Search NASA⌕ Search

SEARCH · Search NASA

Results for “prognostics”

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 91 records · Page 5

MGMT Gene Promoter Methylation as a Potent Prognostic Factor in Glioblastoma Treated With Temozolomide-Based Chemoradiotherapy: A Single-Institution Study

Purpose: Recently, cells deficient in O{sup 6}-methylguanine-DNA methyltransferase (MGMT) were found to show increased sensitivity to temozolomide (TMZ). We evaluated whether hypermethylation of MGMT was associated with survival in patients with glioblastoma multiforme (GBM). Methods and Materials: We retrospectively analyzed 93 patients with histologically confirmed GBM who received involved-field radiotherapy with TMZ from 2001 to 2008. The median age was 58 years (range, 24-78 years). Surgical resection was total in 39 patients (42%), subtotal in 30 patients (32%), and partial in 17 patients (18%); only a biopsy was performed in 7 patients (8%). Postoperative radiotherapy began within 3 weeks of surgery in 87% of the patients. Radiotherapy doses ranged from 50 to 74 Gy (median, 70 Gy). MGMT gene methylation was determined in 78 patients; MGMT was unmethylated in 43 patients (55%) and methylated in 35 patients (45%). The median follow-up period was 22 months (range, 3-88 months) for all patients. Results: The median overall survival (OS) was 22 months, and progression-free survival (PFS) was 11 months. MGMT gene methylation was an independently significant prognostic factor for both OS (p = 0.002) and PFS (p = 0.008) in multivariate analysis. The median OS was 29 months for the methylated group and 20 months for the unmethylated group. In 35 patients with methylated MGMT genes, the 2-year and 5-year OS rates were 54% and 31%, respectively. Six patients with combined prognostic factors of methylated MGMT genes, age {<=}50 years, and total/subtotal resections are all alive 38 to 77 months after operation, whereas the median OS in 8 patients with unmethylated MGMT genes, age >50 years, and less than subtotal resection was 13.2 months. Conclusion: We confirmed that MGMT gene methylation is a potent prognostic factor in patients with GBM. Our results suggest that early postoperative radiotherapy and a high total/subtotal resection rate might further improve the outcome.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Prognostic Utility of Cell Cycle Progression Score in Men With Prostate Cancer After Primary External Beam Radiation Therapy

Purpose: To evaluate the prognostic utility of the cell cycle progression (CCP) score, a RNA signature based on the average expression level of 31 CCP genes, for predicting biochemical recurrence (BCR) in men with prostate cancer treated with external beam radiation therapy (EBRT) as their primary curative therapy. Methods and Materials: The CCP score was derived retrospectively from diagnostic biopsy specimens of men diagnosed with prostate cancer from 1991 to 2006 (n=141). All patients were treated with definitive EBRT; approximately half of the cohort was African American. Outcome was time from EBRT to BCR using the Phoenix definition. Median follow-up for patients without BCR was 4.8 years. Association with outcome was evaluated by Cox proportional hazards survival analysis and likelihood ratio tests. Results: Of 141 patients, 19 (13%) had BCR. The median CCP score for patient samples was 0.12. In univariable analysis, CCP score significantly predicted BCR (P=.0017). The hazard ratio for BCR was 2.55 for 1-unit increase in CCP score (equivalent to a doubling of gene expression). In a multivariable analysis that included Gleason score, prostate-specific antigen, percent positive cores, and androgen deprivation therapy, the hazard ratio for CCP changed only marginally and remained significant (P=.034), indicating that CCP provides prognostic information that is not provided by standard clinical parameters. With 10-year censoring, the CCP score was associated with prostate cancer-specific mortality (P=.013). There was no evidence for interaction between CCP and any clinical variable, including ethnicity. Conclusions: Among men treated with EBRT, the CCP score significantly predicted outcome and provided greater prognostic information than was available with clinical parameters. If validated in a larger cohort, CCP score could identify high-risk men undergoing EBRT who may need more aggressive therapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Residual Tumor After Neoadjuvant Chemoradiation Outside the Radiation Therapy Target Volume: A New Prognostic Factor for Survival in Esophageal Cancer

Purpose/Objective(s): The aim of this study was to analyze the accuracy of gross tumor volume (GTV) delineation and clinical target volume (CTV) margins for neoadjuvant chemoradiation therapy (neo-CRT) in esophageal carcinoma at pathologic examination and to determine the impact on survival. Methods and Materials: The study population consisted of 63 esophageal cancer patients treated with neo-CRT. GTV and CTV borders were demarcated in situ during surgery on the esophagus, using anatomical reference points to provide accurate information regarding tumor location at pathologic evaluation. To identify prognostic factors for disease-free survival (DFS) and overall survival (OS), a Cox regression analysis was performed. Results: After resection, macroscopic residual tumor was found outside the GTV in 7 patients (11%). Microscopic residual tumor was located outside the CTV in 9 patients (14%). The median follow-up was 15.6 months. With multivariate analysis, only microscopic tumor outside the CTV (hazard ratio [HR], 4.96; 95% confidence interval [CI], 1.03-15.36), and perineural growth (HR, 5.77; 95% CI, 1.27-26.13) were identified as independent prognostic factors for OS. The 1-year OS was 20% for patients with tumor outside the CTV and 86% for those without (P<.01). For DFS, microscopic tumor outside the CTV (HR, 5.92; 95% CI, 1.89-18.54) and ypN+ (HR, 3.36; 95% CI, 1.33-8.48) were identified as independent adverse prognostic factors. The 1-year DFS was 23% versus 77% for patients with or without tumor outside the CTV (P<.01). Conclusions: Microscopic tumor outside the CTV is associated with markedly worse OS after neo-CRT. This may either stress the importance of accurate tumor delineation or reflect aggressive tumor behavior requiring new adjuvant treatment modalities.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Prognostic Value and Reproducibility of Pretreatment CT Texture Features in Stage III Non-Small Cell Lung Cancer

Purpose: To determine whether pretreatment CT texture features can improve patient risk stratification beyond conventional prognostic factors (CPFs) in stage III non-small cell lung cancer (NSCLC). Methods and Materials: We retrospectively reviewed 91 cases with stage III NSCLC treated with definitive chemoradiation therapy. All patients underwent pretreatment diagnostic contrast enhanced computed tomography (CE-CT) followed by 4-dimensional CT (4D-CT) for treatment simulation. We used the average-CT and expiratory (T50-CT) images from the 4D-CT along with the CE-CT for texture extraction. Histogram, gradient, co-occurrence, gray tone difference, and filtration-based techniques were used for texture feature extraction. Penalized Cox regression implementing cross-validation was used for covariate selection and modeling. Models incorporating texture features from the 33 image types and CPFs were compared to those with models incorporating CPFs alone for overall survival (OS), local-regional control (LRC), and freedom from distant metastases (FFDM). Predictive Kaplan-Meier curves were generated using leave-one-out cross-validation. Patients were stratified based on whether their predicted outcome was above or below the median. Reproducibility of texture features was evaluated using test-retest scans from independent patients and quantified using concordance correlation coefficients (CCC). We compared models incorporating the reproducibility seen on test-retest scans to our original models and determined the classification reproducibility. Results: Models incorporating both texture features and CPFs demonstrated a significant improvement in risk stratification compared to models using CPFs alone for OS (P=.046), LRC (P=.01), and FFDM (P=.005). The average CCCs were 0.89, 0.91, and 0.67 for texture features extracted from the average-CT, T50-CT, and CE-CT, respectively. Incorporating reproducibility within our models yielded 80.4% (±3.7% SD), 78.3% (±4.0% SD), and 78.8% (±3.9% SD) classification reproducibility in terms of OS, LRC, and FFDM, respectively. Conclusions: Pretreatment tumor texture may provide prognostic information beyond that obtained from CPFs. Models incorporating feature reproducibility achieved classification rates of ∼80%. External validation would be required to establish texture as a prognostic factor.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Radiation Therapy Overcomes Adverse Prognostic Role of Cyclooxygenase-2 Expression on Reed-Sternberg Cells in Early Hodgkin Lymphoma

Purpose: To analyze the role of radiation therapy (RT) on the adverse prognostic influence of cyclooxygenase-2 (COX-2) expression on Reed-Sternberg (RS) cells, in the setting of early Hodgkin lymphoma (HL) treated with ABVD (adriamycin, vinblastine, bleomycin, dacarbazine). Methods and Materials: In the present study we retrospectively investigated the prognostic value of COX-2 expression in a large (n=143), uniformly treated early HL population from the Spanish Network of HL using tissue microarrays. Univariate and multivariate analyses were done, including the most recognized clinical variables and the potential role of administration of adjuvant RT. Results: Median age was 31 years; the expression of COX-2 defined a subgroup with significantly worse prognosis. Considering COX-2{sup +} patients, those who received RT had significantly better 5-year progression-free survival (PFS) (80% vs 54% if no RT; P=.008). In contrast, COX-2{sup −} patients only had a modest, nonsignificant benefit from RT in terms of 5-year PFS (90% vs 79%; P=.13). When we compared the outcome of patients receiving RT considering the expression of COX-2 on RS cells, we found a nonsignificant 10% difference in terms of PFS between COX-2{sup +} and COX-2{sup −} patients (P=.09), whereas the difference between the 2 groups was important (25%) in patients not receiving RT (P=.04). Conclusions: Cyclooxygenase-2 RS cell expression is an adverse independent prognostic factor in early HL. Radiation therapy overcomes the worse prognosis associated with COX-2 expression on RS cells, acting in a chemotherapy-independent way. Cyclooxygenase-2 RS cell expression may be useful for determining patient candidates with early HL to receive consolidation with RT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Metabolic Tumor Volume as a Prognostic Imaging-Based Biomarker for Head-and-Neck Cancer: Pilot Results From Radiation Therapy Oncology Group Protocol 0522

Purpose: To evaluate candidate fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) imaging biomarkers for head-and-neck chemoradiotherapy outcomes in the cooperative group trial setting. Methods and Materials: Radiation Therapy Oncology Group (RTOG) protocol 0522 patients consenting to a secondary FDG-PET/CT substudy were serially imaged at baseline and 8 weeks after radiation. Maximum standardized uptake value (SUVmax), SUV peak (mean SUV within a 1-cm sphere centered on SUVmax), and metabolic tumor volume (MTV) using 40% of SUVmax as threshold were obtained from primary tumor and involved nodes. Results: Of 940 patients entered onto RTOG 0522, 74 were analyzable for this substudy. Neither high baseline SUVmax nor SUVpeak from primary or nodal disease were associated with poor treatment outcomes. However, primary tumor MTV above the cohort median was associated with worse local-regional control (hazard ratio 4.01, 95% confidence interval 1.28-12.52, P=.02) and progression-free survival (hazard ratio 2.34, 95% confidence interval 1.02-5.37, P=.05). Although MTV and T stage seemed to correlate (mean MTV 6.4, 13.2, and 26.8 for T2, T3, and T4 tumors, respectively), MTV remained a strong independent prognostic factor for progression-free survival in bivariate analysis that included T stage. Primary MTV remained prognostic in p16-associated oropharyngeal cancer cases, although sample size was limited. Conclusion: High baseline primary tumor MTV was associated with worse treatment outcomes in this limited patient subset of RTOG 0522. Additional confirmatory work will be required to validate primary tumor MTV as a prognostic imaging biomarker for patient stratification in future trials.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Predicting Overall Survival After Stereotactic Ablative Radiation Therapy in Early-Stage Lung Cancer: Development and External Validation of the Amsterdam Prognostic Model

Purpose: A prognostic model for 5-year overall survival (OS), consisting of recursive partitioning analysis (RPA) and a nomogram, was developed for patients with early-stage non-small cell lung cancer (ES-NSCLC) treated with stereotactic ablative radiation therapy (SABR). Methods and Materials: A primary dataset of 703 ES-NSCLC SABR patients was randomly divided into a training (67%) and an internal validation (33%) dataset. In the former group, 21 unique parameters consisting of patient, treatment, and tumor factors were entered into an RPA model to predict OS. Univariate and multivariate models were constructed for RPA-selected factors to evaluate their relationship with OS. A nomogram for OS was constructed based on factors significant in multivariate modeling and validated with calibration plots. Both the RPA and the nomogram were externally validated in independent surgical (n=193) and SABR (n=543) datasets. Results: RPA identified 2 distinct risk classes based on tumor diameter, age, World Health Organization performance status (PS) and Charlson comorbidity index. This RPA had moderate discrimination in SABR datasets (c-index range: 0.52-0.60) but was of limited value in the surgical validation cohort. The nomogram predicting OS included smoking history in addition to RPA-identified factors. In contrast to RPA, validation of the nomogram performed well in internal validation (r{sup 2}=0.97) and external SABR (r{sup 2}=0.79) and surgical cohorts (r{sup 2}=0.91). Conclusions: The Amsterdam prognostic model is the first externally validated prognostication tool for OS in ES-NSCLC treated with SABR available to individualize patient decision making. The nomogram retained strong performance across surgical and SABR external validation datasets. RPA performance was poor in surgical patients, suggesting that 2 different distinct patient populations are being treated with these 2 effective modalities.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Prognostic Relevance of HPV Infection and p16 Overexpression in Squamous Cell Anal Cancer

Purpose: Human papillomavirus (HPV) DNA and p16 status have both been reported as prognostic factors in anal cancer, but the prognostic relevance of combined detection and particularly HPV−/p16+ and HPV+/p16− signatures is unknown. We evaluated combined HPV DNA and p16 status as a prognostic factor of treatment response in anal cancer. Methods: 106 patients treated with radiochemotherapy (RCT+5-FU/MMC) with available paraffin-embedded tumor tissue specimens were evaluated regarding local control (LC) and overall survival (OS) at 5 years. In addition to HPV DNA/p16 status, the influence of age, gender, previous surgery, initial recurrence, T stage, N status, and tumor localization was analyzed. Results: 63 patients were HPV+/p16+, 9 were HPV+/p16−, 11 were HPV−/p16+, and 23 were HPV−/p16−. In univariate analysis, LC was significantly better in patients with T1/2 stage, female gender, and HPV/p16 status. HPV+/p16+ was associated with significantly better LC (88.1%; 95% confidence interval [CI]: 78.89-97.31) compared with HPV−/p16+ (63.6%; 95% CI: 35.18-92.02; P=.021) and especially HPV−/p16− (55.8%; 95% CI: 33.46-78.14; P=.002) but not with HPV+/p16− (77.8%; 95% CI: 50.56-105.04; P=.270). OS was influenced by T stage and LC. HPV+/p16+ patients showed a trend toward better OS compared with HPV−/p16− patients (HPV+/p16+: 81.1%; 95% CI: 70.12-92.08 vs HPV−/p16−: 68.8%; 95%CI: 47.44-90.16; P=.138). On multivariate analysis, T3/4 stage and HPV/p16 status (HPV−/p16+, HPV−/p16− vs HPV+/p16+) predicted poorer LC (T3/4: 50.3% vs T1/2: 86.6%, hazard ratio [HR] 0.22; 95% CI: 0.09-0.53; P<.001; HPV+/p16+ vs HPV−/p16+: HR 4.73; 95% CI: 1.33-16.82; P=.016, and HPV+/p16+ vs HPV−/p16−: HR 6.40; 95% CI: 2.23-18.35; P<.001), whereas local relapse dramatically influenced OS. Conclusion: Our data suggest that HPV/p16 signature determines prognosis. HPV+/p16+ patients had the best prognosis, and HPV−/p16+ and HPV−/p16− patients showed the worst outcome and therefore require therapy optimization, particularly given that LC is the most important factor for OS.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Delineating transcriptional networks of prognostic gene signatures refines treatment recommendations for lymph node-negative breast cancer patients

The majority of women diagnosed with lymph node-negative breast cancer are unnecessarily treated with damaging chemotherapeutics after surgical resection. This highlights the importance of understanding and more accurately predicting patient prognosis. We define the transcriptional networks regulating well-established prognostic gene expression signatures. We find that the same set of transcriptional regulators consistently lie upstream of both ‘prognosis’ and ‘proliferation’ gene signatures, suggesting that a central transcriptional network underpins a shared phenotype within these signatures. Strikingly, the master transcriptional regulators within this network predict recurrence risk for lymph node-negative breast cancer better than currently used multigene prognostic assays, particularly in estrogen receptor-positive patients. Simultaneous examination of p16 INK4A expression, which predicts tumours that have bypassed cellular senescence, revealed that intermediate levels of p16 INK4A correlate with an intact pRB pathway and improved survival. A combination of these master transcriptional regulators and p16 INK4A , termed the OncoMasTR score, stratifies tumours based on their proliferative and senescence capacity, facilitating a clearer delineation of lymph node-negative breast cancer patients at high risk of recurrence, and thus requiring chemotherapy. Furthermore, OncoMasTR accurately classifies over 60% of patients as ‘low risk’, an improvement on existing prognostic assays, which has the potential to reduce overtreatment in early-stage patients. Taken together, the present study provides new insights into the transcriptional regulation of cellular proliferation in breast cancer and provides an opportunity to enhance and streamline methods of predicting breast cancer prognosis.

59 BASIC BIOLOGICAL SCIENCES↗

Improved prognostic classification of breast cancer defined by antagonistic activation patterns of immune response pathway modules

Background: Elucidating the activation pattern of molecular pathways across a given tumour type is a key challenge necessary for understanding the heterogeneity in clinical response and for developing novel more effective therapies. Gene expression signatures of molecular pathway activation derived from perturbation experiments in model systems as well as structural models of molecular interactions ("model signatures”) constitute an important resource for estimating corresponding activation levels in tumours. However, relatively few strategies for estimating pathway activity from such model signatures exist and only few studies have used activation patterns of pathways to refine molecular classifications of cancer. Methods: Here we propose a novel network-based method for estimating pathway activation in tumours from model signatures. We find that although the pathway networks inferred from cancer expression data are highly consistent with the prior information contained in the model signatures, that they also exhibit a highly modular structure and that estimation of pathway activity is dependent on this modular structure. We apply our methodology to a panel of 438 estrogen receptor negative (ER-) and 785 estrogen receptor positive (ER+) breast cancers to infer activation patterns of important cancer related molecular pathways. Results: We show that in ER negative basal and HER2+ breast cancer, gene expression modules reflecting T-cell helper-1 (Th1) and T-cell helper-2 (Th2) mediated immune responses play antagonistic roles as major risk factors for distant metastasis. Using Boolean interaction Cox-regression models to identify non-linear pathway combinations associated with clinical outcome, we show that simultaneous high activation of Th1 and low activation of a TGF-beta pathway module defines a subtype of particularly good prognosis and that this classification provides a better prognostic model than those based on the individual pathways. In ER+ breast cancer, we find that simultaneous high MYC and RAS activity confers significantly worse prognosis than either high MYC or high RAS activity alone. We further validate these novel prognostic classifications in independent sets of 173 ER- and 567 ER+ breast cancers. Conclusion: We have proposed a novel method for pathway activity estimation in tumours and have shown that pathway modules antagonize or synergize to delineate novel prognostic subtypes. Specifically, our results suggest that simultaneous modulation of T-helper differentiation and TGF-beta pathways may improve clinical outcome of hormone insensitive breast cancers over treatments that target only one of these pathways.

59 BASIC BIOLOGICAL SCIENCES↗

The expression level of HJURP has an independent prognostic impact and predicts the sensitivity to radiotherapy in breast cancer

Introduction: HJURP (Holliday Junction Recognition Protein) is a newly discovered gene reported to function at centromeres and to interact with CENPA. However its role in tumor development remains largely unknown. The goal of this study was to investigate the clinical significance of HJURP in breast cancer and its correlation with radiotherapeutic outcome. Methods: We measured HJURP expression level in human breast cancer cell lines and primary breast cancers by Western blot and/or by Affymetrix Microarray; and determined its associations with clinical variables using standard statistical methods. Validation was performed with the use of published microarray data. We assessed cell growth and apoptosis of breast cancer cells after radiation using high-content image analysis. Results: HJURP was expressed at higher level in breast cancer than in normal breast tissue. HJURP mRNA levels were significantly associated with estrogen receptor (ER), progesterone receptor (PR), Scarff-Bloom-Richardson (SBR) grade, age and Ki67 proliferation indices, but not with pathologic stage, ERBB2, tumor size, or lymph node status. Higher HJURP mRNA levels significantly decreased disease-free and overall survival. HJURP mRNA levels predicted the prognosis better than Ki67 proliferation indices. In a multivariate Cox proportional-hazard regression, including clinical variables as covariates, HJURP mRNA levels remained an independent prognostic factor for disease-free and overall survival. In addition HJURP mRNA levels were an independent prognostic factor over molecular subtypes (normal like, luminal, Erbb2 and basal). Poor clinical outcomes among patients with high HJURP expression werevalidated in five additional breast cancer cohorts. Furthermore, the patients with high HJURP levels were much more sensitive to radiotherapy. In vitro studies in breast cancer cell lines showed that cells with high HJURP levels were more sensitive to radiation treatment and had a higher rate of apoptosis than those with low levels. Knock down of HJURP in human breast cancer cells using shRNA reduced the sensitivity to radiation treatment. HJURP mRNA levels were significantly correlated with CENPA mRNA levels. Conclusions: HJURP mRNA level is a prognostic factor for disease-free and overall survival in patients with breast cancer and is a predictive biomarker for sensitivity to radiotherapy.

59 BASIC BIOLOGICAL SCIENCES↗

Online Monitoring and Prognostics for Passive Components in Nuclear Power Plants

This paper describes research towards developing prognostics technologies for light-water nuclear power reactors components. The focus of this paper is on passive components (those that do not need to change state or move to perform their function), though the technologies are applicable to other classes of components as well. A prototypic failure mechanism (high-cycle fatigue) is used to focus the efforts and provide context for the development effort. A Bayesian framework is proposed for the prognostics of remaining useful life, and applied to simulated data sets representing nondestructive measurements of high-cycle fatigue damage. Initial results of prognostics, based on simulated data sets are presented.

Ramuhalli, Pradeep↗

Prognostic significance of FAM83D gene expression across human cancer types

The family with sequence similarity 83, member D (FAM83D) gene has been proposed as a new prognostic marker for breast cancer. In this work, we further evaluate the prognostic significance of FAM83D expression in different breast cancer subtypes using a meta-analysis. Patients with higher FAM83D mRNA levels have significantly decreased overall and metastatic relapse-free survival, particularly in the group of patients with ER-positive, or luminal subtype tumors. We also assessed FAM83D alterations and its prognostic significance across 22 human cancer types using The Cancer Genome Atlas (TCGA). FAM83D is frequently gained in the majority of human cancer types, resulting in the elevated expression of FAM83D. Higher levels of FAM83D mRNA expression are significantly associated with decreased overall survival in several cancer types. Finally, we demonstrate that TP53 mutation in human cancers is coupled to a significant increase in the expression of FAM83D, and that a higher level of FAM83D expression is positively correlated with an increase in genome instability in many cancer types. These results identify FAM83D as a potential novel oncogene across multiple human cancer types.

59 BASIC BIOLOGICAL SCIENCES↗

Diagnostics, Prognostics, and Optimization for Lithium-Ion Battery Systems

Health management of lithium-ion battery systems presents a host of challenges due to their complex physics, large numbers of components, and a wide variety of degradation behaviors across different battery types. Dr. Paul Gasper will present on research from the Electrochemical Energy Storage Group on Lithium-ion battery diagnostics, prognostics, and optimization. Diagnostics research, including state-estimation via machine-learning from electrochemical impedance spectroscopy and DC pulses as well as continuous state-estimation via Kalman filters, will highlight the ongoing challenges for accurately measuring the state of batteries without performing time-consuming characterization tests. NLR's industry-recognized battery prognostics work, which predicts real-world battery degradation by identifying degradation rate models from accelerated aging data using statistical modeling and machine-learning, will be used to demonstrate the critical impact of battery controls, thermal management, and operating strategy on durability and lifetime. Finally, the use of prognostic models for financial or lifetime optimization will be discussed.

25 ENERGY STORAGE↗

Fleet-Wide Prognostic and Health Management Suite: Asset Fault Signature Database

Proactive online monitoring in the nuclear industry is being explored using the Electric Power Research Institute’s Fleet-Wide Prognostic and Health Management (FW-PHM) Suite software. The FW-PHM Suite is a set of web-based diagnostic and prognostic tools and databases that serves as an integrated health monitoring architecture. The FW-PHM Suite has four main modules: (1) Diagnostic Advisor, (2) Asset Fault Signature (AFS) Database, (3) Remaining Useful Life Advisor, and (4) Remaining Useful Life Database. The paper focuses on the AFS Database of the FW-PHM Suite, which is used to catalog asset fault signatures. A fault signature is a structured representation of the information that an expert would use to first detect and then verify the occurrence of a specific type of fault. The fault signatures developed to assess the health status of generator step-up transformers are described in the paper. The developed fault signatures capture this knowledge and implement it in a standardized approach, thereby streamlining the diagnostic and prognostic process. This will support the automation of proactive online monitoring techniques in nuclear power plants to diagnose incipient faults, perform proactive maintenance, and estimate the remaining useful life of assets.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Local-Level Prognostics Health Management Systems Framework for Passive AdvSMR Components. Interim Report

This report describes research results to date in support of the integration and demonstration of diagnostics technologies for prototypical AdvSMR passive components (to establish condition indices for monitoring) with model-based prognostics methods. The focus of the PHM methodology and algorithm development in this study is at the localized scale. Multiple localized measurements of material condition (using advanced nondestructive measurement methods), along with available measurements of the stressor environment, enhance the performance of localized diagnostics and prognostics of passive AdvSMR components and systems.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Uncertainty quantification in machine learning for engineering design and health prognostics: A tutorial

On top of machine learning (ML) models, uncertainty quantification (UQ) functions as an essential layer of safety assurance that could lead to more principled decision making by enabling sound risk assessment and management. The safety and reliability improvement of ML models empowered by UQ has the potential to significantly facilitate the broad adoption of ML solutions in high-stakes decision settings, such as healthcare, manufacturing, and aviation, to name a few. In this tutorial, we aim to provide a holistic lens on emerging UQ methods for ML models with a particular focus on neural networks and the applications of these UQ methods in tackling engineering design as well as prognostics and health management problems. Towards this goal, we start with a comprehensive classification of uncertainty types, sources, and causes pertaining to UQ of ML models. Next, we provide a tutorial-style description of several state-of-the-art UQ methods: Gaussian process regression, Bayesian neural network, neural network ensemble, and deterministic UQ methods focusing on spectral-normalized neural Gaussian process. Established upon the mathematical formulations, we subsequently examine the soundness of these UQ methods quantitatively and qualitatively (by a toy regression example) to examine their strengths and shortcomings from different dimensions. Then, we review quantitative metrics commonly used to assess the quality of predictive uncertainty in classification and regression problems. Afterward, we discuss the increasingly important role of UQ of ML models in solving challenging problems in engineering design and health prognostics. In conclusion, two case studies with source codes available on GitHub are used to demonstrate these UQ methods and compare their performance in the life prediction of lithium-ion batteries at the early stage (case study 1) and the remaining useful life prediction of turbofan engines (case study 2).

97 MATHEMATICS AND COMPUTING↗

Probabilistic machine learning for battery health diagnostics and prognostics—review and perspectives

Abstract Diagnosing lithium-ion battery health and predicting future degradation is essential for driving design improvements in the laboratory and ensuring safe and reliable operation over a product’s expected lifetime. However, accurate battery health diagnostics and prognostics is challenging due to the unavoidable influence of cell-to-cell manufacturing variability and time-varying operating circumstances experienced in the field. Machine learning approaches informed by simulation, experiment, and field data show enormous promise to predict the evolution of battery health with use; however, until recently, the research community has focused on deterministic modeling methods, largely ignoring the cell-to-cell performance and aging variability inherent to all batteries. To truly make informed decisions regarding battery design in the lab or control strategies for the field, it is critical to characterize the uncertainty in a model’s predictions. After providing an overview of lithium-ion battery degradation, this paper reviews the current state-of-the-art probabilistic machine learning models for health diagnostics and prognostics. Details of the various methods, their advantages, and limitations are discussed in detail with a primary focus on probabilistic machine learning and uncertainty quantification. Last, future trends and opportunities for research and development are discussed.

25 ENERGY STORAGE↗