Shrestha, Deepak K.

Abstract Background Radiotherapy treatment planning requires knowledge of a patient's anatomical surface for use in dose calculations, collision checks, or surface‐guided radiation therapy (SGRT). This information is usually obtained from a computed tomography (CT) scan which may encounter field‐of‐view (FOV) limitations or require additional radiation dose beyond the treatment area to capture the full patient habitus. Purpose This study demonstrates an affordable alternative solution to obtain a patient's surface information without CT or room mounted surface‐guidance imaging cameras for the purpose of detecting patient collisions and providing additional skin surface information needed for appropriate depth calculations of clipped FOV scans. Methods Three‐dimensional (3D) surface models of five patients were acquired immediately prior to CT simulation using an iPad Air 2 equipped with a Structure Sensor Pro and Monocle SS application. The 3D surfaces were manually registered to the CT images and nearest distances between CT surface and 3D surface were calculated. Nearest distance calculations were also performed with a 2 mm CT rind structure that excluded select regions not relevant to surface detection. To demonstrate the advantage of having additional scan length for clearance checks, a collision simulated plan was evaluated with a shortened CT scan length versus with additional 3D surface information. Results CT rind to 3D scan surface mean distances for each patient were 1.65 mm, 0.99 mm, −0.93 mm, −0.79 mm, and −1.53 mm with standard deviations of 6.51 mm, 6.71 mm, 3.38 mm, 2.84 mm, and 2.69 mm. Conclusion This study demonstrates the feasibility of an accessible low‐cost 3D surface scanning solution to supplement clipped field‐of‐view CT scans and reduce CT scan lengths by obtaining important patient surface data for improved calculation accuracy, treatment clearance checks, and SGRT.

This study evaluated the effectiveness of using a traditional computed tomography (CT) simulation scan in the Leksell Gamma Knife® Icon™ workflow for stereotactic radiosurgery. We compared two workflows, one involving a traditional CT simulation co-registered with the magnetic resonance imaging (MRI) and cone-beam CT (CBCT) (MRI-CT-CBCT) and another without the CT scan, co-registering the MRI directly with CBCT (MRI-CBCT). The cohort included 50 frame-based (54 targets) and 55 frameless (124 targets) patients treated between August 2020 and December 2022. Target coverage was assessed by the percentage of each target covered by the prescription isodose line. The mean coverage differences for the frame-based and frameless cohorts were −1.20% ± 1.43% and −1.06% ± 1.51%, respectively, with 95% confidence intervals of −1.59, −0.81 and −1.32, −0.79. The MRI-CBCT workflow was deemed noninferior to the MRI-CT-CBCT workflow. Our data indicate that both workflows yield similar results, suggesting that a traditional CT simulation scan is unnecessary.

AbstractBackgroundRecent advancements in Deep Learning (DL) methodologies have led to state‐of‐the‐art performance in a wide range of applications especially in object recognition, classification, and segmentation of medical images. However, training modern DL models requires a large amount of computation and long training times due to the complex nature of network structures and the large number of training datasets involved. Moreover, it is an intensive, repetitive manual process to select the optimized configuration of hyperparameters for a given DL network.PurposeIn this study, we present a novel approach to accelerate the training time of DL models via the progressive feeding of training datasets based on similarity measures for medical image segmentation. We term this approach Progressive Deep Learning (PDL).MethodsThe two‐stage PDL approach was tested on the auto‐segmentation task for two imaging modalities: CT and MRI. The training datasets were ranked according to similarity measures between each sample based on Mean Square Error (MSE), Peak Signal‐to‐Noise Ratio (PSNR), Structural Similarity Index (SSIM), and the Universal Quality Image Index (UQI) values. At the start of the training process, a relatively coarse sampling of training datasets with higher ranks was used to optimize the hyperparameters of the DL network. Following this, the samples with higher ranks were used in step 1 to yield accelerated loss minimization in early training epochs and the total dataset was added in step 2 for the remainder of training.ResultsOur results demonstrate that the PDL approach can reduce the training time by nearly half (∼49%) and can predict segmentations (CT U‐net/DenseNet dice coefficient: 0.9506/0.9508, MR U‐net/DenseNet dice coefficient: 0.9508/0.9510) without major statistical difference (Wilcoxon signed‐rank test) compared to the conventional DL approach. The total training times with a fixed cutoff at 0.95 DSC for the CT dataset using DenseNet and U‐Net architectures, respectively, were 17 h, 20 min and 4 h, 45 min in the conventional case compared to 8 h, 45 min and 2 h, 20 min with PDL. For the MRI dataset, the total training times using the same architectures were 2 h, 54 min and 52 min in the conventional case and 1 h, 14 min and 25 min with PDL.ConclusionThe proposed PDL training approach offers the ability to substantially reduce the training time for medical image segmentation while maintaining the performance achieved in the conventional case.

Abstract Objective. To investigate synchrotron-based proton pencil beam scanning (PBS) beam delivery time (BDT) using novel continuous scanning mode. Approach. A BDT calculation model was developed for the Hitachi particle therapy system. The model was validated against the measured BDT of 36 representative clinical proton PBS plans with discrete spot scanning (DSS) in the current Hitachi proton therapy system. BDTs were calculated with the next generation using Mayo Clinic Florida system operating parameters for conventional DSS, and novel dose driven continuous scanning (DDCS). BDTs of DDCS with and without Break Spots were investigated. Main results. For DDCS without Break Spots, the use of Stop Ratio to control the transit dose largely reduced the beam intensity and consequently, severely prolonged the BDT. DDCS with Break Spots was able to maintain a sufficiently high beam intensity while controlling transit dose. In DDCS with Break Spots, tradeoffs were made between beam intensity and number of Break Spots. Therefore, BDT decreased with increased beam intensity but reached a plateau for beam intensity larger than 10 MU s −1 . Averaging over all clinical plans, BDT was reduced by 10% for DDCS with Break Spots compared to DSS. Significance. DDCS with Break Spots reduced BDT. DDCS has the potential to further reduce BDT under the ideal scenario which requests both stable beam intensity extraction and accurately modelling the transit dose. Further investigation is warranted.