The InterVision Framework: An Enhanced Fine-Tuning Deep Learning Strategy for Auto-Segmentation in Head and Neck.

Adaptive radiotherapy (ART) workflows are increasingly adopted to achieve dose escalation and tissue sparing under dynamic anatomical conditions. However, recontouring and time constraints hinder the implementation of real-time ART workflows. Various auto-segmentation methods, including deformable image registration, atlas-based segmentation, and deep learning-based segmentation (DLS), have been developed to address these challenges. Despite the potential of DLS methods, clinical implementation remains difficult due to the need for large, high-quality datasets to ensure model generalizability. This study introduces an InterVision framework for segmentation. The InterVision framework can interpolate or create intermediate visuals between existing images to generate specific patient characteristics. The InterVision model is trained in two steps: (1) generating a general model using the dataset, and (2) tuning the general model using the dataset generated from the InterVision framework. The InterVision framework generates intermediate images between existing patient image slides using deformable vectors, effectively capturing unique patient characteristics. By creating a more comprehensive dataset that reflects these individual characteristics, the InterVision model demonstrates the ability to produce more accurate contours compared to general models. Models are evaluated using the volumetric dice similarity coefficient (VDSC) and the Hausdorff distance 95% (HD95%) for 18 structures in 20 test patients. As a result, the Dice score was 0.81 ± 0.05 for the general model, 0.82 ± 0.04 for the general fine-tuning model, and 0.85 ± 0.03 for the InterVision model. The Hausdorff distance was 3.06 ± 1.13 for the general model, 2.81 ± 0.77 for the general fine-tuning model, and 2.52 ± 0.50 for the InterVision model. The InterVision model showed the best performance compared to the general model. The InterVision framework presents a versatile approach adaptable to various tasks where prior information is accessible, such as in ART settings. This capability is particularly valuable for accurately predicting complex organs and targets that pose challenges for traditional deep learning algorithms.

In Situ Quantification of Carbonate Species Concentrations, pH, and pCO2 in Calcite Fluid Inclusions Using Confocal Raman Spectroscopy.

Carbonate minerals are globally distributed on the modern and ancient Earth and are abundant in terrestrial and marine depositional environments. Fluid inclusions hosted by calcite retain primary signatures of the source fluid geochemistry at the time of mineral formation (i.e., pCO2) and can be used to reconstruct paleoenvironments. Confocal laser Raman spectroscopy provides a quick, nondestructive approach to measuring the constituents of fluid inclusions in carbonates and is a reliable method for qualitatively determining composition in both the aqueous and gas phases. Here, we demonstrate a method for accurately quantifying bicarbonate and carbonate ion concentrations (down to 20 mM) and pH (7-11) from calcite fluid inclusions using confocal Raman spectroscopy. Instrument calibrations for carbonate (CO32-) and bicarbonate (HCO3-) concentrations and pH were performed using stock solutions. We show that the calcite host mineral does not affect the accurate quantification of carbonate solution concentrations and that these parameters can be used to estimate the pH and pCO2 of a solution entrapped within a fluid inclusion. We apply the technique to Icelandic spar calcite and find a [CO32-] = 0.11, [HCO3-] = 0.17, pH = 10.1, and CO2 parts per million = 2217. The presence of gaseous Raman bands for CO2, CH4, and H2S suggests that the mineral precipitated in a reducing environment.

Enhanced IDOL segmentation framework using personalized hyperspace learning IDOL.

BACKGROUND: Adaptive radiotherapy (ART) workflows have been increasingly adopted to achieve dose escalation and tissue sparing under shifting anatomic conditions, but the necessity of recontouring and the associated time burden hinders a real-time or online ART workflow. In response to this challenge, approaches to auto-segmentation involving deformable image registration, atlas-based segmentation, and deep learning-based segmentation (DLS) have been developed. Despite the particular promise shown by DLS methods, implementing these approaches in a clinical setting remains a challenge, namely due to the difficulty of curating a data set of sufficient size and quality so as to achieve generalizability in a trained model. PURPOSE: To address this challenge, we have developed an intentional deep overfit learning (IDOL) framework tailored to the auto-segmentation task. However, certain limitations were identified, particularly the insufficiency of the personalized dataset to effectively overfit the model. In this study, we introduce a personalized hyperspace learning (PHL)-IDOL segmentation framework capable of generating datasets that induce the model to overfit specific patient characteristics for medical image segmentation. METHODS: The PHL-IDOL model is trained in two stages. In the first, a conventional, general model is trained with a diverse set of patient data (n = 100 patients) consisting of CT images and clinical contours. Following this, the general model is tuned with a data set consisting of two components: (a) selection of a subset of the patient data (m < n) using the similarity metrics (mean square error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and the universal quality image index (UQI) values); (b) adjust the CT and the clinical contours using a deformed vector generated from the reference patient and the selected patients using (a). After training, the general model, the continual model, the conventional IDOL model, and the proposed PHL-IDOL model were evaluated using the volumetric dice similarity coefficient (VDSC) and the Hausdorff distance 95% (HD95%) computed for 18 structures in 20 test patients. RESULTS: Implementing the PHL-IDOL framework resulted in improved segmentation performance for each patient. The Dice scores increased from 0.81 ± $ \pm $ 0.05 with the general model, 0.83 ± 0.04 $ \pm 0.04$ for the continual model, 0.83 ± 0.04 $ \pm 0.04$ for the conventional IDOL model to an average of 0.87 ± 0.03 $ \pm 0.03$ with the PHL-IDOL model. Similarly, the Hausdorff distance decreased from 3.06 ± 0.99 $ \pm 0.99$ with the general model, 2.84 ± 0.69 $ \pm 0.69$ for the continual model, 2.79 ± 0.79 $ \pm 0.79$ for the conventional IDOL model and 2.36 ± 0.52 $ \pm 0.52$ for the PHL-IDOL model. All the standard deviations were decreased by nearly half of the values comparing the general model and the PHL-IDOL model. CONCLUSION: The PHL-IDOL framework applied to the auto-segmentation task achieves improved performance compared to the general DLS approach, demonstrating the promise of leveraging patient-specific prior information in a task central to online ART workflows.

Optimal Self-Care for Surgeons: Sleep, Diet, and Exercise.

BACKGROUND: Surgeons face intense stress, causing hormonal imbalances that harm their health, leading to burnout, chronic illness, and shorter lifespans due to their demanding careers. PURPOSE: This study explores self-care strategies focusing on sleep, nutrition, and exercise to help surgeons reduce stress and improve their overall well-being and quality of life. RESEARCH DESIGN: A thorough literature review of physiological, metabolic, and psychological principles informed the development of a structured self-care approach. DATA COLLECTION AND/OR ANALYSIS: We reviewed existing research on brain-body interactions, highlighting hormonal balance, nutrition, and exercise to mitigate chronic stress. RESULTS: The review underscores the importance of quality sleep for hormonal balance and overall health. Proper nutrition, emphasizing balanced macronutrients and meal timing, supports health. Exercise should be 80% low-intensity aerobic activities, with 20% high-intensity. Combining these elements strengthens resistance to chronic stress and enhances health. CONCLUSIONS: A structured self-care approach, prioritizing sleep, followed by nutrition and exercise, effectively reduces stress among surgeons. This sequence improves well-being and quality of life. Surgeons should focus on consistent sleep, balanced nutrition, and regular low-intensity exercise to enhance resilience and achieve a fulfilling professional life.

An International Retrospective Early Natural History Study of LAMA2-Related Dystrophies.

Background: LAMA2-related dystrophies (LAMA2-RDs) represent one of the most common forms of congenital muscular dystrophy and have historically been classified into two subtypes: complete or partial deficiency of laminin-211 (merosin). Patients with LAMA2-RD with the typical congenital phenotype manifest severe muscle weakness, delayed motor milestones, joint contractures, failure to thrive, and progressive respiratory insufficiency. Objective: While a comprehensive prospective natural history study has been performed in LAMA2-RD patients over 5 years of age, the early natural history of patients with LAMA2-RD 5 years and younger has not been comprehensively characterized. Methods: We extracted retrospective data for patients with LAMA2-RD ages birth through 5 years via the Congenital Muscle Disease International Registry (CMDIR). We analyzed the data using a phenotypic classification based on maximal motor milestones to divide patients into two phenotypic groups: "Sit" for those patients who attained that ability to remain seated and "Walk" for those patients who attained the ability to walk independently by 3.5 years of age. Results: Sixty patients with LAMA2-RD from 10 countries fulfilled the inclusion criteria. Twenty-four patients had initiated non-invasive ventilation by age 5 years. Hospitalizations during the first years of life were often related to respiratory insufficiency. Feeding/nutritional difficulties and orthopedic issues were commonly reported. Significant elevations of creatine kinase (CK) observed during the neonatal period declined rapidly within the first few months of life. Conclusions: This is the largest international retrospective early natural history study of LAMA2-RD to date, contributing essential data for understanding early clinical findings in LAMA2-RD which, along with the data being collected in international, prospective early natural history studies, will help to establish clinical trial readiness. Our proposed nomenclature of LAMA2-RD1 for patients who attain the ability to sit (remain seated) and LAMA2-RD2 for patients who attain the ability to walk independently is aimed at further improving LAMA2-RD classification.

AAPM Task Group 334: A guidance document to using radiotherapy immobilization devices and accessories in an MR environment.

Use of magnetic resonance (MR) imaging in radiation therapy has increased substantially in recent years as more radiotherapy centers are having MR simulators installed, requesting more time on clinical diagnostic MR systems, or even treating with combination MR linear accelerator (MR-linac) systems. With this increased use, to ensure the most accurate integration of images into radiotherapy (RT), RT immobilization devices and accessories must be able to be used safely in the MR environment and produce minimal perturbations. The determination of the safety profile and considerations often falls to the medical physicist or other support staff members who at a minimum should be a Level 2 personnel as per the ACR. The purpose of this guidance document will be to help guide the user in making determinations on MR Safety labeling (i.e., MR Safe, Conditional, or Unsafe) including standard testing, and verification of image quality, when using RT immobilization devices and accessories in an MR environment.

Impact of proton PBS machine operating parameters on the effectiveness of layer rescanning for interplay effect mitigation in lung SBRT treatment.

BACKGROUND: Rescanning is a common technique used in proton pencil beam scanning to mitigate the interplay effect. Advances in machine operating parameters across different generations of particle therapy systems have led to improvements in beam delivery time (BDT). However, the potential impact of these improvements on the effectiveness of rescanning remains an underexplored area in the existing research. METHODS: We systematically investigated the impact of proton machine operating parameters on the effectiveness of layer rescanning in mitigating interplay effect during lung SBRT treatment, using the CIRS phantom. Focused on the Hitachi synchrotron particle therapy system, we explored machine operating parameters from our institution's current (2015) and upcoming systems (2025A and 2025B). Accumulated dynamic 4D dose were reconstructed to assess the interplay effect and layer rescanning effectiveness. RESULTS: Achieving target coverage and dose homogeneity within 2% deviation required 6, 6, and 20 times layer rescanning for the 2015, 2025A, and 2025B machine parameters, respectively. Beyond this point, further increasing the number of layer rescanning did not further improve the dose distribution. BDTs without rescanning were 50.4, 24.4, and 11.4 s for 2015, 2025A, and 2025B, respectively. However, after incorporating proper number of layer rescanning (six for 2015 and 2025A, 20 for 2025B), BDTs increased to 67.0, 39.6, and 42.3 s for 2015, 2025A, and 2025B machine parameters. Our data also demonstrated the potential problem of false negative and false positive if the randomness of the respiratory phase at which the beam is initiated is not considered in the evaluation of interplay effect. CONCLUSION: The effectiveness of layer rescanning for mitigating interplay effect is affected by machine operating parameters. Therefore, past clinical experiences may not be applicable to modern machines.

Does Epidural Corticosteroid Application During Spinal Surgery Reduce Postoperative Pain?: An Adjunct to Multimodal Analgesia.

STUDY DESIGN: A prospective, randomized, placebo-controlled, double-blinded study. OBJECTIVE: To examine the effect of intraoperative epidural administration of Depo-Medrol on postoperative back pain and radiculitis symptoms in patients undergoing Transforaminal Lumbar Interbody Fusion (TLIF). SUMMARY OF BACKGROUND DATA: Postoperative pain is commonly experienced by patients undergoing spinal fusion surgery. Adequate management of intense pain is necessary to encourage early ambulation, increase patient satisfaction, and limit opioid consumption. Intraoperative steroid application has been shown to improve postoperative pain in patients undergoing lumbar decompression surgeries. There have been no studies examining the effect of epidural steroids on both back pain and radicular pain in patients undergoing TLIF. METHOD: In all, 151 patients underwent TLIF surgery using rh-BMP2 with 3 surgeons at a single institution. Of those, 116 remained in the study and were included in the final analysis. Based on a 1:1 randomization, a collagen sponge saturated with either Saline (1 cc) or Depo-Medrol (40 mg/1 cc) was placed at the annulotomy site on the TLIF level. Follow-up occurred on postoperative days 1, 2, 3, 7, and postoperative months 1, 2, and 3. Lumbar radiculopathy was measured by a modified symptom- and laterality-specific Visual Analog Scale (VAS) regarding the severity of back pain and common radiculopathy symptoms. RESULTS: The patients who received Depo-Medrol, compared with those who received saline, experienced significantly less back pain on postoperative days 1, 2, 3, and 7 ( P <0.05). There was no significant difference in back pain beyond day 7. Radiculopathy-related symptoms such as leg pain, numbness, tingling, stiffness, and weakness tended to be reduced in the steroid group at most time points. CONCLUSION: This study provides Level 1 evidence that intraoperative application of Depo-Medrol during a TLIF surgery with rh-BMP2 significantly reduces back pain for the first week after TLIF surgery. The use of epidural Depo-Medrol may be a useful adjunct to multimodal analgesia for pain relief in the postoperative period.

A high-resolution cone beam computed tomography (HRCBCT) reconstruction framework for CBCT-guided online adaptive therapy.

BACKGROUND: Kilo-voltage cone-beam computed tomography (CBCT) is a prevalent modality used for adaptive radiotherapy (ART) due to its compatibility with linear accelerators and ability to provide online imaging. However, the widely-used Feldkamp-Davis-Kress (FDK) reconstruction algorithm has several limitations, including potential streak aliasing artifacts and elevated noise levels. Iterative reconstruction (IR) techniques, such as total variation (TV) minimization, dictionary-based methods, and prior information-based methods, have emerged as viable solutions to address these limitations and improve the quality and applicability of CBCT in ART. PURPOSE: One of the primary challenges in IR-based techniques is finding the right balance between minimizing image noise and preserving image resolution. To overcome this challenge, we have developed a new reconstruction technique called high-resolution CBCT (HRCBCT) that specifically focuses on improving image resolution while reducing noise levels. METHODS: The HRCBCT reconstruction technique builds upon the conventional IR approach, incorporating three components: the data fidelity term, the resolution preservation term, and the regularization term. The data fidelity term ensures alignment between reconstructed values and measured projection data, while the resolution preservation term exploits the high resolution of the initial Feldkamp-Davis-Kress (FDK) algorithm. The regularization term mitigates noise during the IR process. To enhance convergence and resolution at each iterative stage, we applied Iterative Filtered Backprojection (IFBP) to the data fidelity minimization process. RESULTS: We evaluated the performance of the proposed HRCBCT algorithm using data from two physical phantoms and one head and neck patient. The HRCBCT algorithm outperformed all four different algorithms; FDK, Iterative Filtered Back Projection (IFBP), Compressed Sensing based Iterative Reconstruction (CSIR), and Prior Image Constrained Compressed Sensing (PICCS) methods in terms of resolution and noise reduction for all data sets. Line profiles across three line pairs of resolution revealed that the HRCBCT algorithm delivered the highest distinguishable line pairs compared to the other algorithms. Similarly, the Modulation Transfer Function (MTF) measurements, obtained from the tungsten wire insert on the CatPhan 600 physical phantom, showed a significant improvement with HRCBCT over traditional algorithms. CONCLUSION: The proposed HRCBCT algorithm offers a promising solution for enhancing CBCT image quality in adaptive radiotherapy settings. By addressing the challenges inherent in traditional IR methods, the algorithm delivers high-definition CBCT images with improved resolution and reduced noise throughout each iterative step. Implementing the HR CBCT algorithm could significantly impact the accuracy of treatment planning during online adaptive therapy.

Assessing the reliability of pediatric emergency medicine billing code assignment for future consideration as a proxy workload measure.

OBJECTIVES: Prediction of pediatric emergency department (PED) workload can allow for optimized allocation of resources to improve patient care and reduce physician burnout. A measure of PED workload is thus required, but to date no variable has been consistently used or could be validated against for this purpose. Billing codes, a variable assigned by physicians to reflect the complexity of medical decision making, have the potential to be a proxy measure of PED workload but must be assessed for reliability. In this study, we investigated how reliably billing codes are assigned by PED physicians, and factors that affect the inter-rater reliability of billing code assignment. METHODS: A retrospective cross-sectional study was completed to determine the reliability of billing code assigned by physicians (n = 150) at a quaternary-level PED between January 2018 and December 2018. Clinical visit information was extracted from health records and presented to a billing auditor, who independently assigned a billing code-considered as the criterion standard. Inter-rater reliability was calculated to assess agreement between the physician-assigned versus billing auditor-assigned billing codes. Unadjusted and adjusted logistic regression models were used to assess the association between covariables of interest and inter-rater reliability. RESULTS: Overall, we found substantial inter-rater reliability (AC2 0.72 [95% CI 0.64-0.8]) between the billing codes assigned by physicians compared to those assigned by the billing auditor. Adjusted logistic regression models controlling for Pediatric Canadian Triage and Acuity scores, disposition, and time of day suggest that clinical trainee involvement is significantly associated with increased inter-rater reliability. CONCLUSIONS: Our work identified that there is substantial agreement between PED physician and a billing auditor assigned billing codes, and thus are reliably assigned by PED physicians. This is a crucial step in validating billing codes as a potential proxy measure of pediatric emergency physician workload.

Assessing the Accuracy of Responses by the Language Model ChatGPT to Questions Regarding Bariatric Surgery.

PURPOSE: ChatGPT is a large language model trained on a large dataset covering a broad range of topics, including the medical literature. We aim to examine its accuracy and reproducibility in answering patient questions regarding bariatric surgery. MATERIALS AND METHODS: Questions were gathered from nationally regarded professional societies and health institutions as well as Facebook support groups. Board-certified bariatric surgeons graded the accuracy and reproducibility of responses. The grading scale included the following: (1) comprehensive, (2) correct but inadequate, (3) some correct and some incorrect, and (4) completely incorrect. Reproducibility was determined by asking the model each question twice and examining difference in grading category between the two responses. RESULTS: In total, 151 questions related to bariatric surgery were included. The model provided "comprehensive" responses to 131/151 (86.8%) of questions. When examined by category, the model provided "comprehensive" responses to 93.8% of questions related to "efficacy, eligibility and procedure options"; 93.3% related to "preoperative preparation"; 85.3% related to "recovery, risks, and complications"; 88.2% related to "lifestyle changes"; and 66.7% related to "other". The model provided reproducible answers to 137 (90.7%) of questions. CONCLUSION: The large language model ChatGPT often provided accurate and reproducible responses to common questions related to bariatric surgery. ChatGPT may serve as a helpful adjunct information resource for patients regarding bariatric surgery in addition to standard of care provided by licensed healthcare professionals. We encourage future studies to examine how to leverage this disruptive technology to improve patient outcomes and quality of life.

Phase 2 Dose-Finding Study in Patients with Gout Using SEL-212, a Novel PEGylated Uricase (SEL-037) Combined with Tolerogenic Nanoparticles (SEL-110).

INTRODUCTION: SEL-212 is a developmental treatment for uncontrolled gout characterized by serum uric acid (sUA) levels ≥ 6 mg/dl despite treatment. It comprises a novel PEGylated uricase (SEL-037; also called pegadricase) co-administered with tolerogenic nanoparticles containing sirolimus (rapamycin) (SEL-110; also called ImmTOR®), which mitigates the formation of anti-drug antibodies (ADAs) against uricase and SEL-037 (PEGylated uricase), thereby enabling sustained sUA control (sUA < 6 mg/dl). The aim of this study was to identify appropriate dosing for SEL-037 and SEL-110 for use in phase 3 clinical trials. METHODS: This open-label phase 2 study was conducted in adults with symptomatic gout and sUA ≥ 6 mg/dl. Participants received five monthly infusions of SEL-037 (0.2 or 0.4 mg/kg) alone or in combination with three or five monthly infusions of SEL-110 (0.05-0.15 mg/kg). Safety, tolerability, sUA, ADAs, and tophi were monitored for 6 months. RESULTS: A total of 152 adults completed the study. SEL-037 alone resulted in rapid sUA reductions that were not sustained beyond 30 days in most participants due to ADA formation and loss of uricase activity. Levels of ADAs decreased with increasing doses of SEL-110 up to 0.1 mg/kg, with anti-uricase titers < 1080 correlating with sustained sUA control and reductions in tophi. Overall, 66% of evaluable participants achieved sUA control at week 20 following five monthly doses of SEL-037 0.2 mg/kg + SEL-110 0.1-0.15 mg/kg, whereas only 26% achieved sUA control at week 20 when SEL-110 was withdrawn after week 12. Compared to other dose combinations, SEL-037 0.2 mg/kg + SEL-110 0.15 mg/kg achieved the greatest sUA control at week 12 and was well-tolerated with no safety concerns. CONCLUSION: Results provide continued support for the use of multiple monthly administrations of SEL-037 0.2 mg/kg + SEL-110 0.1-0.15 mg/kg in clinical trials for SEL-212. TRIAL REGISTRATION: ClinicalTrials.gov identifier, NCT02959918.

Racial and Ethnic Disparities in Years of Potential Life Loss Among Patients With Cirrhosis During the COVID-19 Pandemic in the United States.

INTRODUCTION: Our aim was to evaluate the impact of race/ethnicity on cirrhosis-related premature death during the COVID-19 pandemic. METHODS: We obtained cirrhosis-related death data (n = 872,965, January 1, 2012-December 31, 2021) from the US National Vital Statistic System to calculate age-standardized mortality rates and years of potential life lost (YPLL) for premature death aged 25-64 years. RESULTS: Significant racial/ethnic disparity in cirrhosis-related age-standardized mortality rates was noted prepandemic but widened during the pandemic, with the highest excess YPLL for the non-Hispanic American Indian/American Native (2020: 41.0%; 2021: 68.8%) followed by other minority groups (28.7%-45.1%), and the non-Hispanic White the lowest (2020: 20.7%; 2021: 31.6%). COVID-19 constituted >30% of the excess YPLLs for Hispanic and non-Hispanic American Indian/American Native in 2020, compared with 11.1% for non-Hispanic White. DISCUSSION: Ethnic minorities with cirrhosis experienced a disproportionate excess death and YPLLs in 2020-2021.

Technical note: Progressive deep learning: An accelerated training strategy for medical image segmentation.

BACKGROUND: Recent 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. PURPOSE: In 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). METHODS: The 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. RESULTS: Our 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. CONCLUSION: The 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.

Segmentation by test-time optimization for CBCT-based adaptive radiation therapy.

PURPOSE: Online adaptive radiotherapy (ART) requires accurate and efficient auto-segmentation of target volumes and organs-at-risk (OARs) in mostly cone-beam computed tomography (CBCT) images, which often have severe artifacts and lack soft-tissue contrast, making direct segmentation very challenging. Propagating expert-drawn contours from the pretreatment planning CT through traditional or deep learning (DL)-based deformable image registration (DIR) can achieve improved results in many situations. Typical DL-based DIR models are population based, that is, trained with a dataset for a population of patients, and so they may be affected by the generalizability problem. METHODS: In this paper, we propose a method called test-time optimization (TTO) to refine a pretrained DL-based DIR population model, first for each individual test patient, and then progressively for each fraction of online ART treatment. Our proposed method is less susceptible to the generalizability problem and thus can improve overall performance of different DL-based DIR models by improving model accuracy, especially for outliers. Our experiments used data from 239 patients with head-and-neck squamous cell carcinoma to test the proposed method. First, we trained a population model with 200 patients and then applied TTO to the remaining 39 test patients by refining the trained population model to obtain 39 individualized models. We compared each of the individualized models with the population model in terms of segmentation accuracy. RESULTS: The average improvement of the Dice similarity coefficient (DSC) and 95% Hausdorff distance (HD95) of segmentation can be up to 0.04 (5%) and 0.98 mm (25%), respectively, with the individualized models compared to the population model over 17 selected OARs and a target of 39 patients. Although the average improvement may seem mild, we found that the improvement for outlier patients with structures of large anatomical changes is significant. The number of patients with at least 0.05 DSC improvement or 2 mm HD95 improvement by TTO averaged over the 17 selected structures for the state-of-the-art architecture VoxelMorph is 10 out of 39 test patients. By deriving the individualized model using TTO from the pretrained population model, TTO models can be ready in about 1 min. We also generated the adapted fractional models for each of the 39 test patients by progressively refining the individualized models using TTO to CBCT images acquired at later fractions of online ART treatment. When adapting the individualized model to a later fraction of the same patient, the model can be ready in less than a minute with slightly improved accuracy. CONCLUSIONS: The proposed TTO method is well suited for online ART and can boost segmentation accuracy for DL-based DIR models, especially for outlier patients where the pretrained models fail.

Ensemble learning and personalized training for the improvement of unsupervised deep learning-based synthetic CT reconstruction.

BACKGROUND: The growing adoption of magnetic resonance imaging (MRI)-guided radiation therapy (RT) platforms and a focus on MRI-only RT workflows have brought the technical challenge of synthetic computed tomography (sCT) reconstruction to the forefront. Unpaired-data deep learning-based approaches to the problem offer the attractive characteristic of not requiring paired training data, but the gap between paired- and unpaired-data results can be limiting. PURPOSE: We present two distinct approaches aimed at improving unpaired-data sCT reconstruction results: a cascade ensemble that combines multiple models and a personalized training strategy originally designed for the paired-data setting. METHODS: Comparisons are made between the following models: (1) the paired-data fully convolutional DenseNet (FCDN), (2) the FCDN with the Intentional Deep Overfit Learning (IDOL) personalized training strategy, (3) the unpaired-data CycleGAN, (4) the CycleGAN with the IDOL training strategy, and (5) the CycleGAN as an intermediate model in a cascade ensemble approach. Evaluation of the various models over 25 total patients is carried out using a five-fold cross-validation scheme, with the patient-specific IDOL models being trained for the five patients of fold 3, chosen at random. RESULTS: In both the paired- and unpaired-data settings, adopting the IDOL training strategy led to improvements in the mean absolute error (MAE) between true CT images and sCT outputs within the body contour (mean improvement, paired- and unpaired-data approaches, respectively: 38%, 9%) and in regions of bone (52%, 5%), the peak signal-to-noise ratio (PSNR; 15%, 7%), and the structural similarity index (SSIM; 6%, <1%). The ensemble approach offered additional benefits over the IDOL approach in all three metrics (mean improvement over unpaired-data approach in fold 3; MAE: 20%; bone MAE: 16%; PSNR: 10%; SSIM: 2%), and differences in body MAE between the ensemble approach and the paired-data approach are statistically insignificant. CONCLUSIONS: We have demonstrated that both a cascade ensemble approach and a personalized training strategy designed initially for the paired-data setting offer significant improvements in image quality metrics for the unpaired-data sCT reconstruction task. Closing the gap between paired- and unpaired-data approaches is a step toward fully enabling these powerful and attractive unpaired-data frameworks.

Oral cancer patients achieve comparable survival at high safety-net burden hospitals.

PURPOSE: To evaluate the impact of hospital safety-net burden and social demographics on the overall survival of patients with oral cavity squamous cell carcinoma. MATERIALS AND METHODS: We identified 48,176 oral cancer patients diagnosed between the years 2004 to 2015 from the National Cancer Database and categorized treatment facilities as no, low, or high safety-net burden hospitals based on the percentage of uninsured or Medicaid patients treated. Using the Kaplan Meier method and multivariate analysis, we examined the effect of hospital safety-net burden, sociodemographic variables, and clinical factors on overall survival. RESULTS: Of the 1269 treatment facilities assessed, the median percentage of uninsured/Medicaid patients treated was 0% at no, 11.6% at low, and 23.5% at high safety-net burden hospitals and median survival was 68.6, 74.8, and 55.0 months, respectively (p < 0.0001). High safety-net burden hospitals treated more non-white populations (15.4%), lower median household income (<$30,000) (23.2%), and advanced stage cancers (AJCC III/IV) (54.6%). Patients treated at low (aHR = 0.97; 95% CI = 0.91-1.04, p = 0.405) and high (aHR = 1.05; 95% CI = 0.98-1.13, p = 0.175) safety-net burden hospitals did not experience worse survival outcomes compared to patients treated at no safety-net burden hospitals. CONCLUSION: High safety-net burden hospitals treated more oral cancer patients of lower socioeconomic status and advanced disease. Multivariate analysis showed high safety-net burden hospitals achieved comparable patient survival to lower burden hospitals.

Real-time liver tumor localization via a single x-ray projection using deep graph neural network-assisted biomechanical modeling.

Objective.Real-time imaging is highly desirable in image-guided radiotherapy, as it provides instantaneous knowledge of patients' anatomy and motion during treatments and enables online treatment adaptation to achieve the highest tumor targeting accuracy. Due to extremely limited acquisition time, only one or few x-ray projections can be acquired for real-time imaging, which poses a substantial challenge to localize the tumor from the scarce projections. For liver radiotherapy, such a challenge is further exacerbated by the diminished contrast between the tumor and the surrounding normal liver tissues. Here, we propose a framework combining graph neural network-based deep learning and biomechanical modeling to track liver tumor in real-time from a single onboard x-ray projection.Approach.Liver tumor tracking is achieved in two steps. First, a deep learning network is developed to predict the liver surface deformation using image features learned from the x-ray projection. Second, the intra-liver deformation is estimated through biomechanical modeling, using the liver surface deformation as the boundary condition to solve tumor motion by finite element analysis. The accuracy of the proposed framework was evaluated using a dataset of 10 patients with liver cancer.Main results.The results show accurate liver surface registration from the graph neural network-based deep learning model, which translates into accurate, fiducial-less liver tumor localization after biomechanical modeling (<1.2 (±1.2) mm average localization error).Significance.The method demonstrates its potentiality towards intra-treatment and real-time 3D liver tumor monitoring and localization. It could be applied to facilitate 4D dose accumulation, multi-leaf collimator tracking and real-time plan adaptation. The method can be adapted to other anatomical sites as well.

Intentional deep overfit learning (IDOL): A novel deep learning strategy for adaptive radiation therapy.

PURPOSE: Applications of deep learning (DL) are essential to realizing an effective adaptive radiotherapy (ART) workflow. Despite the promise demonstrated by DL approaches in several critical ART tasks, there remain unsolved challenges to achieve satisfactory generalizability of a trained model in a clinical setting. Foremost among these is the difficulty of collecting a task-specific training dataset with high-quality, consistent annotations for supervised learning applications. In this study, we propose a tailored DL framework for patient-specific performance that leverages the behavior of a model intentionally overfitted to a patient-specific training dataset augmented from the prior information available in an ART workflow-an approach we term Intentional Deep Overfit Learning (IDOL). METHODS: Implementing the IDOL framework in any task in radiotherapy consists of two training stages: (1) training a generalized model with a diverse training dataset of N patients, just as in the conventional DL approach, and (2) intentionally overfitting this general model to a small training dataset-specific the patient of interest ( N + 1 ) generated through perturbations and augmentations of the available task- and patient-specific prior information to establish a personalized IDOL model. The IDOL framework itself is task-agnostic and is, thus, widely applicable to many components of the ART workflow, three of which we use as a proof of concept here: the autocontouring task on replanning CTs for traditional ART, the MRI super-resolution (SR) task for MRI-guided ART, and the synthetic CT (sCT) reconstruction task for MRI-only ART. RESULTS: In the replanning CT autocontouring task, the accuracy measured by the Dice similarity coefficient improves from 0.847 with the general model to 0.935 by adopting the IDOL model. In the case of MRI SR, the mean absolute error (MAE) is improved by 40% using the IDOL framework over the conventional model. Finally, in the sCT reconstruction task, the MAE is reduced from 68 to 22 HU by utilizing the IDOL framework. CONCLUSIONS: In this study, we propose a novel IDOL framework for ART and demonstrate its feasibility using three ART tasks. We expect the IDOL framework to be especially useful in creating personally tailored models in situations with limited availability of training data but existing prior information, which is usually true in the medical setting in general and is especially true in ART.