Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “regression”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,405 records · Page 78
All sourcesNASA NTRSDOE OSTI
DOE OSTI2014

Outcomes and Prognostic Factors in Women With 1 to 3 Breast Cancer Brain Metastases Treated With Definitive Stereotactic Radiosurgery

Background: With the continuing increase in the use of definitive stereotactic radiosurgery (SRS) for patients with limited brain metastases (BM), clinicians need more specific prognostic tools. We investigated clinical predictors of outcomes in patients with limited breast cancer BM treated with SRS alone. Methods and Materials: We identified 136 patients with breast cancer and 1-3 BM who underwent definitive SRS for 186 BM between 2000 and 2012. The Kaplan-Meier method was used to assess overall survival (OS), regional failure (RF), and local failure (LF). Associations between clinical factors and outcomes were tested using Cox regression. A point scoring system was used to stratify patients based on OS, and the predictive power was tested with concordance probability estimate (CPE). Results: The median OS was 17.6 months. The 12-month RF and LF rates were 45% and 10%, respectively. On multivariate analysis, >1 lesion (hazard ratio [HR] = 1.6, P=.02), triple-negative (TN) disease (HR=2.0, P=.006), and active extracranial disease (ED) (HR=2.7, P<.0001) were significantly associated with worse OS. The point score system was defined using proportional simplification of the multivariate Cox proportional hazards regression function. The median OS for patients with 3.0-4.0 points (n=37), 4.5-5.5 points (n=28), 6.0-6.5 points (n=37), and 8-8.5 points (n=34) were 9.2, 15.6, 25.1, and 45.1 months, respectively (P<.0001, CPE = 0.72). Active ED (HR=2.4, P=.0007) was significantly associated with RF. Higher risk for LF was significantly associated with larger BM size (HR=3.1, P=.0001). Conclusion: Patients with >1 BM, active ED, and TN had the highest risk of death after SRS. Active ED is an important prognostic factor for OS and intracranial control.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Gastrointestinal Dose-Histogram Effects in the Context of Dose-Volume–Constrained Prostate Radiation Therapy: Analysis of Data From the RADAR Prostate Radiation Therapy Trial

Purpose: To use a high-quality multicenter trial dataset to determine dose-volume effects for gastrointestinal (GI) toxicity following radiation therapy for prostate carcinoma. Influential dose-volume histogram regions were to be determined as functions of dose, anatomical location, toxicity, and clinical endpoint. Methods and Materials: Planning datasets for 754 participants in the TROG 03.04 RADAR trial were available, with Late Effects of Normal Tissues (LENT) Subjective, Objective, Management, and Analytic (SOMA) toxicity assessment to a median of 72 months. A rank sum method was used to define dose-volume cut-points as near-continuous functions of dose to 3 GI anatomical regions, together with a comprehensive assessment of significance. Univariate and multivariate ordinal regression was used to assess the importance of cut-points at each dose. Results: Dose ranges providing significant cut-points tended to be consistent with those showing significant univariate regression odds-ratios (representing the probability of a unitary increase in toxicity grade per percent relative volume). Ranges of significant cut-points for rectal bleeding validated previously published results. Separation of the lower GI anatomy into complete anorectum, rectum, and anal canal showed the impact of mid-low doses to the anal canal on urgency and tenesmus, completeness of evacuation and stool frequency, and mid-high doses to the anorectum on bleeding and stool frequency. Derived multivariate models emphasized the importance of the high-dose region of the anorectum and rectum for rectal bleeding and mid- to low-dose regions for diarrhea and urgency and tenesmus, and low-to-mid doses to the anal canal for stool frequency, diarrhea, evacuation, and bleeding. Conclusions: Results confirm anatomical dependence of specific GI toxicities. They provide an atlas summarizing dose-histogram effects and derived constraints as functions of anatomical region, dose, toxicity, and endpoint for informing future radiation therapy planning.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Total Laryngectomy Versus Larynx Preservation for T4a Larynx Cancer: Patterns of Care and Survival Outcomes

Purpose: To examine practice patterns and compare survival outcomes between total laryngectomy (TL) and larynx preservation chemoradiation (LP-CRT) in the setting of T4a larynx cancer, using a large national cancer registry. Methods and Materials: Using the National Cancer Database, we identified 969 patients from 2003 to 2006 with T4a squamous cell larynx cancer receiving definitive treatment with either initial TL plus adjuvant therapy or LP-CRT. Univariate and multivariable logistic regression were used to assess predictors of undergoing surgery. Survival outcomes were compared using Kaplan-Meier and propensity score–adjusted and inverse probability of treatment–weighted Cox proportional hazards methods. Sensitivity analyses were performed to account for unmeasured confounders. Results: A total of 616 patients (64%) received LP-CRT, and 353 (36%) received TL. On multivariable logistic regression, patients with advanced nodal disease were less likely to receive TL (N2 vs N0, 26.6% vs 43.4%, odds ratio [OR] 0.52, 95% confidence interval [CI] 0.37-0.73; N3 vs N0, 19.1% vs 43.4%, OR 0.23, 95% CI 0.07-0.77), whereas patients treated in high case-volume facilities were more likely to receive TL (46.1% vs 31.5%, OR 1.78, 95% CI 1.27-2.48). Median survival for TL versus LP was 61 versus 39 months (P<.001). After controlling for potential confounders, LP-CRT had inferior overall survival compared with TL (hazard ratio 1.31, 95% CI 1.10-1.57), and with the inverse probability of treatment–weighted model (hazard ratio 1.25, 95% CI 1.05-1.49). This survival difference was shown to be robust on additional sensitivity analyses. Conclusions: Most patients with T4a larynx cancer receive LP-CRT, despite guidelines suggesting TL as the preferred initial approach. Patients receiving LP-CRT had more advanced nodal disease and worse overall survival. Previous studies of (non-T4a) locally advanced larynx cancer showing no difference in survival between LP-CRT and TL may not apply to T4a disease, and patients should be counseled accordingly.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Competing Risk Analysis of Neurologic versus Nonneurologic Death in Patients Undergoing Radiosurgical Salvage After Whole-Brain Radiation Therapy Failure: Who Actually Dies of Their Brain Metastases?

Purpose: To estimate the hazard for neurologic (central nervous system, CNS) and nonneurologic (non-CNS) death associated with patient, treatment, and systemic disease status in patients receiving stereotactic radiosurgery after whole-brain radiation therapy (WBRT) failure, using a competing risk model. Patients and Methods: Of 757 patients, 293 experienced recurrence or new metastasis following WBRT. Univariate Cox proportional hazards regression identified covariates for consideration in the multivariate model. Competing risks multivariable regression was performed to estimate the adjusted hazard ratio (aHR) and 95% confidence interval (CI) for both CNS and non-CNS death after adjusting for patient, disease, and treatment factors. The resultant model was converted into an online calculator for ease of clinical use. Results: The cumulative incidence of CNS and non-CNS death at 6 and 12 months was 20.6% and 21.6%, and 34.4% and 35%, respectively. Patients with melanoma histology (relative to breast) (aHR 2.7, 95% CI 1.5-5.0), brainstem location (aHR 2.1, 95% CI 1.3-3.5), and number of metastases (aHR 1.09, 95% CI 1.04-1.2) had increased aHR for CNS death. Progressive systemic disease (aHR 0.55, 95% CI 0.4-0.8) and increasing lowest margin dose (aHR 0.97, 95% CI 0.9-0.99) were protective against CNS death. Patients with lung histology (aHR 1.3, 95% CI 1.1-1.9) and progressive systemic disease (aHR 2.14, 95% CI 1.5-3.0) had increased aHR for non-CNS death. Conclusion: Our nomogram provides individual estimates of neurologic death after salvage stereotactic radiosurgery for patients who have failed prior WBRT, based on histology, neuroanatomical location, age, lowest margin dose, and number of metastases after adjusting for their competing risk of death from other causes.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

A Quantitative Analysis of the Relationship Between Radiation Therapy Use and Travel Time

Purpose: To model and quantify the relationship between radiation therapy (RT) use and travel time to RT services. Methods and Materials: Population-based registries and databases were used to identify both incident cancer patient and patients receiving RT within 1 year of diagnosis (RT1y) in British Columbia, Canada, between 1992 and 2011. The effects of age, gender, diagnosis year, income, prevailing wait time, and travel duration for RT on RT1y were assessed. Significant factors from univariate analyses were included in a multivariable logistic regression model. The shape of the travel time–RT1y curve was represented by generalized additive and segmented regression models. Analyses were conducted for breast, lung, and genitourinary cancer separately and for all cancer sites combined. Results: After adjustment for age, gender, diagnosis year, income, and prevailing wait times, increasing travel time to the closest RT facility had a negative impact RT1y. The shape of the travel time–RT1y curve varied with cancer type. For breast cancer, the odds of RT1y were constant for the first 2 driving hours and decreased at 17% per hour thereafter. For lung cancer, the odds of RT1y decreased by 16% after 20 minutes and then decreased at 6% per hour. Genitourinary cancer RT1y was relatively independent of travel time. For all cancer sites combined, the odds of RT1y were constant within the first 2 driving hours and decreased at 7% per hour thereafter. Conclusions: Travel time to receive RT has a different impact on RT1y for different tumor sites. The results provide evidence-based insights for the configuration of catchment areas for new and existing cancer centers providing RT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Impact of Insurance Status on Radiation Treatment Modality Selection Among Potential Candidates for Prostate, Breast, or Gynecologic Brachytherapy

Purpose: The Patient Protection and Affordable Care Act looks to expand both private and Medicaid insurance. To evaluate how these changes may affect the field of radiation oncology, we evaluated the association of insurance status with the use of brachytherapy in cancers for which this treatment technique is used. Methods and Materials: A total of 190,467 patients met the inclusion criteria, of whom 95,292 (50.0%) had breast cancer, 61,096 (32.1%) had prostate cancer, 28,194 (14.8%) had endometrial cancer, and 5885 (3.1%) had cervical cancer. A multivariate logistic regression model was used to determine the association between insurance status and receipt of brachytherapy among patients treated definitively for prostate and cervical cancer or postoperatively for breast and endometrial cancer. Results: The rates of non-Medicaid insurance were 49.9% (cervical), 85.3% (endometrial), 87.4% (breast), and 90.9% (prostate) (P<.001). In a logistic regression, patients who received radiation therapy were less likely to receive brachytherapy if they had Medicaid coverage (odds ratio [OR] 0.57, 95% confidence interval [CI] 0.53-0.61, P<.001) or did not have insurance coverage (OR 0.50, 95% CI 0.45-0.56, P<.001) compared with those with non-Medicaid insurance. On subset analysis, patients with Medicaid coverage or without insurance coverage were significantly less likely to receive brachytherapy than were those with non-Medicaid insurance for all 4 sites, except for patients with endometrial cancer. Conclusions: Despite being a cost-effective treatment modality, brachytherapy is less often used in the definitive or postoperative management of cancer in patients with Medicaid coverage or without insurance. Upcoming health policy changes resulting in the expansion of private insurance and Medicaid will likely increase access to and demand for brachytherapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2015

Prospective, Multicenter Validation Study of Magnetic Resonance Volumetry for Response Assessment After Preoperative Chemoradiation in Rectal Cancer: Can the Results in the Literature be Reproduced?

Purpose: To review the available literature on tumor size/volume measurements on magnetic resonance imaging for response assessment after chemoradiotherapy, and validate these cut-offs in an independent multicenter patient cohort. Methods and Materials: The study included 2 parts. (1) Review of the literature: articles were included that assessed the accuracy of tumor size/volume measurements on magnetic resonance imaging for tumor response assessment. Size/volume cut-offs were extracted; (2) Multicenter validation: extracted cut-offs from the literature were tested in a multicenter cohort (n=146). Accuracies were calculated and compared with reported results from the literature. Results: The review included 14 articles, in which 3 different measurement methods were assessed: (1) tumor length; (2) 3-dimensonial tumor size; and (3) whole volume. Study outcomes consisted of (1) complete response (ypT0) versus residual tumor; (2) tumor regression grade 1 to 2 versus 3 to 5; and (3) T-downstaging (ypT<cT). In the multicenter cohort, best results were obtained for the validation of the whole-volume measurements, in particular for the outcome ypT0 (accuracy 44%-80%), with the optimal cut-offs being 1.6 cm{sup 3} (after chemoradiation therapy) and a volume reduction of Δ80% to 86.6%. Accuracies for whole-volume measurements to assess tumor regression grade 1 to 2 were 52% to 61%, and for T-downstaging 51% to 57%. Overall accuracies for tumor length ranged between 48% and 53% and for 3D size measurement between 52% and 56%. Conclusions: Magnetic resonance volumetry using whole-tumor volume measurements can be helpful in rectal cancer response assessment with selected cut-off values. Measurements of tumor length or 3-dimensional tumor size are not helpful. Magnetic resonance volumetry is mainly accurate to assess a complete tumor response (ypT0) after chemoradiation therapy (accuracies up to 80%).

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2016

Long-Term Bone Marrow Suppression During Postoperative Chemotherapy in Rectal Cancer Patients After Preoperative Chemoradiation Therapy

Purpose/Objective(s): To quantify ensuing bone marrow (BM) suppression during postoperative chemotherapy resulting from preoperative chemoradiation (CRT) therapy for rectal cancer. Methods and Materials: We retrospectively evaluated 35 patients treated with preoperative CRT followed by postoperative 5-Fluorouracil and oxaliplatin (OxF) chemotherapy for locally advanced rectal cancer. The pelvic bone marrow (PBM) was divided into ilium (IBM), lower pelvis (LPBM), and lumbosacrum (LSBM). Dose volume histograms (DVH) measured the mean doses and percentage of BM volume receiving between 5-40 Gy (i.e.: PBM-V5, LPBM-V5). The Wilcoxon signed rank tests evaluated the differences in absolute hematologic nadirs during neoadjuvant vs. adjuvant treatment. Logistic regressions evaluated the association between dosimetric parameters and ≥ grade 3 hematologic toxicity (HT3) and hematologic event (HE) defined as ≥ grade 2 HT and a dose reduction in OxF. Receiver Operator Characteristic (ROC) curves were constructed to determine optimal threshold values leading to HT3. Results: During OxF chemotherapy, 40.0% (n=14) and 48% (n=17) of rectal cancer patients experienced HT3 and HE, respectively. On multivariable logistic regression, increasing pelvic mean dose (PMD) and lower pelvis mean dose (LPMD) along with increasing PBM-V (25-40), LPBM-V25, and LPBM-V40 were significantly associated with HT3 and/or HE during postoperative chemotherapy. Exceeding ≥36.6 Gy to the PMD and ≥32.6 Gy to the LPMD strongly correlated with causing HT3 during postoperative chemotherapy. Conclusions: Neoadjuvant RT for rectal cancer has lasting effects on the pelvic BM, which are demonstrable during adjuvant OxF. Sparing of the BM during preoperative CRT can aid in reducing significant hematologic adverse events and aid in tolerance of postoperative chemotherapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

Patterns of Care in Proton Radiation Therapy for Pediatric Central Nervous System Malignancies

Purpose: Proton beam therapy (PBT) potentially allows for improved sparing of normal tissues, hopefully leading to decreased late side effects in children. Using a national registry, we sought to perform a patterns-of-care analysis for children receiving PBT for primary malignancies of the central nervous system (CNS). Methods and Materials: Using the National Cancer Data Base, we identified pediatric patients with primary CNS malignancies that were diagnosed between 2004 and 2012. We used a standard t test for comparison of means and χ{sup 2} testing to identify differences in demographic and clinical characteristics. Univariate and multivariate logistical regression was applied to identify predictors of PBT use. Results: We identified 4637 pediatric patients receiving radiation therapy from 2004 to 2012, including a subset of 267 patients treated with PBT. We found that PBT use increased with time from <1% in 2004 to 15% in 2012. In multivariate logistical regression, we found the following to be predictors of receipt of PBT: private insurance, the highest income bracket, younger age, living in a metropolitan area, and residing >200 miles from a radiation treatment facility (P<.05). Conclusions: We noted the proportion of children receiving PBT to be significantly increasing over time from <1% to 15% from 2004 to 2012. We also observed important disparities in receipt of PBT based on socioeconomic status. Children from higher-income households and with private insurance were more likely to use this expensive technology. As we continue to demonstrate the potential benefits of PBT in children, efforts are needed to expand the accessibility of PBT for children of all socioeconomic backgrounds and regions of the country.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

A Comparison Between Low-Dose-Rate Brachytherapy With or Without Androgen Deprivation, External Beam Radiation Therapy With or Without Androgen Deprivation, and Radical Prostatectomy With or Without Adjuvant or Salvage Radiation Therapy for High-Risk Prostate Cancer

Purpose: We compare the efficacy and toxicity among the 3 major modalities available used to treat high-risk prostate cancer (HRCaP). Methods and Materials: From 1996 to 2012, 2557 HRCaP patients were treated: 734 received external beam radiation therapy (EBRT) with or without androgen deprivation therapy (ADT), 515 received low-dose-rate prostate brachytherapy (LDR) with or without ADT, and 1308 received radical prostatectomy (RP) with or without EBRT. Biochemical relapse-free survival (bRFS), clinical relapse-free survival (cRFS), and prostate cancer–specific mortality (PCSM) were assessed. Toxicity was assessed using the Common Terminology Criteria for Adverse Events, version 4.03. The log-rank test compared bRFS and cRFS among the modalities, and Cox regression identified factors associated with bRFS and cRFS. Gray's test compared differences in late toxicity and PSCM among the modalities. Competing risk regression identified factors associated with PCSM. Results: The median follow-up time and age were 63.5 months and 65 years, respectively. The bRFS at 5 and 10 years, respectively, was 74% and 53% for EBRT, 74% and 52% for LDR, and 65% and 47% for RP (P=.0001). The cRFS at 5 and 10 years, respectively, was 85% and 73% for EBRT, 90% and 76% for LDR, and 89% and 75% for RP (P=.121). The PCSM at 5 and 10 years, respectively, was 5.3% and 11.2% for EBRT, 3.2% and 3.6% for LDR, and 2.8% and 6.8% for RP (P=.0004). The 10-year cumulative incidence of ≥grade 3 genitourinary toxicity was 8.1% for EBRT, 7.2% for LDR, and 16.4% for RP (P<.0001). The 10-year cumulative incidence of ≥grade 3 gastrointestinal toxicity was 4.6% for EBRT, 1.1% for LDR, and 1.0% for RP (P<.0001). Conclusion: HRCaP treated with EBRT, LDR, or RP yields efficacy showing better bRFS for LDR and EBRT relative to RP, equivalence for cRFS, and a PCSM advantage of LDR and RP over EBRT. The toxicity is lowest for LDR.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

Single Nucleotide Polymorphism TGFβ1 R25P Correlates with Acute Toxicity during Neoadjuvant Chemoradiotherapy in Rectal Cancer Patients

Purpose: To validate the finding of an association between single nucleotide polymorphisms (SNPs) and toxicity during chemoradiotherapy (CRT) in rectal cancer patients, in an independent population. Methods and Materials: The cohort consisted of 165 patients who received CRT for rectal cancer from 2006 to 2012. Prospectively recorded toxicity information, graded according to the Common Terminology Criteria for Adverse Events version 3.0, was retrieved from the medical record. Additionally, a subset of 52 patients recorded their gastrointestinal symptoms weekly during CRT, using the 7-item Bowel Problems Scale. Deoxyribonucleic acid was extracted from normal tissue in the proctectomy specimens and screened for 3 SNPs: XRCC1 R399Q, XPD K751Q, and TGFβ1 R25P. Univariable and multivariable logistic regression models were constructed. Results: The median radiation dose was 50.4 Gy, and all patients received concurrent chemotherapy. Toxicities measured by the Common Terminology Criteria for Adverse Events were closely associated with patient-reported outcomes for the patients who completed the 7-item Bowel Problems Scale. Grade ≥3 toxicity occurred during CRT in 14 patients (8%). All 14 patients had either XRCC1 R399Q or TGFβ1 R25P polymorphisms. The TGFβ1 R25P polymorphism was significantly associated with grade ≥3 toxicity (odds ratio [OR] 3.47, P=.04) and, in patients who completed the Bowel Problems Scale, with grade ≥4 toxicity (OR 5.61, P=.02). The latter finding persisted in a multivariable logistic regression model controlling for ethnicity, age, and sex (adjusted OR 1.83, P=.02). Conclusions: We have validated the correlation between the TGFβ1 R25P SNP and acute toxicity during CRT in an independent cohort using both clinician- and patient-reported toxicity. The information from our study could be used as a basis to formulate a prospective trial testing the utility of this SNP as a biomarker of acute toxicity during neoadjuvant treatment in locally advanced rectal cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

Brachytherapy Improves Survival in Stage III Endometrial Cancer With Cervical Involvement

Purpose: To evaluate the survival benefit of adding vaginal brachytherapy (BT) to pelvic external beam radiation therapy (EBRT) in women with stage III endometrial cancer. Methods and Materials: The National Cancer Data Base was used to identify patients with stage III endometrial cancer from 2004 to 2013. Only women who received adjuvant EBRT were analyzed. Women were grouped according to receipt of BT. Logistic regression modeling was used to identify predictors of receiving BT. Log–rank statistics were used to compare survival outcomes. Cox proportional hazards modeling was used to evaluate the effect of BT on survival. A propensity score–matched analysis was also conducted among women with cervical involvement. Results: We evaluated 12,988 patients with stage III endometrial carcinoma, 39% of whom received EBRT plus BT. Women who received BT were more likely to have endocervical or cervical stromal involvement (odds ratios 2.03 and 1.77; P<.01, respectively). For patients receiving EBRT alone, the 5-year survival was 66% versus 69% with the addition of BT at 5 years (P<.01). Brachytherapy remained significantly predictive of decreased risk of death (hazard ratio 0.86; P<.01) on multivariate Cox regression. The addition of BT to EBRT did not affect survival among women without cervical involvement (P=.84). For women with endocervical or cervical stromal invasion, the addition of BT significantly improved survival (log–rank P<.01). Receipt of EBRT plus BT was associated with improved survival in women with positive and negative surgical margins, and receiving chemotherapy did not alter the benefit of BT. Propensity score–matched analysis results confirmed the benefit of BT among women with cervical involvement (hazard ratio 0.80; P=.01). Conclusions: In this population of women with stage III endometrial cancer the addition of BT to EBRT was associated with an improvement in survival for women with endocervical or cervical stromal invasion.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

Predictors of Radiation Therapy–Related Gastrointestinal Toxicity From Anal Cancer Dose-Painted Intensity Modulated Radiation Therapy: Secondary Analysis of NRG Oncology RTOG 0529

Purpose: NRG Oncology RTOG 0529 assessed the feasibility of dose-painted intensity modulated radiation therapy (DP-IMRT) to reduce the acute morbidity of chemoradiation with 5-fluorouracil (5FU) and mitomycin-C (MMC) for T2-4N0-3M0 anal cancer. This secondary analysis was performed to identify patient and treatment factors associated with acute and late gastrointestinal (GI) adverse events (AEs). Methods and Materials: NRG Oncology RTOG 0529 treatment plans were reviewed to extract dose-volume data for tightly contoured small bowel, loosely contoured anterior pelvic contents (APC), and uninvolved colon outside the target volume (UC). Univariate logistic regression was performed to evaluate association between volumes of each structure receiving doses ≥5 to 60 Gy (V5-V60) in 5-Gy increments between patients with and without grade ≥2 acute and late GI AEs, and grade ≥3 acute GI AEs. Additional patient and treatment factors were evaluated in multivariate logistic regression (acute AEs) or Cox proportional hazards models (late AEs). Results: Among 52 evaluable patients, grade ≥2 acute, grade ≥2 late, and grade ≥3 acute GI AEs were observed in 35, 17, and 10 patients, respectively. Trends (P<.05) toward statistically significant associations were observed between grade ≥2 acute GI AEs and small bowel dose (V20-V40), grade ≥2 late GI AEs and APC dose (V60), grade ≥3 acute GI AEs and APC dose (V5-V25), increasing age, tumor size >4 cm, and worse Zubrod performance status. Small bowel volumes of 186.0 cc, 155.0 cc, 41.0 cc, and 30.4 cc receiving doses greater than 25, 30, 35, and 40 Gy, respectively, correlated with increased risk of acute grade ≥2 GI AEs. Conclusions: Acute and late GI AEs from 5FU/MMC chemoradiation using DP-IMRT correlate with radiation dose to the small bowel and APC. Such associations will be incorporated in the dose-volume normal tissue constraint design for future NRG oncology anal cancer studies.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

National Trends and Predictors of Androgen Deprivation Therapy Use in Low-Risk Prostate Cancer

Purpose: Androgen deprivation therapy (ADT) is not recommended for low-risk prostate cancer because of its lack of benefit and potential for harm. We evaluated the incidence and predictors of ADT use in low-risk disease. Methods and Materials: Using the National Cancer Database, we identified 197,957 patients with low-risk prostate cancer (Gleason score of 3 + 3 = 6, prostate-specific antigen level <10 ng/mL, and cT1-T2a) diagnosed from 2004 to 2012 with complete demographic and treatment information. We used multiple logistic regression to evaluate predictors of ADT use and Cox regression to examine its association with all-cause mortality. Results: Overall ADT use decreased from 17.6% in 2004 to 3.5% in 2012. In 2012, 11.5% of low-risk brachytherapy patients and 7.6% of external beam radiation therapy patients received ADT. Among 82,352 irradiation-managed patients, predictors of ADT use included treatment in a community versus academic cancer program (adjusted odds ratio [AOR], 1.60; 95% confidence interval [CI], 1.50-1.71; P<.001; incidence, 14.0% vs 6.0% in 2012); treatment in the South (AOR, 1.51), Midwest (AOR, 1.81), or Northeast (AOR, 1.90) versus West (P<.001); and brachytherapy use versus external beam radiation therapy (AOR, 1.32; 95% CI, 1.27-1.37; P<.001). Among 25,196 patients who did not receive local therapy, predictors of primary ADT use included a Charlson-Deyo comorbidity score of ≥2 versus 0 (AOR, 1.42; 95% CI, 1.06-1.91; P=.018); treatment in a community versus academic cancer program (AOR, 1.61; 95% CI, 1.37-1.90; P<.001); and treatment in the South (AOR, 1.26), Midwest (AOR, 1.52), or Northeast (AOR, 1.28) versus West (P≤.008). Primary ADT use was associated with increased all-cause mortality in patients who did not receive local therapy (adjusted hazard ratio, 1.28; 95% CI, 1.14-1.43; P<.001) after adjustment for age and comorbidity. Conclusions: ADT use in low-risk prostate cancer has declined nationally but may remain an issue of concern in certain populations and regions.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2021

Bubbles in US gasoline prices: Assessing the role of hurricanes and anti–price gouging laws

This paper examines the presence of bubbles — price changes that cannot be attributed to market fundamentals — in weekly US retail gasoline prices for ten US cities, and whether these bubbles are affected by hurricane events and anti–price gouging laws. Rolling-window, right-tailed unit root tests were used to identify and date bubbles. The analysis found 6 to 22 bubbles, lasting 4 weeks or longer, in city-level retail gasoline prices between 2000 and 2017. A linear regression model of the adjustment speed coefficients from the bubble identification step found that hurricane events increase the rate of price explosivity (or slow the rate of return to equilibrium) following a shock. Furthermore, the regression estimates also imply that state anti–price gouging laws can counteract these effects.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2014

Selection of polynomial chaos bases via Bayesian model uncertainty methods with applications to sparse approximation of PDEs with stochastic inputs

Generalized polynomial chaos (gPC) expansions allow us to represent the solution of a stochastic system using a series of polynomial chaos basis functions. The number of gPC terms increases dramatically as the dimension of the random input variables increases. When the number of the gPC terms is larger than that of the available samples, a scenario that often occurs when the corresponding deterministic solver is computationally expensive, evaluation of the gPC expansion can be inaccurate due to over-fitting. We propose a fully Bayesian approach that allows for global recovery of the stochastic solutions, in both spatial and random domains, by coupling Bayesian model uncertainty and regularization regression methods. It allows the evaluation of the PC coefficients on a grid of spatial points, via (1) the Bayesian model average (BMA) or (2) the median probability model, and their construction as spatial functions on the spatial domain via spline interpolation. The former accounts for the model uncertainty and provides Bayes-optimal predictions; while the latter provides a sparse representation of the stochastic solutions by evaluating the expansion on a subset of dominating gPC bases. Moreover, the proposed methods quantify the importance of the gPC bases in the probabilistic sense through inclusion probabilities. We design a Markov chain Monte Carlo (MCMC) sampler that evaluates all the unknown quantities without the need of ad-hoc techniques. The proposed methods are suitable for, but not restricted to, problems whose stochastic solutions are sparse in the stochastic space with respect to the gPC bases while the deterministic solver involved is expensive. We demonstrate the accuracy and performance of the proposed methods and make comparisons with other approaches on solving elliptic SPDEs with 1-, 14- and 40-random dimensions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials

We present a new interatomic potential for solids and liquids called Spectral Neighbor Analysis Potential (SNAP). The SNAP potential has a very general form and uses machine-learning techniques to reproduce the energies, forces, and stress tensors of a large set of small configurations of atoms, which are obtained using high-accuracy quantum electronic structure (QM) calculations. The local environment of each atom is characterized by a set of bispectrum components of the local neighbor density projected onto a basis of hyperspherical harmonics in four dimensions. The bispectrum components are the same bond-orientational order parameters employed by the GAP potential [1]. The SNAP potential, unlike GAP, assumes a linear relationship between atom energy and bispectrum components. The linear SNAP coefficients are determined using weighted least-squares linear regression against the full QM training set. This allows the SNAP potential to be fit in a robust, automated manner to large QM data sets using many bispectrum components. The calculation of the bispectrum components and the SNAP potential are implemented in the LAMMPS parallel molecular dynamics code. We demonstrate that a previously unnoticed symmetry property can be exploited to reduce the computational cost of the force calculations by more than one order of magnitude. We present results for a SNAP potential for tantalum, showing that it accurately reproduces a range of commonly calculated properties of both the crystalline solid and the liquid phases. In addition, unlike simpler existing potentials, SNAP correctly predicts the energy barrier for screw dislocation migration in BCC tantalum.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2019

A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems

Presently the training of neural networks relies on data of comparable accuracy but in real applications only a very small set of high-fidelity data is available while inexpensive lower fidelity data may be plentiful. We propose a new composite neural network (NN) that can be trained based on multi-fidelity data. It is comprised of three NNs, with the first NN trained using the low-fidelity data and coupled to two high-fidelity NNs, one with activation functions and another one without, in order to discover and exploit nonlinear and linear correlations, respectively, between the low-fidelity and the high-fidelity data. We first demonstrate the accuracy of the new multi-fidelity NN for approximating some standard benchmark functions but also a 20-dimensional function that is not easy to approximate with other methods, e.g. Gaussian process regression. Subsequently, we extend the recently developed physics-informed neural networks (PINNs) to be trained with multi-fidelity data sets (MPINNs). MPINNs contain four fully-connected neural networks, where the first one approximates the low-fidelity data, while the second and third construct the correlation between the low- and high-fidelity data and produce the multi-fidelity approximation, which is then used in the last NN that encodes the partial differential equations (PDEs). Specifically, by decomposing the correlation into a linear and nonlinear part, the present model is capable of learning both the linear and complex nonlinear correlations between the low- and high-fidelity data adaptively. By training the MPINNs, we can: (1) obtain the correlation between the low- and high-fidelity data, (2) infer the quantities of interest based on a few scattered data, and (3) identify the unknown parameters in the PDEs. In particular, we employ the MPINNs to learn the hydraulic conductivity field for unsaturated flows as well as the reactive models for reactive transport. The results demonstrate that MPINNs can achieve relatively high accuracy based on a very small set of high-fidelity data. Despite the relatively low dimension and limited number of fidelities (two-fidelity levels) for the benchmark problems in the present study, the proposed model can be readily extended to very high-dimensional regression and classification problems involving multi-fidelity data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗