Deep learning and atlas-based models to streamline the segmentation workflow of total marrow and lymphoid irradiation.

PURPOSE: To improve the workflow of total marrow and lymphoid irradiation (TMLI) by enhancing the delineation of organs at risk (OARs) and clinical target volume (CTV) using deep learning (DL) and atlas-based (AB) segmentation models. MATERIALS AND METHODS: Ninety-five TMLI plans optimized in our institute were analyzed. Two commercial DL software were tested for segmenting 18 OARs. An AB model for lymph node CTV (CTV_LN) delineation was built using 20 TMLI patients. The AB model was evaluated on 20 independent patients, and a semiautomatic approach was tested by correcting the automatic contours. The generated OARs and CTV_LN contours were compared to manual contours in terms of topological agreement, dose statistics, and time workload. A clinical decision tree was developed to define a specific contouring strategy for each OAR. RESULTS: The two DL models achieved a median [interquartile range] dice similarity coefficient (DSC) of 0.84 [0.71;0.93] and 0.85 [0.70;0.93] across the OARs. The absolute median Dmean difference between manual and the two DL models was 2.0 [0.7;6.6]% and 2.4 [0.9;7.1]%. The AB model achieved a median DSC of 0.70 [0.66;0.74] for CTV_LN delineation, increasing to 0.94 [0.94;0.95] after manual revision, with minimal Dmean differences. Since September 2022, our institution has implemented DL and AB models for all TMLI patients, reducing from 5 to 2 h the time required to complete the entire segmentation process. CONCLUSION: DL models can streamline the TMLI contouring process of OARs. Manual revision is still necessary for lymph node delineation using AB models.

The role of EPID in vivo dosimetry in the risk management of stereotactic lung treatments.

BACKGROUND AND OBJECTIVE: In this work we report our experience with the use of in vivo dosimetry (IVD) in the risk management of stereotactic lung treatments. METHODS: A commercial software based on the electronic portal imaging device (EPID) signal was used to reconstruct the actual planning target volume (PTV) dose of stereotactic lung treatments. The study was designed in two phases: i) in the observational phase, the IVD results of 41 consecutive patients were reviewed and out-of-tolerance cases were studied for root cause analysis; ii) in the active phase, the IVD results of 52 patients were analyzed and corrective actions were taken when needed. Moreover, proactive preventions were further introduced to reduce the risk of future failures. The error occurrence rate was analyzed to evaluate the effectiveness of proactive actions. RESULTS: A total of 330 fractions were analyzed. In the first phase, 13 errors were identified. In the active phase, 12 errors were detected, 5 of which needed corrective actions; in 4 patients the actions taken corrected the error. Several preventions and barriers were introduced to reduce the risk of future failures: the planning checklist was updated, the procedure for vacuum pillows was improved, and use of the respiratory compression belt was optimized. A decrease in the failure rate was observed, showing the effectiveness of procedural adjustment. CONCLUSION: The use of IVD allowed the quality of lung stereotactic body radiation therapy (SBRT) treatments to be improved. Patient-specific and procedural corrective actions were successfully taken as part of risk management, leading to an overall improvement in the dosimetric accuracy.

Automatic planning of the lower extremities for total marrow irradiation using volumetric modulated arc therapy.

PURPOSE: Total marrow (and lymphoid) irradiation (TMI-TMLI) is limited by the couch travel range of modern linacs, which forces the treatment delivery to be split into two plans with opposite orientations: a head-first supine upper-body plan, and a feet-first supine lower extremities plan. A specific field junction is thus needed to obtain adequate target coverage in the overlap region of the two plans. In this study, an automatic procedure was developed for field junction creation and lower extremities plan optimization. METHODS: Ten patients treated with TMI-TMLI at our institution were selected retrospectively. The planning of the lower extremities was performed automatically. Target volume parameters (CTV_J­V98% > 98%) at the junction region and several dose statistics (D98%, Dmean, and D2%) were compared between automatic and manual plans. The modulation complexity score (MCS) was used to assess plan complexity. RESULTS: The automatic procedure required 60-90 min, depending on the case. All automatic plans achieved clinically acceptable dosimetric results (CTV_J­V98% > 98%), with significant differences found at the junction region, where Dmean and D2% increased on average by 2.4% (p < 0.03) and 3.0% (p < 0.02), respectively. Similar plan complexity was observed (median MCS = 0.12). Since March 2022, the automatic procedure has been introduced in our clinic, reducing the TMI-TMLI simulation-to-delivery schedule by 2 days. CONCLUSION: The developed procedure allowed treatment planning of TMI-TMLI to be streamlined, increasing efficiency and standardization, preventing human errors, while maintaining the dosimetric plan quality and complexity of manual plans. Automated strategies can simplify the future adoption and clinical implementation of TMI-TMLI treatments in new centers.

Evaluation of plan complexity and dosimetric plan quality of total marrow and lymphoid irradiation using volumetric modulated arc therapy.

PURPOSE: To assess the impact of the planner's experience and optimization algorithm on the plan quality and complexity of total marrow and lymphoid irradiation (TMLI) delivered by means of volumetric modulated arc therapy (VMAT) over 2010-2022 at our institute. METHODS: Eighty-two consecutive TMLI plans were considered. Three complexity indices were computed to characterize the plans in terms of leaf gap size, irregularity of beam apertures, and modulation complexity. Dosimetric points of the target volume (D2%) and organs at risk (OAR) (Dmean) were automatically extracted to combine them with plan complexity and obtain a global quality score (GQS). The analysis was stratified based on the different optimization algorithms used over the years, including a knowledge-based (KB) model. Patient-specific quality assurance (QA) using Portal Dosimetry was performed retrospectively, and the gamma agreement index (GAI) was investigated in conjunction with plan complexity. RESULTS: Plan complexity significantly reduced over the years (r = -0.50, p < 0.01). Significant differences in plan complexity and plan dosimetric quality among the different algorithms were observed. Moreover, the KB model allowed to achieve significantly better dosimetric results to the OARs. The plan quality remained similar or even improved during the years and when moving to a newer algorithm, with GQS increasing from 0.019 ± 0.002 to 0.025 ± 0.003 (p < 0.01). The significant correlation between GQS and time (r = 0.33, p = 0.01) indicated that the planner's experience was relevant to improve the plan quality of TMLI plans. Significant correlations between the GAI and the complexity metrics (r = -0.71, p < 0.01) were also found. CONCLUSION: Both the planner's experience and algorithm version are crucial to achieve an optimal plan quality in TMLI plans. Thus, the impact of the optimization algorithm should be carefully evaluated when a new algorithm is introduced and in system upgrades. Knowledge-based strategies can be useful to increase standardization and improve plan quality of TMLI treatments.

Dose coverage impacts local control in ultra-central lung oligometastases treated with stereotactic radiotherapy.

INTRODUCTION: The use of Stereotactic Body Radiotherapy (SBRT) is controversial in Ultra-Central lung tumors, a subset of central lung tumors characterized by proximity to critical mediastinal structures. This is of interest in oligometastatic (≤3 metastases) patients, who can yield survival benefit from local treatments. The aim of our study is to assess the determinants of efficacy and toxicity in this setting. MATERIALS AND METHODS: Clinical and dosimetric parameters were reviewed in a cohort of oligometastatic patients treated with SBRT for ultra-central tumors. Local control rate (LC) and toxicity were assessed. Statistical Analysis was carried out to assess the impact of those predictors on local recurrence and adverse events. RESULTS: One-hundred-nine consecutive patients were included. A median Biologic Effective Dose (BED) of 105 (75-132) Gy10 was prescribed. At a median follow-up of 17 (range 3-78) months, 2-year LC was 87%. Improved LC was correlated to Planning Treatment Volume (PTV) covered by 95% of the prescription dose (V95% PTV) > 85% (HR 0.15, 95%CI 0.05-0.49, p = 0.0017) and to Gross Tumor Volume (GTV) < 90 cm3 (HR 0.2, 95%CI 0.07-0.56, p = 0.0021). Overall and grade ≥ 3 toxicity incidence was 20% and 5%, respectively. Patients experiencing acute and late toxicities received significantly higher dose to 1 cm3 (D1cm3) of esophagus and lung volume receiving ≥5 Gy (V5Gy) (p = 0.016 and p = 0.013), and higher dose to 0.1 cm3 (D0.1cm3) of heart (p = 0.036), respectively. CONCLUSION: V95% PTV > 85% and GTV < 90 cm3 are independent predictors of LC. Dose to esophagus, lung and heart should be carefully assessed to minimize treatment-related toxicities.

Total marrow and total lymphoid irradiation in bone marrow transplantation for acute leukaemia.

The use of total body irradiation as part of conditioning regimens for acute leukaemia is progressively declining because of concerns of late toxic effects and the introduction of radiation-free regimens. Total marrow irradiation and total marrow and lymphoid irradiation represent more targeted forms of radiotherapy compared with total body irradiation that have the potential to decrease toxicity and escalate the dose to the bone marrow for high-risk patients. We review the technological basis and the clinical development of total marrow irradiation and total marrow and lymphoid irradiation, highlighting both the possible advantages as well as the current roadblocks for widespread implementation among transplantation units. The exact role of total marrow irradiation or total marrow and lymphoid irradiation in new conditioning regimens seems dependent on its technological implementation, aiming to make the whole procedure less time consuming, more streamlined, and easier to integrate into the clinical workflow. We also foresee a role for computer-assisted planning, as a way to improve planning and delivery and to incorporate total marrow irradiation and total marrow and lymphoid irradiation in multi-centric phase 2-3 trials.

Role of stereotactic body radiation therapy in the treatment of liver metastases: clinical results and prognostic factors.

PURPOSE: To evaluate feasibility and efficacy of Stereotactic Body Radiation Therapy (SBRT) for unresectable liver metastasis in oligometastatic patients. METHODS: Oligometastatic patients with up to three liver metastases of a maximum diameter of 6 cm were treated with SBRT. Total dose was 75 Gy in three consecutive fractions. Study endpoints were efficacy of this fractionation in terms of local control (LC), overall survival (OS), toxicity, and prognostic factors affecting OS and LC. RESULTS: Between February 2010 and December 2016, we enrolled 202 patients, with a total of 268 unresectable liver metastases. Median follow-up time from SBRT was 33 months (5-87 months). One-, 3­, and 5­year LC rates were 92%, 84%, and 84%, respectively. In univariate analysis, the primary histology and previous local ablative therapies were significant. Median OS was 21 months and the survival rates were 79%, 27%, and 15% at 1, 3, and 5 years after SBRT, respectively. At univariate analysis, sex, primary disease histology, intra-, and extra-hepatic progression were significant prognostic factors. This analysis confirmed the absence of late toxicity >G3. CONCLUSION: This study confirms the efficacy and safety of SBRT for unresectable liver metastases. Selection of cases may improve survival and LC.

Linac-based stereotactic body radiation therapy for low and intermediate-risk prostate cancer : Long-term results and factors predictive for outcome and toxicity.

INTRODUCTION: Stereotactic body radiation therapy (SBRT) is considered an effective and safe treatment in patients with low- and intermediate-risk prostate cancer (PC). However, due to a lack of long-term follow-up and late toxicity data, this treatment is not universally accepted. The present study aimed to evaluate outcome and early and late toxicity in a cohort of patients with low- and intermediate-risk PC treated prospectively with linear accelerator (linac)-based SBRT. PATIENTS AND METHODS: Patients with low- or intermediate-risk (NCCN criteria) PC were included. All patients received linac-based SBRT to 35 Gy in 5 fractions delivered on alternate days. Endpoints were toxicity, biochemical relapse-free survival (BRFS), metastatic progression-free survival (mPFS), and overall survival (OS). RESULTS: From 2012 to 2018, 178 patients were treated. Median baseline prostate-specific antigen (iPSA) was 6.37 ng/ml (range 1.78-20). Previous transurethral resection of the prostate (TURP) was present in 23 (12.9%) patients. Median follow-up was 58.9 months (range 9.7-89.9). BRFS rates at 1, 3, and 5 years were 98.3 (95% confidence interval, CI, 94.7-99.4%), 94.4 (95%CI 89.4-97), and 91.6% (95%CI 85.4-95.2), respectively. In univariate analysis, performance status (PS), iPSA, and nadir PSA (nPSA) were correlated with BRFS. In multivariable analysis iPSA and nPSA remained significant. BRFS rates at 5 years were 94.9% (95%CI 86.8-98) for International Society of Urological Pathology (ISUP) grade group 1, 93.2% (95%CI 80.5-97.7) for ISUP group 2, and 74.8% (95%CI 47.1-89.5) for ISUP group 3. At 1, 3, and 5 years, mPFS rates were 98.8 (95%CI 95.5-99.7), 96.2 (95%CI 91.9-98.3), and 92.9% (95%CI 87.2-96.2), respectively; OS rates were 100, 97.2 (95%CI 92.9-98.9), and 95.1% (95%CI 90-97.6), respectively. One (0.56%) case of grade 3 acute genitourinary (GU), one case of acute gastrointestinal (GI), and one case of grade 3 late GU toxicity were observed. GI toxicity positively correlated with prostate volume. CONCLUSION: At long-term follow-up, linac-based SBRT continues to be a valid option for the management localized PC. Biochemical control remains high at 5 years, albeit with some concerns regarding the optimal schedule for unfavorable intermediate-risk PC. Considering the excellent prognosis, patient selection is crucial for prevention of severe late toxicity.

Role of Stereotactic Body Radiation Therapy for the Management of Oligometastatic Renal Cell Carcinoma.

PURPOSE: Kidney cancer has been increasing 1.7% annually. Renal cell carcinoma is the most common kidney cancer and it can metastasize. Our aim was to analyze patients treated with stereotactic body radiation therapy of renal cell carcinoma metastases. MATERIALS AND METHODS: A total of 58 patients (73 lesions) were treated from 2004 to 2016. Patients were candidates for analysis if a maximum of 3 metastases were diagnosed and the primary tumor was resected. Toxicity was classified according to Common Terminology Criteria for Adverse Events version 3. RESULTS: All patients had renal cell carcinoma, in particular the clear cell type in 82.7%. A total of 39 metastases (53.4%) were located in the lungs and 19 (26%) were in the lymph nodes. Less common were metastases to bone (9.5% of cases), the liver (4.1%) and the adrenal gland (6.8%). Median followup was 16.1 months (range 3.5 to 157.1). The local control rate at 12 and 18 months was 90.2% and 90.2%, respectively. The progression-free survival rate at 12 and 18 months was 46.2% (95% CI 32.2-59) and 35% (95% CI 21.4-48.9), respectively. On univariate and multivariable analyses metachronous and single metastases predicted better progression-free survival. Systemic therapy before stereotactic body radiation therapy predicted improved local control in clear cell cases. CONCLUSIONS: Stereotactic body radiation therapy can be considered a safe approach and it provides effective local control of oligometastatic renal cell carcinoma. However, future prospective studies are necessary to evaluate the impact on survival and quality of life.

Role of Stereotactic Body Radiation Therapy for the Management of Oligometastatic Renal Cell Carcinoma.

PURPOSE: Kidney cancer has been increasing 1.7% annually. Renal cell carcinoma is the most common kidney cancer and it can metastasize. Our aim was to analyze patients treated with stereotactic body radiation therapy of renal cell carcinoma metastases. MATERIALS AND METHODS: A total of 58 patients (73 lesions) were treated from 2004 to 2016. Patients were candidates for analysis if a maximum of 3 metastases were diagnosed and the primary tumor was resected. Toxicity was classified according to Common Terminology Criteria for Adverse Events version 3. RESULTS: All patients had renal cell carcinoma, in particular the clear cell type in 82.7%. A total of 39 metastases (53.4%) were located in the lungs and 19 (26%) were in the lymph nodes. Less common were metastases to bone (9.5% of cases), the liver (4.1%) and the adrenal gland (6.8%). Median followup was 16.1 months (range 3.5 to 157.1). The local control rate at 12 and 18 months was 90.2% and 90.2%, respectively. The progression-free survival rate at 12 and 18 months was 46.2% (95% CI 32.2-59) and 35% (95% CI 21.4-48.9), respectively. On univariate and multivariable analyses metachronous and single metastases predicted better progression-free survival. Systemic therapy before stereotactic body radiation therapy predicted improved local control in clear cell cases. CONCLUSIONS: Stereotactic body radiation therapy can be considered a safe approach and it provides effective local control of oligometastatic renal cell carcinoma. However, future prospective studies are necessary to evaluate the impact on survival and quality of life.

SBRT planning for spinal metastasis: indications from a large multicentric study.

BACKGROUND: The dosimetric variability in spine stereotactic body radiation therapy (SBRT) planning was investigated in a large number of centres to identify crowd knowledge-based solutions. METHODS: Two spinal cases were planned by 48 planners (38 centres). The required prescription dose (PD) was 3â€¯× 10 Gy and the planning target volume (PTV) coverage request was: VPD > 90% (minimum request: VPD > 80%). The dose constraints were: planning risk volume (PRV) spinal cord: V18Gy < 0.35 cm3, V21.9 Gy < 0.03 cm3; oesophagus: V17.7 Gy < 5 cm3, V25.2 Gy < 0.03 cm3. Planners who did not fulfil the protocol requirements were asked to re-optimize the plans, using the results of planners with the same technology. Statistical analysis was performed to assess correlations between dosimetric results and planning parameters. A quality index (QI) was defined for scoring plans. RESULTS: In all, 12.5% of plans did not meet the protocol requirements. After re-optimization, 98% of plans fulfilled the constraints, showing the positive impact of knowledge sharing. Statistical analysis showed a significant correlation (p < 0.05) between the homogeneity index (HI) and PTV coverage for both cases, while the correlation between HI and spinal cord sparing was significant only for the single dorsal PTV case. Moreover, the multileaf collimator leaf thickness correlated with the spinal cord sparing. Planners using comparable delivery/planning system techniques produced different QI, highlighting the impact of the planner's skills in the optimization process. CONCLUSION: Both the technology and the planner's skills are fundamentally important in spine SBRT planning optimization. Knowledge sharing helped to follow the plan objectives.

11C-Choline-Pet Guided Stereotactic Body Radiation Therapy for Lymph Node Metastases in Oligometastatic Prostate Cancer.

INTRODUCTION: aim is outcome of 11C-Choline-PET guided SBRT on lymph node metastases. MATERIALS AND METHODS: patients with 1 - 4 lymph node metastases detected by 11C-choline-PET were treated with SBRT. Toxicity, treated metastases control and Progression Free Survival were computed. RESULTS: twenty-six patients, 38 lymph node metastases were irradiated. No grade ≥ 2 toxicity. Median PSA-nadir after RT was 1.02 ng/mL. Post-treatment 11C-Choline-PET showed metabolic complete response in 17 metastases (44,7%), partial response in 9 metastases (38%). CONCLUSION: SBRT is effective and safe for lymph node metastases. PET is important in identification of gross tumor and evaluation of the response.

Role of Stereotactic Body Radiation Therapy with Volumetric-Modulated Arcs and High-Intensity Photon Beams for the Treatment of Abdomino-Pelvic Lymph-Node Metastases.

AIM: To study clinical outcome for oligometastatic patients (abdominal lymph-node metastases) treated with stereotactic body radiation therapy. MATERIALS AND METHODS: Seventy-one patients were studied retrospectively. Dose prescription was 45 Gy in six fractions. Clinical outcome was assessed with actuarial analysis. RESULTS: The median follow-up was 1.5 years; 45 patients (63.3%) had solitary metastasis, and 26 (36.6%) had multiple lesions. Local control was achieved in 97.5% with a 1-year actuarial rate of 83%. Two-year progression-free survival was 63.1%, and the overall survival was 76.9%. Two patients (3%) developed grade 2 gastro-enteric toxicity. CONCLUSIONS: The treatment provided adequate clinical response in the management of oligometastatic cases.

A feasibility dosimetric study on prostate cancer : are we ready for a multicenter clinical trial on SBRT?

PURPOSE: The Italian Association of Medical Physics (AIFM) started a working group dedicated to stereotactic body radiotherapy (SBRT) treatment. In this work, we performed a multicenter planning study on patients who were candidates for SBRT in the treatment of prostate cancer with the aim of evaluating the dosimetric consistency among the different hospitals. METHODS AND MATERIALS: Fourteen centers were provided the contours of 5 patients. Plans were performed following the dose prescription and constraints for organs at risk (OARs) of a reference paper. The dose prescription was 35 Gy in five fractions for the planning target volume (PTV). Different techniques were used (3D-CRT, fixed-Field IMRT, VMAT, CyberKnife). Plans were compared in terms of dose-volume histogram (DVH) parameters. Furthermore, the median DVH was calculated and one patient was re-planned. RESULTS: A total of 70 plans were compared. The maximum dose to the body was 107.9 ± 4.5 % (range 101.5-116.3 %). Dose at 98 % (D98 %) and mean dose to the clinical target volume (CTV) were 102.0 ± 0.9 % (global range 101.1-102.9 %) and 105.1 ± 0.6 % (range 98.6-124.6 %). Similar trends were found for D95 % and mean dose to the PTV. Important differences were found in terms of the homogeneity index. Doses to OARs were heterogeneous. The subgroups with the same treatment planning system showed differences comparable to the differences of the whole group. In the re-optimized plans, DVH differences among institutes were reduced and OAR sparing improved. CONCLUSION: Important dosimetric differences with possible clinical implications, in particular related to OARs, were found. Replanning allowed a reduction in the OAR dose and decreased standard deviations. Multicenter clinical trials on SBRT should require a preplanning study to standardize the optimization procedure.

Radiation therapy of anal canal cancer: from conformal therapy to volumetric modulated arc therapy.

BACKGROUND: To appraise the role of volumetric modulated arc (RapidArc, RA) in the treatment of anal canal carcinoma (ACC). METHODS: A retrospective analysis has been conducted on 36 patients treated with RA since 2009 comparing outcome against a group of 28 patients treated with conformal therapy (CRT). RA treatments were prescribed with SIB technique with 59.4 Gy to the primary tumor and nodes and 49.5 Gy to the elective nodes. CRT was sequentially delivered with 45 Gy to the pelvic target and a boost of 14.4 Gy to the primary tumor. RESULTS: Median age of patients was 65 yrs for RA (59 yrs for CRT); 90% had Stage II-III (93% in the CRT group). No statistically significant differences were observed concerning survival or control. 5 yrs disease specific survival was 85.7% and 81.2%, loco-regional control was of 78.1% and 82.1% for RA and CRT respectively. RA treatments lead to lower incidence of higher grade of toxicity events (all retrospectively retrieved from charts as worse events). Grade 2-3 toxicity, compared to CRT, reduced from 89% to 68% for GI, from 39% to 33% for GU and from 82% to 75% for the skin. Late toxicity was as follows: 5/36 (14%) and 3/36 (8%) patients had G1 or G2 GI toxicity in the RA group (1/28 (4%) and 4/28 (14%) in the CRT group). GU late toxicity was observed only in 4/28 (14%) patients of the CRT group: 3/28 (11%) had G2 and 1/28 (4%) had G1. CONCLUSIONS: RA treatments of ACC patients proved to be equally effective than CRT but it was associated to a reduction of toxicity.

Stereotactic Ablative Radiotherapy (SABR) in inoperable oligometastatic disease from colorectal cancer: a safe and effective approach.

BACKGROUND: To assess the safety and efficacy of Stereotactic Ablative Radiotherapy (SABR) in oligometastatic patients from colorectal cancer. METHODS: 82 patients with 1-3 inoperable metastases confined to one organ (liver or lung), were treated with SABR for a total of 112 lesions in an observational study. Prescription dose ranged between 48 and 75 Gy in 3 or 4 consecutive fractions. Primary end-points were local control (LC), overall survival (OS) and progression-free survival (PFS). Secondary end-point was toxicity. RESULTS: Median follow-up was 24 months (range 3-47). One, two and three years LC rate was 90%,80% and 75% (85%,75% and 70% for lung and 95%, 90% and 85% for liver metastases; no statistically significance was found). The difference in LC between the subgroup of lesions treated with ≥ 60 Gy (n = 58) and those irradiated with <60 Gy (n = 52) was statistically significant, with a 1, 2 and 3 yrs LC of 97%,92% and 83% for the higher dose, compared to 85%,70% and 70% for the lower dose (p < 0.04). Median OS was 32 months. Actuarial OS rate at 1, 2 and 3 yrs was 85%,65% and 43%. Univariate analysis showed a correlation only between OS and cumulative GTV > 3 cm (p < 0.02). Median PFS was 14 months, with a PFS rate of 56% at 1 yr and 40% at 2-3 yrs, without correlation with the site and prescription dose (p < 0.48 and p < 0.56). No patients experienced radiation-induced liver disease or grade >3 toxicity. CONCLUSIONS: SABR is a safe and feasible alternative treatment of oligometastatic colorectal liver and lung metastases in patients not amenable to surgery or other ablative treatments.

Recommendation on the technical and dosimetric data to be included in stereotactic body radiation therapy clinical trial publications based on a systematic review.

The results of phase II and III trials on Stereotactic Body Radiation Therapy (SBRT) increased adoption of SBRT worldwide. The ability to replicate clinical trial outcomes in routine practice depends on the capability to reproduce technical and dosimetric procedures used in the clinical trial. In this systematic review, we evaluated if peer-reviewed publications of clinical trials in SBRT reported sufficient technical data to ensure safe and robust implementation in real world clinics. Twenty papers were selected for inclusion, and data was extracted by a working group of medical physicists created following the ESTRO 2021 physics workshop. A large variability in technical and dosimetric data were observed, with frequent lack of required information for reproducing trial procedures. None of the evaluated studies were judged completely reproducible from a technical perspective. A list of recommendations has been provided by the group, based on the analysis and consensus process, to ensure an adequate reproducibility of technical parameters in primary SBRT clinical trials. Future publications should consider these recommendations to assist transferability of the clinical trial in real world practice.

Machine learning and lean six sigma for targeted patient-specific quality assurance of volumetric modulated arc therapy plans.

Background and purpose: Radiotherapy plans with excessive complexity exhibit higher uncertainties and worse patient-specific quality assurance (PSQA) results, while the workload of measurement-based PSQA can impact the efficiency of the radiotherapy workflow. Machine Learning (ML) and Lean Six Sigma, a process optimization method, were implemented to adopt a targeted PSQA approach, aiming to reduce workload, risk of failures, and monitor complexity. Materials and methods: Lean Six Sigma was applied using DMAIC (define, measure, analyze, improve, and control) steps. Ten complexity metrics were computed for 69,811 volumetric modulated arc therapy (VMAT) arcs from 28,612 plans delivered in our Institute (2013-2021). Outlier complexities were defined as >95th-percentile of the historical distributions, stratified by treatment. An ML model was trained to predict the gamma passing rate (GPR-3 %/1mm) of an arc given its complexity. A decision support system was developed to monitor the complexity and expected GPR. Plans at risk of PSQA failure, either extremely complex or with average GPR <90 %, were identified. The tool's impact was assessed after nine months of clinical use. Results: Among 1722 VMAT plans monitored prospectively, 29 (1.7 %) were found at risk of failure. Planners reacted by performing PSQA measurement and re-optimizing the plan. Occurrences of outlier complexities remained stable within 5 %. The expected GPR increased from a median of 97.4 % to 98.2 % (Mann-Whitney p < 0.05) due to plan re-optimization. Conclusions: ML and Lean Six Sigma have been implemented in clinical practice enabling a targeted measurement-based PSQA approach for plans at risk of failure to improve overall quality and patient safety.

Developing an ensemble machine learning study: Insights from a multi-center proof-of-concept study.

BACKGROUND: To address the numerous unmeet clinical needs, in recent years several Machine Learning models applied to medical images and clinical data have been introduced and developed. Even when they achieve encouraging results, they lack evolutionary progression, thus perpetuating their status as autonomous entities. We postulated that different algorithms which have been proposed in the literature to address the same diagnostic task, can be aggregated to enhance classification performance. We suggested a proof of concept to define an ensemble approach useful for integrating different algorithms proposed to solve the same clinical task. METHODS: The proposed approach was developed starting from a public database consisting of radiomic features extracted from CT images relating to 535 patients suffering from lung cancer. Seven algorithms were trained independently by participants in the AI4MP working group on Artificial Intelligence of the Italian Association of Physics in Medicine to discriminate metastatic from non-metastatic patients. The classification scores generated by these algorithms are used to train SVM classifier. The Explainable Artificial Intelligence approach is applied to the final model. The ensemble model was validated following an 80-20 hold-out and leave-one-out scheme on the training set. RESULTS: Compared to individual algorithms, a more accurate result was achieved. On the independent test the ensemble model achieved an accuracy of 0.78, a F1-score of 0.57 and a log-loss of 0.49. Shapley values representing the contribution of each algorithm to the final classification result of the ensemble model were calculated. This information represents an added value for the end user useful for evaluating the appropriateness of the classification result on a particular case. It also allows us to evaluate on a global level which methodological approaches of the individual algorithms are likely to have the most impact. CONCLUSION: Our proposal represents an innovative approach useful for integrating different algorithms that populate the literature and which lays the foundations for future evaluations in broader application scenarios.