PURPOSE:This study explored Continuous Positive Airway Pressure (CPAP) as an alternative to breath-holding (DIBH) techniques by comparing it with free-breathing (FB) in mediastinal lymphoma patients treated with optimized volumetric arc therapy (VMAT). METHODS:Patients underwent computed tomography (CT) simulation in both FB and CPAP (18 cm H2O). The amount of residual respiratory motion with CPAP was assessed using 4D-CT in 27 patients. Lungs, breasts, heart, and cardiac structures were contoured and included in VMAT optimization. Maximum and mean doses were compared for all organs, and CPAP was used if beneficial. Risks of coronary artery disease (CAD) and chronic heart failure (CHF) were estimated via excess relative risk (ERR). RESULTS:The study included 58 patients (22 males, 36 females; mean age 29 years) with Hodgkin (45) or large B-cell lymphoma (13). The median prescribed dose was 30 Gy (range 30-40 Gy). CPAP was well tolerated by all patients. Only nine patients (16%) had no dosimetric benefit from CPAP and were treated in FB. CPAP increased mean lung volume by 1.6 L (4460 vs 2810 cc, p < 0.01) minimized respiratory motion in patients assessed with 4D-CT (mean residual excursion: 388 cc), and significantly reduced lung V20 (10% vs 12.7%, p < 0.01) and V5 (38.2% vs 42.7%, p < 0.01). Moreover, CPAP decreased the intersection of the whole heart with planning target volume (27.9 to 20.9 cc, p < 0.01), which resulted in lower mean doses to the heart (5.3 vs 6.1 Gy, p < 0.01), coronary arteries (7.6 vs 9.5 Gy, p < 0.01), and left ventricle (2.8 vs 3.5 Gy, p < 0.01). CAD and CHF risks significantly reduced with CPAP (p < 0.01). CONCLUSIONS:CPAP significantly reduced radiation doses to lungs and heart, was well tolerated and easy to implement and to use also in department with high patients workload, representing a valid and reproducible alternative to DIBH in mediastinal lymphoma radiotherapy.

Background/Objectives: Total body irradiation (TBI) is widely used in conditioning regimens before hematopoietic stem cell transplantation. In conventional opposed-field TBI, monitor unit (MU) calculation and lung shielding definition are often based on manual procedures that may introduce operator-dependent variability. This study aimed to develop machine learning (ML) models to support the prediction of these treatment parameters using routinely available clinical and imaging data. Methods: A retrospective analysis was performed on 80 patients treated with conventional opposed-field TBI. Clinical, geometric, and CT-derived variables were used to train regression models for MU prediction. Feature selection was performed using LASSO regression, followed by Ridge regression for final modeling. Lung shielding thickness prediction was developed using planning CT data from 66 patients through recursive feature elimination and Random Forest regression. Model performance was assessed using nested 5-fold cross-validation and mean absolute error (MAE). Results: The final Ridge model achieved an MAE of 74.0 ± 6.9 MU, improving compared with the full-feature model 115.6 ± 44.0 MU. The Random Forest benchmark achieved an MAE of 81.1 ± 10.3 MU. For lung shielding thickness prediction (6-9 mm), the Random Forest model achieved an MAE of 0.60 mm. Prediction uncertainties were consistent with clinically accepted in vivo dosimetric tolerances. Conclusions: ML-based models can support the estimation of key TBI treatment parameters, potentially improving workflow efficiency and reducing operator-dependent variability while complementing standard treatment planning and verification procedures.

Nuclear medicine has acquired a crucial role in the management of patients with neuroendocrine neoplasms (NENs) by improving the accuracy of diagnosis and staging as well as their risk stratification and personalized therapies, including radioligand therapies (RLT). Artificial intelligence (AI) and radiomics can enable physicians to further improve the overall efficiency and accuracy of the use of these tools in both diagnostic and therapeutic settings by improving the prediction of the tumor grade, differential diagnosis from other malignancies, assessment of tumor behavior and aggressiveness, and prediction of treatment response. This systematic review aims to describe the state-of-the-art AI and radiomics applications in the molecular imaging of NENs.

Radiotherapy accelerators have undergone continuous technological developments. We investigated the differences between Radixact™ and VMAT treatment plans. Sixty patients were included in this study. Dosimetric comparison between the Radixact™ and VMAT plans was performed for six cancer sites: whole-brain, head and neck, lymphoma, lung, prostate, and rectum. The VMAT plans were generated with two Elekta linear accelerators (Synergy® and Versa HD™). The planning target volume (PTV) coverage, organs-at-risk dose constraints, and four dosimetric indexes were considered. The deliverability of the plans was assessed using quality assurance (gamma index evaluation) measurements; clinical judgment was included in the assessment. The mean AAPM TG218 (3

In neuroendocrine neoplasms (NENs), the use of new radiopharmaceuticals has improved the accuracy of diagnosis and staging, refined surveillance approaches, and introduced specific and personalized radiation therapies. Nuclear medicine has therefore acquired a crucial role in the management of NENs patients by improving their risk stratification and personalized therapies. Artificial intelligence (AI) and radiomics can enable physicians to further improve the overall efficiency and accuracy of the use of these tools in both di-agnostic and therapeutic settings by improving the prediction of tumor grade, differential diagnosis from other malignancies, assessment of tumor behavior and aggressiveness, and prediction of treatment response. This systematic review aims to describe the state-of-the-art on AI and radiomics applications in molecular imaging of NENs.

Purpose Different methods are available to identify haematopoietically active bone marrow (ActBM). However, their use can be challenging for radiotherapy routine treatments, since they require specific equipment and dedicated time. A machine learning (ML) approach, based on radiomic features as inputs to three different classifiers, was applied to computed tomography (CT) images to identify haematopoietically active bone marrow in anal cancer patients. Methods A total of 40 patients was assigned to the construction set (training set + test set). Fluorine-18-Fluorodeoxyglucose Positron Emission Tomography (18FDG-PET) images were used to detect the active part of the pelvic bone marrow (ActPBM) and stored as ground-truth for three subregions: iliac, lower pelvis and lumbosacral bone marrow (ActIBM, ActLPBM, ActLSBM). Three parameters were used for the correspondence analyses between 18FDG-PET and ML classifiers: DICE index, Precision and Recall. Results For the 40-patient cohort, median values [min; max] of the Dice index were 0.69 [0.20; 0.84], 0.76 [0.25; 0.89], and 0.36 [0.15; 0.67] for ActIBM, ActLSBM, and ActLPBM, respectively. The Precision/Recall (P/R) ratio median value for the ActLPBM structure was 0.59 [0.20; 1.84] (over segmentation), while for the other two subregions the P/R ratio median has values of 1.249 [0.43; 4.15] for ActIBM and 1.093 [0.24; 1.91] for ActLSBM (under segmentation). Conclusion A satisfactory degree of overlap compared to 18FDG-PET was found for 2 out of the 3 subregions within pelvic bones. Further optimization and generalization of the process is required before clinical implementation.