Cluster-Based Toxicity Estimation of Osteoradionecrosis Via Unsupervised Machine Learning: Moving Beyond Single Dose-Parameter Normal Tissue Complication Probability by Using Whole Dose-Volume Histograms for Cohort Risk Stratification.

PURPOSE: Given the limitations of extant models for normal tissue complication probability estimation for osteoradionecrosis (ORN) of the mandible, the purpose of this study was to enrich statistical inference by exploiting structural properties of data and provide a clinically reliable model for ORN risk evaluation through an unsupervised-learning analysis that incorporates the whole radiation dose distribution on the mandible. METHODS AND MATERIALS: The analysis was conducted on retrospective data of 1259 patients with head and neck cancer treated at The University of Texas MD Anderson Cancer Center between 2005 and 2015. During a minimum 12-month posttherapy follow-up period, 173 patients in this cohort (13.7%) developed ORN (grades I to IV). The (structural) clusters of mandibular dose-volume histograms (DVHs) for these patients were identified using the K-means clustering method. A soft-margin support vector machine was used to determine the cluster borders and partition the dose-volume space. The risk of ORN for each dose-volume region was calculated based on incidence rates and other clinical risk factors. RESULTS: The K-means clustering method identified 6 clusters among the DVHs. Based on the first 5 clusters, the dose-volume space was partitioned by the soft-margin support vector machine into distinct regions with different risk indices. The sixth cluster entirely overlapped with the others; the region of this cluster was determined by its envelopes. For each region, the ORN incidence rate per preradiation dental extraction status (a statistically significant, nondose related risk factor for ORN) was reported as the corresponding risk index. CONCLUSIONS: This study presents an unsupervised-learning analysis of a large-scale data set to evaluate the risk of mandibular ORN among patients with head and neck cancer. The results provide a visual risk-assessment tool for ORN (based on the whole DVH and preradiation dental extraction status) as well as a range of constraints for dose optimization under different risk levels.

Comparison of Machine Leaning Models for Prediction of Acute Pain Severity and On-Treatment Opioid Utilization in Oral Cavity and Oropharyngeal Cancer Patients Receiving Radiation Therapy: Exploratory Analysis from a Large-Scale Retrospective Cohort.

Background: Acute pain is a common and debilitating symptom experienced by oral cavity and oropharyngeal cancer (OC/OPC) patients undergoing radiation therapy (RT). Uncontrolled pain can result in opioid overuse and increased risks of long-term opioid dependence. The specific aim of this exploratory analysis was the prediction of severe acute pain and opioid use in the acute on-treatment setting, to develop risk-stratification models for pragmatic clinical trials. Materials and Methods: A retrospective study was conducted on 900 OC/OPC patients treated with RT during 2017 to 2023. Clinical data including demographics, tumor data, pain scores and medication data were extracted from patient records. On-treatment pain intensity scores were assessed using a numeric rating scale (0-none, 10-worst) and total opioid doses were calculated using morphine equivalent daily dose (MEDD) conversion factors. Analgesics efficacy was assessed based on the combined pain intensity and the total required MEDD. ML models, including Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and Gradient Boosting Model (GBM) were developed and validated using ten-fold cross-validation. Performance of models were evaluated using discrimination and calibration metrics. Feature importance was investigated using bootstrap and permutation techniques. Results: For predicting acute pain intensity, the GBM demonstrated superior area under the receiver operating curve (AUC) (0.71), recall (0.39), and F1 score (0.48). For predicting the total MEDD, LR outperformed other models in the AUC (0.67). For predicting the analgesics efficacy, SVM achieved the highest specificity (0.97), and best calibration (ECE of 0.06), while RF and GBM achieved the same highest AUC, 0.68. RF model emerged as the best calibrated model with ECE of 0.02 for pain intensity prediction and 0.05 for MEDD prediction. Baseline pain scores and vital signs demonstrated the most contributed features for the different predictive models. Conclusion: These ML models are promising in predicting end-of-treatment acute pain and opioid requirements and analgesics efficacy in OC/OPC patients undergoing RT. Baseline pain score, vital sign changes were identified as crucial predictors. Implementation of these models in clinical practice could facilitate early risk stratification and personalized pain management. Prospective multicentric studies and external validation are essential for further refinement and generalizability.

Investigation of autosegmentation techniques on T2-weighted MRI for off-line dose reconstruction in MR-linac workflow for head and neck cancers.

BACKGROUND: In order to accurately accumulate delivered dose for head and neck cancer patients treated with the Adapt to Position workflow on the 1.5T magnetic resonance imaging (MRI)-linear accelerator (MR-linac), the low-resolution T2-weighted MRIs used for daily setup must be segmented to enable reconstruction of the delivered dose at each fraction. PURPOSE: In this pilot study, we evaluate various autosegmentation methods for head and neck organs at risk (OARs) on on-board setup MRIs from the MR-linac for off-line reconstruction of delivered dose. METHODS: Seven OARs (parotid glands, submandibular glands, mandible, spinal cord, and brainstem) were contoured on 43 images by seven observers each. Ground truth contours were generated using a simultaneous truth and performance level estimation (STAPLE) algorithm. Twenty total autosegmentation methods were evaluated in ADMIRE: 1-9) atlas-based autosegmentation using a population atlas library (PAL) of 5/10/15 patients with STAPLE, patch fusion (PF), random forest (RF) for label fusion; 10-19) autosegmentation using images from a patient's 1-4 prior fractions (individualized patient prior [IPP]) using STAPLE/PF/RF; 20) deep learning (DL) (3D ResUNet trained on 43 ground truth structure sets plus 45 contoured by one observer). Execution time was measured for each method. Autosegmented structures were compared to ground truth structures using the Dice similarity coefficient, mean surface distance (MSD), Hausdorff distance (HD), and Jaccard index (JI). For each metric and OAR, performance was compared to the inter-observer variability using Dunn's test with control. Methods were compared pairwise using the Steel-Dwass test for each metric pooled across all OARs. Further dosimetric analysis was performed on three high-performing autosegmentation methods (DL, IPP with RF and 4 fractions [IPP_RF_4], IPP with 1 fraction [IPP_1]), and one low-performing (PAL with STAPLE and 5 atlases [PAL_ST_5]). For five patients, delivered doses from clinical plans were recalculated on setup images with ground truth and autosegmented structure sets. Differences in maximum and mean dose to each structure between the ground truth and autosegmented structures were calculated and correlated with geometric metrics. RESULTS: DL and IPP methods performed best overall, all significantly outperforming inter-observer variability and with no significant difference between methods in pairwise comparison. PAL methods performed worst overall; most were not significantly different from the inter-observer variability or from each other. DL was the fastest method (33 s per case) and PAL methods the slowest (3.7-13.8 min per case). Execution time increased with a number of prior fractions/atlases for IPP and PAL. For DL, IPP_1, and IPP_RF_4, the majority (95%) of dose differences were within ± 250 cGy from ground truth, but outlier differences up to 785 cGy occurred. Dose differences were much higher for PAL_ST_5, with outlier differences up to 1920 cGy. Dose differences showed weak but significant correlations with all geometric metrics (R2 between 0.030 and 0.314). CONCLUSIONS: The autosegmentation methods offering the best combination of performance and execution time are DL and IPP_1. Dose reconstruction on on-board T2-weighted MRIs is feasible with autosegmented structures with minimal dosimetric variation from ground truth, but contours should be visually inspected prior to dose reconstruction in an end-to-end dose accumulation workflow.

Patterns of failure for recurrent head and neck squamous cell carcinoma treated with salvage surgery and postoperative IMRT reirradiation.

Purpose/Objectives: The purpose of this study was to evaluate patterns of locoregional recurrence (LRR) after surgical salvage and adjuvant reirradiation with IMRT for recurrent head and neck squamous cell cancer (HNSCC). Materials/Methods: Patterns of LRR for 61 patients treated consecutively between 2003 and 2014 who received post-operative IMRT reirradiation to ≥ 60 Gy for recurrent HNSCC were determined by 2 methods: 1) physician classification via visual comparison of post-radiotherapy imaging to reirradiation plans; and 2) using deformable image registration (DIR). Those without evaluable CT planning image data were excluded. All recurrences were verified by biopsy or radiological progression. Failures were defined as in-field, marginal, or out-of-field. Logistic regression analyses were performed to identify predictors for LRR. Results: A total of 55 patients were eligible for analysis and 23 (42 %) had documented LRR after reirradiation. Location of recurrent disease prior to salvage surgery (lymphatic vs. mucosal) was the most significant predictor of LRR after post-operative reirradiation with salvage rate of 67 % for lymphatic vs. 33 % for mucosal sites (p = 0.037). Physician classification of LRR yielded 14 (61 %) in-field failures, 3 (13 %) marginal failures, and 6 (26 %) out-of-field failures, while DIR yielded 10 (44 %) in-field failures, 4 (17 %) marginal failures, and 9 (39 %) out-of-field failures. Most failures (57 %) occurred within the original site of recurrence or first echelon lymphatic drainage. Of patients who had a free flap placed during salvage surgery, 56 % of failures occurred within 1 cm of the surgical flap. Conclusion: Our study highlights the role of DIR in enhancing the accuracy and consistency of POF analysis. Compared to traditional visual inspection, DIR reduces interobserver variability and provides more nuanced insights into dose-specific and spatial parameters of locoregional recurrences. Additionally, the study identifies the location of the initial recurrence as a key predictor of subsequent locoregional recurrence after salvage surgery and re-IMRT.

Artificial Intelligence and Machine Learning in Cancer Related Pain: A Systematic Review.

Background/objective: Pain is a challenging multifaceted symptom reported by most cancer patients, resulting in a substantial burden on both patients and healthcare systems. This systematic review aims to explore applications of artificial intelligence/machine learning (AI/ML) in predicting pain-related outcomes and supporting decision-making processes in pain management in cancer. Methods: A comprehensive search of Ovid MEDLINE, EMBASE and Web of Science databases was conducted using terms including "Cancer", "Pain", "Pain Management", "Analgesics", "Opioids", "Artificial Intelligence", "Machine Learning", "Deep Learning", and "Neural Networks" published up to September 7, 2023. The screening process was performed using the Covidence screening tool. Only original studies conducted in human cohorts were included. AI/ML models, their validation and performance and adherence to TRIPOD guidelines were summarized from the final included studies. Results: This systematic review included 44 studies from 2006-2023. Most studies were prospective and uni-institutional. There was an increase in the trend of AI/ML studies in cancer pain in the last 4 years. Nineteen studies used AI/ML for classifying cancer patients' pain development after cancer therapy, with median AUC 0.80 (range 0.76-0.94). Eighteen studies focused on cancer pain research with median AUC 0.86 (range 0.50-0.99), and 7 focused on applying AI/ML for cancer pain management decisions with median AUC 0.71 (range 0.47-0.89). Multiple ML models were investigated with. median AUC across all models in all studies (0.77). Random forest models demonstrated the highest performance (median AUC 0.81), lasso models had the highest median sensitivity (1), while Support Vector Machine had the highest median specificity (0.74). Overall adherence of included studies to TRIPOD guidelines was 70.7%. Lack of external validation (14%) and clinical application (23%) of most included studies was detected. Reporting of model calibration was also missing in the majority of studies (5%). Conclusion: Implementation of various novel AI/ML tools promises significant advances in the classification, risk stratification, and management decisions for cancer pain. These advanced tools will integrate big health-related data for personalized pain management in cancer patients. Further research focusing on model calibration and rigorous external clinical validation in real healthcare settings is imperative for ensuring its practical and reliable application in clinical practice.

Development and implementation of optimized endogenous contrast sequences for delineation in adaptive radiotherapy on a 1.5T MR-linear-accelerator: a prospective R-IDEAL stage 0-2a quantitative/qualitative evaluation of in vivo site-specific quality-assurance using a 3D T2 fat-suppressed platform for head and neck cancer.

Purpose: To improve segmentation accuracy in head and neck cancer (HNC) radiotherapy treatment planning for the 1.5T hybrid magnetic resonance imaging/linear accelerator (MR-Linac), three-dimensional (3D), T2-weighted, fat-suppressed magnetic resonance imaging sequences were developed and optimized. Approach: After initial testing, spectral attenuated inversion recovery (SPAIR) was chosen as the fat suppression technique. Five candidate SPAIR sequences and a nonsuppressed, T2-weighted sequence were acquired for five HNC patients using a 1.5T MR-Linac. MR physicists identified persistent artifacts in two of the SPAIR sequences, so the remaining three SPAIR sequences were further analyzed. The gross primary tumor volume, metastatic lymph nodes, parotid glands, and pterygoid muscles were delineated using five segmentors. A robust image quality analysis platform was developed to objectively score the SPAIR sequences on the basis of qualitative and quantitative metrics. Results: Sequences were analyzed for the signal-to-noise ratio and the contrast-to-noise ratio and compared with fat and muscle, conspicuity, pairwise distance metrics, and segmentor assessments. In this analysis, the nonsuppressed sequence was inferior to each of the SPAIR sequences for the primary tumor, lymph nodes, and parotid glands, but it was superior for the pterygoid muscles. The SPAIR sequence that received the highest combined score among the analysis categories was recommended to Unity MR-Linac users for HNC radiotherapy treatment planning. Conclusions: Our study led to two developments: an optimized, 3D, T2-weighted, fat-suppressed sequence that can be disseminated to Unity MR-Linac users and a robust image quality analysis pathway that can be used to objectively score SPAIR sequences and can be customized and generalized to any image quality optimization protocol. Improved segmentation accuracy with the proposed SPAIR sequence will potentially lead to improved treatment outcomes and reduced toxicity for patients by maximizing the target coverage and minimizing the radiation exposure of organs at risk.

Determining The Role Of Radiation Oncologist Demographic Factors On Segmentation Quality: Insights From A Crowd-Sourced Challenge Using Bayesian Estimation.

BACKGROUND: Medical image auto-segmentation is poised to revolutionize radiotherapy workflows. The quality of auto-segmentation training data, primarily derived from clinician observers, is of utmost importance. However, the factors influencing the quality of these clinician-derived segmentations have yet to be fully understood or quantified. Therefore, the purpose of this study was to determine the role of common observer demographic variables on quantitative segmentation performance. METHODS: Organ at risk (OAR) and tumor volume segmentations provided by radiation oncologist observers from the Contouring Collaborative for Consensus in Radiation Oncology public dataset were utilized for this study. Segmentations were derived from five separate disease sites comprised of one patient case each: breast, sarcoma, head and neck (H&N), gynecologic (GYN), and gastrointestinal (GI). Segmentation quality was determined on a structure-by-structure basis by comparing the observer segmentations with an expert-derived consensus gold standard primarily using the Dice Similarity Coefficient (DSC); surface DSC was investigated as a secondary metric. Metrics were stratified into binary groups based on previously established structure-specific expert-derived interobserver variability (IOV) cutoffs. Generalized linear mixed-effects models using Markov chain Monte Carlo Bayesian estimation were used to investigate the association between demographic variables and the binarized segmentation quality for each disease site separately. Variables with a highest density interval excluding zero - loosely analogous to frequentist significance - were considered to substantially impact the outcome measure. RESULTS: After filtering by practicing radiation oncologists, 574, 110, 452, 112, and 48 structure observations remained for the breast, sarcoma, H&N, GYN, and GI cases, respectively. The median percentage of observations that crossed the expert DSC IOV cutoff when stratified by structure type was 55% and 31% for OARs and tumor volumes, respectively. Bayesian regression analysis revealed tumor category had a substantial negative impact on binarized DSC for the breast (coefficient mean ± standard deviation: -0.97 ± 0.20), sarcoma (-1.04 ± 0.54), H&N (-1.00 ± 0.24), and GI (-2.95 ± 0.98) cases. There were no clear recurring relationships between segmentation quality and demographic variables across the cases, with most variables demonstrating large standard deviations and wide highest density intervals. CONCLUSION: Our study highlights substantial uncertainty surrounding conventionally presumed factors influencing segmentation quality. Future studies should investigate additional demographic variables, more patients and imaging modalities, and alternative metrics of segmentation acceptability.

Multi-organ spatial stratification of 3-D dose distributions improves risk prediction of long-term self-reported severe symptoms in oropharyngeal cancer patients receiving radiotherapy: development of a pre-treatment decision support tool.

Purpose: Identify Oropharyngeal cancer (OPC) patients at high-risk of developing long-term severe radiation-associated symptoms using dose volume histograms for organs-at-risk, via unsupervised clustering. Material and methods: All patients were treated using radiation therapy for OPC. Dose-volume histograms of organs-at-risk were extracted from patients' treatment plans. Symptom ratings were collected via the MD Anderson Symptom Inventory (MDASI) given weekly during, and 6 months post-treatment. Drymouth, trouble swallowing, mucus, and vocal dysfunction were selected for analysis in this study. Patient stratifications were obtained by applying Bayesian Mixture Models with three components to patient's dose histograms for relevant organs. The clusters with the highest total mean doses were translated into dose thresholds using rule mining. Patient stratifications were compared against Tumor staging information using multivariate likelihood ratio tests. Model performance for prediction of moderate/severe symptoms at 6 months was compared against normal tissue complication probability (NTCP) models using cross-validation. Results: A total of 349 patients were included for long-term symptom prediction. High-risk clusters were significantly correlated with outcomes for severe late drymouth (p <.0001, OR = 2.94), swallow (p = .002, OR = 5.13), mucus (p = .001, OR = 3.18), and voice (p = .009, OR = 8.99). Simplified clusters were also correlated with late severe symptoms for drymouth (p <.001, OR = 2.77), swallow (p = .01, OR = 3.63), mucus (p = .01, OR = 2.37), and voice (p <.001, OR = 19.75). Proposed cluster stratifications show better performance than NTCP models for severe drymouth (AUC.598 vs.559, MCC.143 vs.062), swallow (AUC.631 vs.561, MCC.20 vs -.030), mucus (AUC.596 vs.492, MCC.164 vs -.041), and voice (AUC.681 vs.555, MCC.181 vs -.019). Simplified dose thresholds also show better performance than baseline models for predicting late severe ratings for all symptoms. Conclusion: Our results show that leveraging the 3-D dose histograms from radiation therapy plan improves stratification of patients according to their risk of experiencing long-term severe radiation associated symptoms, beyond existing NTPC models. Our rule-based method can approximate our stratifications with minimal loss of accuracy and can proactively identify risk factors for radiation-associated toxicity.

Spatially-aware clustering improves AJCC-8 risk stratification performance in oropharyngeal carcinomas.

OBJECTIVE: Evaluate the effectiveness of machine learning tools that incorporate spatial information such as disease location and lymph node metastatic patterns-of-spread, for prediction of survival and toxicity in HPV+ oropharyngeal cancer (OPC). MATERIALS & METHODS: 675 HPV+ OPC patients that were treated at MD Anderson Cancer Center between 2005 and 2013 with curative intent IMRT were retrospectively collected under IRB approval. Risk stratifications incorporating patient radiometric data and lymph node metastasis patterns via an anatomically-adjacent representation with hierarchical clustering were identified. These clusterings were combined into a 3-level patient stratification and included along with other known clinical features in a Cox model for predicting survival outcomes, and logistic regression for toxicity, using independent subsets for training and validation. RESULTS: Four groups were identified and combined into a 3-level stratification. The inclusion of patient stratifications in predictive models for 5-yr Overall survival (OS), 5-year recurrence free survival, (RFS) and Radiation-associated dysphagia (RAD) consistently improved model performance measured using the area under the curve (AUC). Test set AUC improvements over models with clinical covariates, was 9 % for predicting OS, and 18 % for predicting RFS, and 7 % for predicting RAD. For models with both clinical and AJCC covariates, AUC improvement was 7 %, 9 %, and 2 % for OS, RFS, and RAD, respectively. CONCLUSION: Including data-driven patient stratifications considerably improve prognosis for survival and toxicity outcomes over the performance achieved by clinical staging and clinical covariates alone. These stratifications generalize well to across cohorts, and sufficient information for reproducing these clusters is included.

Prospective evaluation of in vivo and phantom repeatability and reproducibility of diffusion-weighted MRI sequences on 1.5 T MRI-linear accelerator (MR-Linac) and MR simulator devices for head and neck cancers.

INTRODUCTION: Diffusion-weighted imaging (DWI) on MRI-linear accelerator (MR-linac) systems can potentially be used for monitoring treatment response and adaptive radiotherapy in head and neck cancers (HNC) but requires extensive validation. We performed technical validation to compare six total DWI sequences on an MR-linac and MR simulator (MR sim) in patients, volunteers, and phantoms. METHODS: Ten human papillomavirus-positive oropharyngeal cancer patients and ten healthy volunteers underwent DWI on a 1.5 T MR-linac with three DWI sequences: echo planar imaging (EPI), split acquisition of fast spin echo signals (SPLICE), and turbo spin echo (TSE). Volunteers were also imaged on a 1.5 T MR sim with three sequences: EPI, BLADE (vendor tradename), and readout segmentation of long variable echo trains (RESOLVE). Participants underwent two scan sessions per device and two repeats of each sequence per session. Repeatability and reproducibility within-subject coefficient of variation (wCV) of mean ADC were calculated for tumors and lymph nodes (patients) and parotid glands (volunteers). ADC bias, repeatability/reproducibility metrics, SNR, and geometric distortion were quantified using a phantom. RESULTS: In vivo repeatability/reproducibility wCV for parotids were 5.41%/6.72%, 3.83%/8.80%, 5.66%/10.03%, 3.44%/5.70%, 5.04%/5.66%, 4.23%/7.36% for EPIMR-linac, SPLICE, TSE, EPIMR sim, BLADE, RESOLVE. Repeatability/reproducibility wCV for EPIMR-linac, SPLICE, TSE were 9.64%/10.28%, 7.84%/8.96%, 7.60%/11.68% for tumors and 7.80%/9.95%, 7.23%/8.48%, 10.82%/10.44% for nodes. All sequences except TSE had phantom ADC biases within ± 0.1x10-3 mm2/s for most vials (EPIMR-linac, SPLICE, and BLADE had 2, 3, and 1 vials out of 13 with larger biases, respectively). SNR of b = 0 images was 87.3, 180.5, 161.3, 171.0, 171.9, 130.2 for EPIMR-linac, SPLICE, TSE, EPIMR sim, BLADE, RESOLVE. CONCLUSION: MR-linac DWI sequences demonstrated near-comparable performance to MR sim sequences and warrant further clinical validation for treatment response assessment in HNC.

Cluster-Based Toxicity Estimation of Osteoradionecrosis via Unsupervised Machine Learning: Moving Beyond Single Dose-Parameter Normal Tissue Complication Probability by Using Whole Dose-Volume Histograms for Cohort Risk Stratification.

Purpose: Given the limitations of extant models for normal tissue complication probability estimation for osteoradionecrosis (ORN) of the mandible, the purpose of this study was to enrich statistical inference by exploiting structural properties of data and provide a clinically reliable model for ORN risk evaluation through an unsupervised-learning analysis. Materials and Methods: The analysis was conducted on retrospective data of 1,259 head and neck cancer (HNC) patients treated at the University of Texas MD Anderson Cancer Center between 2005 and 2015. The (structural) clusters of mandibular dose-volume histograms (DVHs) were identified through the K-means clustering method. A soft-margin support vector machine (SVM) was used to determine the cluster borders and partition the dose-volume space. The risk of ORN for each dose-volume region was calculated based on the clinical risk factors and incidence rates. Results: The K-means clustering method identified six clusters among the DVHs. Based on the first five clusters, the dose-volume space was partitioned almost perfectly by the soft-margin SVM into distinct regions with different risk indices. The sixth cluster overlapped the others entirely; the region of this cluster was determined by its envelops. These regions and the associated risk indices provide a range of constraints for dose optimization under different risk levels. Conclusion: This study presents an unsupervised-learning analysis of a large-scale data set to evaluate the risk of mandibular ORN among HNC patients. The results provide a visual risk-assessment tool (based on the whole DVH) and a spectrum of dose constraints for radiation planning.

Prospective validation of diffusion-weighted MRI as a biomarker of tumor response and oncologic outcomes in head and neck cancer: Results from an observational biomarker pre-qualification study.

PURPOSE: To determine DWI parameters associated with tumor response and oncologic outcomes in head and neck (HNC) patients treated with radiotherapy (RT). METHODS: HNC patients in a prospective study were included. Patients had MRIs pre-, mid-, and post-RT completion. We used T2-weighted sequences for tumor segmentation which were co-registered to respective DWIs for extraction of apparent diffusion coefficient (ADC) measurements. Treatment response was assessed at mid- and post-RT and was defined as: complete response (CR) vs. non-complete response (non-CR). The Mann-Whitney U test was used to compare ADC between CR and non-CR. Recursive partitioning analysis (RPA) was performed to identify ADC threshold associated with relapse. Cox proportional hazards models were done for clinical vs. clinical and imaging parameters and internal validation was done using bootstrapping technique. RESULTS: Eighty-one patients were included. Median follow-up was 31 months. For patients with post-RT CR, there was a significant increase in mean ADC at mid-RT compared to baseline ((1.8 ± 0.29) × 10-3 mm2/s vs. (1.37 ± 0.22) × 10-3 mm2/s, p < 0.0001), while patients with non-CR had no significant increase (p > 0.05). RPA identified GTV-P delta (Δ)ADCmean < 7% at mid-RT as the most significant parameter associated with worse LC and RFS (p = 0.01). Uni- and multi-variable analysis showed that GTV-P ΔADCmean at mid-RT ≥ 7% was significantly associated with better LC and RFS. The addition of ΔADCmean significantly improved the c-indices of LC and RFS models compared with standard clinical variables (0.85 vs. 0.77 and 0.74 vs. 0.68 for LC and RFS, respectively, p < 0.0001 for both). CONCLUSION: ΔADCmean at mid-RT is a strong predictor of oncologic outcomes in HNC. Patients with no significant increase of primary tumor ADC at mid-RT are at high risk of disease relapse.

Large scale crowdsourced radiotherapy segmentations across a variety of cancer anatomic sites.

Clinician generated segmentation of tumor and healthy tissue regions of interest (ROIs) on medical images is crucial for radiotherapy. However, interobserver segmentation variability has long been considered a significant detriment to the implementation of high-quality and consistent radiotherapy dose delivery. This has prompted the increasing development of automated segmentation approaches. However, extant segmentation datasets typically only provide segmentations generated by a limited number of annotators with varying, and often unspecified, levels of expertise. In this data descriptor, numerous clinician annotators manually generated segmentations for ROIs on computed tomography images across a variety of cancer sites (breast, sarcoma, head and neck, gynecologic, gastrointestinal; one patient per cancer site) for the Contouring Collaborative for Consensus in Radiation Oncology challenge. In total, over 200 annotators (experts and non-experts) contributed using a standardized annotation platform (ProKnow). Subsequently, we converted Digital Imaging and Communications in Medicine data into Neuroimaging Informatics Technology Initiative format with standardized nomenclature for ease of use. In addition, we generated consensus segmentations for experts and non-experts using the Simultaneous Truth and Performance Level Estimation method. These standardized, structured, and easily accessible data are a valuable resource for systematically studying variability in segmentation applications.

E pluribus unum: prospective acceptability benchmarking from the Contouring Collaborative for Consensus in Radiation Oncology crowdsourced initiative for multiobserver segmentation.

Purpose: Contouring Collaborative for Consensus in Radiation Oncology (C3RO) is a crowdsourced challenge engaging radiation oncologists across various expertise levels in segmentation. An obstacle to artificial intelligence (AI) development is the paucity of multiexpert datasets; consequently, we sought to characterize whether aggregate segmentations generated from multiple nonexperts could meet or exceed recognized expert agreement. Approach: Participants who contoured ≥ 1 region of interest (ROI) for the breast, sarcoma, head and neck (H&N), gynecologic (GYN), or gastrointestinal (GI) cases were identified as a nonexpert or recognized expert. Cohort-specific ROIs were combined into single simultaneous truth and performance level estimation (STAPLE) consensus segmentations. STAPLE nonexpert ROIs were evaluated against STAPLE expert contours using Dice similarity coefficient (DSC). The expert interobserver DSC ( IODSC expert ) was calculated as an acceptability threshold between STAPLE nonexpert and STAPLE expert . To determine the number of nonexperts required to match the IODSC expert for each ROI, a single consensus contour was generated using variable numbers of nonexperts and then compared to the IODSC expert . Results: For all cases, the DSC values for STAPLE nonexpert versus STAPLE expert were higher than comparator expert IODSC expert for most ROIs. The minimum number of nonexpert segmentations needed for a consensus ROI to achieve IODSC expert acceptability criteria ranged between 2 and 4 for breast, 3 and 5 for sarcoma, 3 and 5 for H&N, 3 and 5 for GYN, and 3 for GI. Conclusions: Multiple nonexpert-generated consensus ROIs met or exceeded expert-derived acceptability thresholds. Five nonexperts could potentially generate consensus segmentations for most ROIs with performance approximating experts, suggesting nonexpert segmentations as feasible cost-effective AI inputs.

Dysphagia and shortness-of-breath as markers for treatment failure and survival in oropharyngeal cancer after radiation.

BACKGROUND: Post-treatment symptoms are a focal point of follow-up visits for head and neck cancer patients. While symptoms such as dysphagia and shortness-of-breath early after treatment may motivate additional work up, their precise association with disease control and survival outcomes is not well established. METHODS: This prospective data cohort study of 470 oropharyngeal cancer patients analyzed patient-reported swallowing, choking and shortness-of-breath symptoms at 3-to-6 months following radiotherapy to evaluate their association with overall survival and disease control. Associations between the presence of moderate-to-severe swallowing, choking and mild-to-severe shortness-of-breath and treatment outcomes were analyzed via Cox regression and Kaplan-Meier. The main outcome was overall survival (OS), and the secondary outcomes were local, regional, and distant disease control. RESULTS: The majority of patients (91.3%) were HPV-positive. Median follow-up time was 31.7 months (IQR: 21.9-42.1). Univariable analysis showed significant associations between OS and all three symptoms of swallowing, choking, and shortness-of-breath. A composite variable integrating scores of all three symptoms was significantly associated with OS on multivariable Cox regression (p = 0.0018). Additionally, this composite symptom score showed the best predictive value for OS (c-index = 0.75). Multivariable analysis also revealed that the composite score was significantly associated with local (p = 0.044) and distant (p = 0.035) recurrence/progression. Notably, the same significant associations with OS were seen for HPV-positive only subset analysis (p < 0.01 for all symptoms). CONCLUSIONS: Quantitative patient-reported measures of dysphagia and shortness-of-breath 3-to-6 months post-treatment are significant predictors of OS and disease recurrence/progression in OPC patients and in HPV-positive OPC only.

Quality assurance assessment of intra-acquisition diffusion-weighted and T2-weighted magnetic resonance imaging registration and contour propagation for head and neck cancer radiotherapy.

BACKGROUND/PURPOSE: Adequate image registration of anatomical and functional magnetic resonance imaging (MRI) scans is necessary for MR-guided head and neck cancer (HNC) adaptive radiotherapy planning. Despite the quantitative capabilities of diffusion-weighted imaging (DWI) MRI for treatment plan adaptation, geometric distortion remains a considerable limitation. Therefore, we systematically investigated various deformable image registration (DIR) methods to co-register DWI and T2-weighted (T2W) images. MATERIALS/METHODS: We compared three commercial (ADMIRE, Velocity, Raystation) and three open-source (Elastix with default settings [Elastix Default], Elastix with parameter set 23 [Elastix 23], Demons) post-acquisition DIR methods applied to T2W and DWI MRI images acquired during the same imaging session in twenty immobilized HNC patients. In addition, we used the non-registered images (None) as a control comparator. Ground-truth segmentations of radiotherapy structures (tumour and organs at risk) were generated by a physician expert on both image sequences. For each registration approach, structures were propagated from T2W to DWI images. These propagated structures were then compared with ground-truth DWI structures using the Dice similarity coefficient and mean surface distance. RESULTS: 19 left submandibular glands, 18 right submandibular glands, 20 left parotid glands, 20 right parotid glands, 20 spinal cords, and 12 tumours were delineated. Most DIR methods took <30 s to execute per case, with the exception of Elastix 23 which took ∼458 s to execute per case. ADMIRE and Elastix 23 demonstrated improved performance over None for all metrics and structures (Bonferroni-corrected p < 0.05), while the other methods did not. Moreover, ADMIRE and Elastix 23 significantly improved performance in individual and pooled analysis compared to all other methods. CONCLUSIONS: The ADMIRE DIR method offers improved geometric performance with reasonable execution time so should be favoured for registering T2W and DWI images acquired during the same scan session in HNC patients. These results are important to ensure the appropriate selection of registration strategies for MR-guided radiotherapy.

Optimized decision support for selection of transoral robotic surgery or (chemo)radiation therapy based on posttreatment swallowing toxicity.

BACKGROUND: A primary goal in transoral robotic surgery (TORS) for oropharyngeal squamous cell cancer (OPSCC) survivors is to optimize swallowing function. However, the uncertainty in the outcomes of TORS including postoperative residual positive margin (PM) and extranodal extension (ENE), may necessitate adjuvant therapy, which may cause significant swallowing toxicity to survivors. METHODS: A secondary analysis was performed on a prospective registry data with low- to intermediate-risk human papillomavirus-related OPSCC possibly resectable by TORS. Decision trees were developed to model the uncertainties in TORS compared with definitive radiation therapy (RT) and chemoradiation therapy (CRT). Swallowing toxicities were measured by Dynamic Imaging Grade of Swallowing Toxicity (DIGEST), MD Anderson Dysphagia Inventory (MDADI), and the MD Anderson Symptom Inventory-Head and Neck (MDASI-HN) instruments. The likelihoods of PM/ENE were varied to determine the thresholds within which each therapy remains optimal. RESULTS: Compared with RT, TORS resulted in inferior swallowing function for moderate likelihoods of PM/ENE (>60% in short term for all instruments, >75% in long term for DIGEST and MDASI) leaving RT as the optimal treatment. Compared with CRT, TORS remained the optimal therapy based on MDADI and MDASI but showed inferior swallowing outcomes based on DIGEST for moderate-to-high likelihoods of PM/ENE (>75% for short-term and >40% for long-term outcomes). CONCLUSION: In the absence of reliable estimation of postoperative PM/ENE concurrent with significant postoperative PM, the overall toxicity level in OPSCC patients undergoing TORS with adjuvant therapy may become more severe compared with patients receiving nonsurgical treatments thus advocating definitive (C)RT protocols.

Temporal characterization of acute pain and toxicity kinetics during radiation therapy for head and neck cancer. A retrospective study.

Objectives: Pain during Radiation Therapy (RT) for oral cavity/oropharyngeal cancer (OC/OPC) is a clinical challenge due to its multifactorial etiology and variable management. The objective of this study was to define complex pain profiles through temporal characterization of pain descriptors, physiologic state, and RT-induced toxicities for pain trajectories understanding. Materials and methods: Using an electronic health record registry, 351 OC/OPC patients treated with RT from 2013 to 2021 were included. Weekly numeric scale pain scores, pain descriptors, vital signs, physician-reported toxicities, and analgesics were analyzed using linear mixed effect models and Spearman's correlation. Area under the pain curve (AUCpain) was calculated to measure pain burden over time. Results: Median pain scores increased from 0 during the weekly visit (WSV)-1 to 5 during WSV-7. By WSV-7, 60% and 74% of patients reported mouth and throat pain, respectively, with a median pain score of 5. Soreness and burning pain peaked during WSV-6/7 (51%). Median AUCpain was 16% (IQR (9.3-23)), and AUCpain significantly varied based on gender, tumor site, surgery, drug use history, and pre-RT pain. A temporal increase in mucositis and dermatitis, declining mean bodyweight (-7.1%; P < 0.001) and mean arterial pressure (MAP) 6.8 mmHg; P < 0.001 were detected. Pulse rate was positively associated while weight and MAP were negatively associated with pain over time (P < 0.001). Conclusion: This study provides insight on in-depth characterization and associations between dynamic pain, physiologic, and toxicity kinetics. Our findings support further needs of optimized pain control through temporal data-driven clinical decision support systems for acute pain management.

Improving Prediction of Late Symptoms using LSTM and Patient-reported Outcomes for Head and Neck Cancer Patients.

Patient-Reported Outcomes (PRO) are collected directly from the patients using symptom questionnaires. In the case of head and neck cancer patients, PRO surveys are recorded every week during treatment with each patient's visit to the clinic and at different follow-up times after the treatment has concluded. PRO surveys can be very informative regarding the patient's status and the effect of treatment on the patient's quality of life (QoL). Processing PRO data is challenging for several reasons. First, missing data is frequent as patients might skip a question or a questionnaire altogether. Second, PROs are patient-dependent, a rating of 5 for one patient might be a rating of 10 for another patient. Finally, most patients experience severe symptoms during treatment which usually subside over time. However, for some patients, late toxicities persist negatively affecting the patient's QoL. These long-term severe symptoms are hard to predict and are the focus of this study. In this work, we model PRO data collected from head and neck cancer patients treated at the MD Anderson Cancer Center using the MD Anderson Symptom Inventory (MDASI) questionnaire as time series. We impute missing values with a combination of K nearest neighbor (KNN) and Long Short-Term Memory (LSTM) neural networks, and finally, apply LSTM to predict late symptom severity 12 months after treatment. We compare performance against clinical and ARIMA models. We show that the LSTM model combined with KNN imputation is effective in predicting late-stage symptom ratings for occurrence and severity under the AUC and F1 score metrics.

Deep-learning-based generation of synthetic 6-minute MRI from 2-minute MRI for use in head and neck cancer radiotherapy.

Background: Quick magnetic resonance imaging (MRI) scans with low contrast-to-noise ratio are typically acquired for daily MRI-guided radiotherapy setup. However, for patients with head and neck (HN) cancer, these images are often insufficient for discriminating target volumes and organs at risk (OARs). In this study, we investigated a deep learning (DL) approach to generate high-quality synthetic images from low-quality images. Methods: We used 108 unique HN image sets of paired 2-minute T2-weighted scans (2mMRI) and 6-minute T2-weighted scans (6mMRI). 90 image sets (~20,000 slices) were used to train a 2-dimensional generative adversarial DL model that utilized 2mMRI as input and 6mMRI as output. Eighteen image sets were used to test model performance. Similarity metrics, including the mean squared error (MSE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR) were calculated between normalized synthetic 6mMRI and ground-truth 6mMRI for all test cases. In addition, a previously trained OAR DL auto-segmentation model was used to segment the right parotid gland, left parotid gland, and mandible on all test case images. Dice similarity coefficients (DSC) were calculated between 2mMRI and either ground-truth 6mMRI or synthetic 6mMRI for each OAR; two one-sided t-tests were applied between the ground-truth and synthetic 6mMRI to determine equivalence. Finally, a visual Turing test using paired ground-truth and synthetic 6mMRI was performed using three clinician observers; the percentage of images that were correctly identified was compared to random chance using proportion equivalence tests. Results: The median similarity metrics across the whole images were 0.19, 0.93, and 33.14 for MSE, SSIM, and PSNR, respectively. The median of DSCs comparing ground-truth vs. synthetic 6mMRI auto-segmented OARs were 0.86 vs. 0.85, 0.84 vs. 0.84, and 0.82 vs. 0.85 for the right parotid gland, left parotid gland, and mandible, respectively (equivalence p<0.05 for all OARs). The percent of images correctly identified was equivalent to chance (p<0.05 for all observers). Conclusions: Using 2mMRI inputs, we demonstrate that DL-generated synthetic 6mMRI outputs have high similarity to ground-truth 6mMRI, but further improvements can be made. Our study facilitates the clinical incorporation of synthetic MRI in MRI-guided radiotherapy.