrolled in this study. Each delivered treatment plan was optimized in the Pinnacle3 treatment planning system with the Auto-Planning module for the Trilogy and Synergy systems. Similar planning dose objectives and beam configurations were used for each site in the two different delivery systems to produce clinically acceptable plans. The beam complexity was evaluated in terms of the segment area (SA), segment width (SW), leaf sequence variability (LSV), aperture area variability (AAV), and modulation complexity score (MCS) based on the multileaf collimator sequence and MU. Plan delivery and a gamma evaluation were performed using a helical diode array.

RESULTS

With similar plan quality, the average SAs for the Trilogy plans were smaller than those for the Synergy plans: 55.5 ± 21.3 cm2 vs. 66.3 ± 17.9 cm2 (p < 0.05) for the NPC cases and 100.7 ± 49.2 cm2 vs. 108.5 ± 42.7 cm2 (p < 0.05) for the BC cases, respectively. The SW was statistically significant for the two delivery systems (NPC: 6.87 ± 1.95 cm vs. 6.72 ± 2.71 cm, p < 0.05; BC: 8.84 ± 2.56 cm vs. 8.09 ± 2.63 cm, p < 0.05). The LSV was significantly smaller for Trilogy (NPC: 0.84 ± 0.033 vs. 0.86 ± 0.033, p < 0.05; BC: 0.89 ± 0.026 vs. 0.90 ± 0.26, p < 0.05). The mean AAV was significantly larger for Trilogy than for Synergy (NPC: 0.18 ± 0.064 vs. 0.14 ± 0.037, p < 0.05; BC: 0.46 ± 0.15 vs. 0.33 ± 0.13, p < 0.05). The MCS values for Trilogy were higher than those for Synergy: 0.14 ± 0.016 vs. 0.12 ± 0.017 (p < 0.05) for the NPC cases and 0.42 ± 0.106 vs. 0.30 ± 0.087 (p < 0.05) for the BC cases. Compared with the Synergy plans, the average MUs for the Trilogy plans were larger: 828.6 ± 74.1 MU and 782.9 ± 85.2 MU (p > 0.05) for the NPC cases and 444.8 ± 61.3 MU and 393.8 ± 75.3 MU (p > 0.05) for the BC cases. The gamma index agreement scores were never below 91% using 3 mm/3% (global) distance to agreement and dose difference criteria and a 10% lower dose exclusion threshold.

CONCLUSIONS

The Pinnacle3 Auto-Planning system can optimize BC and NPC plans to achieve the same plan quality using both the Trilogy and Synergy systems. We found that these two systems resulted in different SAs, SWs, LSVs, AAVs and MCSs. As a result, we suggested that the beam complexity should be considered in the development of further methodologies while optimizing VMAT autoplanning.

An Artificial Intelligence-Based Full-Process Solution for Radiotherapy: A Proof of Concept Study on Rectal Cancer

BACKGROUND AND PURPOSE

To develop an artificial intelligence-based full-process solution for rectal cancer radiotherapy.

MATERIALS AND METHODS

A full-process solution that integrates autosegmentation and automatic treatment planning was developed under a single deep-learning framework. A convolutional neural network (CNN) was used to generate segmentations of the target and the organs at risk (OAR) as well as dose distribution. A script in Pinnacle that simulates the treatment planning process was used to execute plan optimization. A total of 172 rectal cancer patients were used for model training, and 18 patients were used for model validation. Another 40 rectal cancer patients were used for an end-to-end evaluation for both autosegmentation and treatment planning. The PTV and OAR segmentation was compared with manual segmentation. The planning results was evaluated by both objective and subjective assessment.

RESULTS

The total time for full-process planning without contour modification was 7 min, and an additional 15 min may require for contour modification and re-optimization. The PTV DICE similarity coefficient was greater than 0.85 for all 40 patients in the evaluation dataset while the DICE indices of the OARs also indicated good performance. There were no significant differences between the auto plans and manual plans. The physician accepted 80% of the auto plans without any further operation.

CONCLUSION

We developed a deep learning-based automatic solution for rectal cancer treatment that can improve the efficiency of treatment planning.

Automated planning of whole breast irradiation using hybrid IMRT improves efficiency and quality

Purpose

To develop an automated workflow for whole breast irradiation treatment planning using hybrid intensity modulated radiation therapy (IMRT) approach and to demonstrate that this workflow can improve planning quality and efficiency when compared to manual planning.

Methods

The auto planning framework was built based on scripting with MIM and Pinnacle systems. MIM workflows were developed to automatically segment normal structures and targets, identify landmarks for beam placement, select beam energies, and set beam configurations. Pinnacle scripts were generated from the MIM workflow to create hybrid IMRT plans automatically. Each hybrid IMRT plan included two prescriptions: a three‐dimensional (3D) prescription consisted of two open tangent beams, and an IMRT prescription consisted of two step‐and‐shoot IMRT beams. The 3D prescription delivered a full prescription dose to the maximum dose point, and the IMRT prescription was optimized to deliver a uniform dose to the entire breast while sparing dose to the normal structures. For 30 patients, the auto plans were compared with clinically accepted manual plans using the paired sample t‐test.

Results

The auto planning process took approximately 8 min to complete. The mean dice coefficients between auto‐segmentation and manual contours were 0.98, 0.94 and 0.88 for the lungs, heart, and PTVeval_Breast, respectively. The MUs of the auto plans was on average 13% higher than that of the manual plans. Auto planning improved plan quality significantly: percentage volume receiving 95% of the prescription dose (V95%) of the PTVeval_Breast increased from 91.5% to 93.2% (P = 0.001), V105% of the PTVeval_Breast decreased from 7.2% to 1.2% (P = 0.013), V20Gy of the ipsilateral lung decreased from 13.1% to 10.4% (P = 0.001) and mean heart dose for left‐sided breast patients decreased from 1.2 Gy to 0.9 Gy (P < 0.001).

Conclusion

An automated treatment planning process can make the planning process efficient with improved plan quality.

Approach and assessment of automated stereotactic radiotherapy planning for early stage non-small-cell lung cancer

BACKGROUND

Intensity-modulated radiotherapy (IMRT) and volumetric-modulated arc therapy (VMAT) are standard physical technologies of stereotactic body radiotherapy (SBRT) that are used for patients with non-small-cell lung cancer (NSCLC). The treatment plan quality depends on the experience of the planner and is limited by planning time. An automated planning process can save time and ensure a high-quality plan. This study aimed to introduce and demonstrate an automated planning procedure for SBRT for patients with NSCLC based on machine-learning algorithms. The automated planning was conducted in two steps: (1) determining patient-specific optimized beam orientations; (2) calculating the organs at risk (OAR) dose achievable for a given patient and setting these dosimetric parameters as optimization objectives. A model was developed using data of historical expertise plans based on support vector regression. The study cohort comprised patients with NSCLC who were treated using SBRT. A training cohort (N = 125) was used to calculate the beam orientations and dosimetric parameters for the lung as functions of the geometrical feature of each case. These plan-geometry relationships were used in a validation cohort (N = 30) to automatically establish the SBRT plan. The automatically generated plans were compared with clinical plans established by an experienced planner.

RESULTS

All 30 automated plans (100%) fulfilled the dose criteria for OARs and planning target volume (PTV) coverage, and were deemed acceptable according to evaluation by experienced radiation oncologists. An automated plan increased the mean maximum dose for ribs (31.6 ± 19.9 Gy vs. 36.6 ± 18.1 Gy, P < 0.05). The minimum, maximum, and mean dose; homogeneity index; conformation index to PTV; doses to other organs; and the total monitor units showed no significant differences between manual plans established by experts and automated plans (P > 0.05). The hands-on planning time was reduced from 40-60 min to 10-15 min.

CONCLUSION

An automated planning method using machine learning was proposed for NSCLC SBRT. Validation results showed that the proposed method decreased planning time without compromising plan quality. Plans generated by this method were acceptable for clinical use.