Trustworthy data-driven prognostics in gas turbine engines are crucial for safety, cost-efficiency, and sustainability. Accurate predictions depend on data quality, model accuracy, uncertainty estimation, and practical implementation. This work discusses data quality attributes to build trust using anonymized real-world engine data, focusing on traceability, completeness, and representativeness. A significant challenge is handling missing data, which introduces bias and affects training and predictions. The study compares the accuracy of predictions using Exhaust Gas Temperature (EGT) margin, a key health indicator, by keeping missing values, using KNN-imputation, and employing a Generalized Additive Model (GAM). Preliminary results indicate that while KNN-imputation can be useful for identifying general trends, it may not be as effective for specific predictions compared to GAM, which considers the context of missing data. The choice of method depends on the study’s objective: broad trend forecasting or specific event prediction, each requiring different approaches to manage missing data.
DOCUMENT
Estimating the remaining useful life (RUL) of an asset lies at the heart of prognostics and health management (PHM) of many operations-critical industries such as aviation. Mod- ern methods of RUL estimation adopt techniques from deep learning (DL). However, most of these contemporary tech- niques deliver only single-point estimates for the RUL without reporting on the confidence of the prediction. This practice usually provides overly confident predictions that can have severe consequences in operational disruptions or even safety. To address this issue, we propose a technique for uncertainty quantification (UQ) based on Bayesian deep learning (BDL). The hyperparameters of the framework are tuned using a novel bi-objective Bayesian optimization method with objectives the predictive performance and predictive uncertainty. The method also integrates the data pre-processing steps into the hyperparameter optimization (HPO) stage, models the RUL as a Weibull distribution, and returns the survival curves of the monitored assets to allow informed decision-making. We vali- date this method on the widely used C-MAPSS dataset against a single-objective HPO baseline that aggregates the two ob- jectives through the harmonic mean (HM). We demonstrate the existence of trade-offs between the predictive performance and the predictive uncertainty and observe that the bi-objective HPO returns a larger number of hyperparameter configurations compared to the single-objective baseline. Furthermore, we see that with the proposed approach, it is possible to configure models for RUL estimation that exhibit better or comparable performance to the single-objective baseline when validated on the test sets.
DOCUMENT
Objective To systematically summarize the literature on the course of pain in patients with knee osteoarthritis (OA), prognostic factors that predict deterioration of pain, the course of physical functioning, and prognostic factors that predict deterioration of physical functioning in persons with knee OA. Methods A search was conducted in PubMed, CINAHL, Embase, Psych‐INFO, and SPORTDiscus up to January 2014. A meta‐analysis and a qualitative data synthesis were performed. Results Of the 58 studies included, 39 were of high quality. High heterogeneity across studies (I2 >90%) and within study populations (reflected by large SDs of change scores) was found. Therefore, the course of pain and physical functioning was interpreted to be indistinct. We found strong evidence for a number of prognostic factors predicting deterioration in pain (e.g., higher knee pain at baseline, bilateral knee symptoms, and depressive symptoms). We also found strong evidence for a number of prognostic factors predicting deterioration in physical functioning (e.g., worsening in radiographic OA, worsening of knee pain, lower knee extension muscle strength, lower walking speed, and higher comorbidity count). Conclusion Because of high heterogeneity across studies and within study populations, no conclusions can be drawn with regard to the course of pain and physical functioning. These findings support current research efforts to define subgroups or phenotypes within knee OA populations. Strong evidence was found for knee characteristics, clinical factors, and psychosocial factors as prognostics of deterioration of pain and physical functioning.
DOCUMENT
The full potential of predictive maintenance has not yet been utilised. Current solutions focus on individual steps of the predictive maintenance cycle and only work for very specific settings. The overarching challenge of predictive maintenance is to leverage these individual building blocks to obtain a framework that supports optimal maintenance and asset management. The PrimaVera project has identified four obstacles to tackle in order to utilise predictive maintenance at its full potential: lack of orchestration and automation of the predictive maintenance workflow, inaccurate or incomplete data and the role of human and organisational factors in data-driven decision support tools. Furthermore, an intuitive generic applicable predictive maintenance process model is presented in this paper to provide a structured way of deploying predictive maintenance solutions https://doi.org/10.3390/app10238348 LinkedIn: https://www.linkedin.com/in/john-bolte-0856134/
DOCUMENT
The Heating Ventilation and Air Conditioning (HVAC) sector is responsible for a large part of the total worldwide energy consumption, a significant part of which is caused by incorrect operation of controls and maintenance. HVAC systems are becoming increasingly complex, especially due to multi-commodity energy sources, and as a result, the chance of failures in systems and controls will increase. Therefore, systems that diagnose energy performance are of paramount importance. However, despite much research on Fault Detection and Diagnosis (FDD) methods for HVAC systems, they are rarely applied. One major reason is that proposed methods are different from the approaches taken by HVAC designers who employ process and instrumentation diagrams (P&IDs). This led to the following main research question: Which FDD architecture is suitable for HVAC systems in general to support the set up and implementation of FDD methods, including energy performance diagnosis? First, an energy performance FDD architecture based on information embedded in P&IDs was elaborated. The new FDD method, called the 4S3F method, combines systems theory with data analysis. In the 4S3F method, the detection and diagnosis phases are separated. The symptoms and faults are classified into 4 types of symptoms (deviations from balance equations, operating states (OS) and energy performance (EP), and additional information) and 3 types of faults (component, control and model faults). Second, the 4S3F method has been tested in four case studies. In the first case study, the symptom detection part was tested using historical Building Management System (BMS) data for a whole year: the combined heat and power plant of the THUAS (The Hague University of Applied Sciences) building in Delft, including an aquifer thermal energy storage (ATES) system, a heat pump, a gas boiler and hot and cold water hydronic systems. This case study showed that balance, EP and OS symptoms can be extracted from the P&ID and the presence of symptoms detected. In the second case study, a proof of principle of the fault diagnosis part of the 4S3F method was successfully performed on the same HVAC system extracting possible component and control faults from the P&ID. A Bayesian Network diagnostic, which mimics the way of diagnosis by HVAC engineers, was applied to identify the probability of all possible faults by interpreting the symptoms. The diagnostic Bayesian network (DBN) was set up in accordance with the P&ID, i.e., with the same structure. Energy savings from fault corrections were estimated to be up to 25% of the primary energy consumption, while the HVAC system was initially considered to have an excellent performance. In the third case study, a demand-driven ventilation system (DCV) was analysed. The analysis showed that the 4S3F method works also to identify faults on an air ventilation system.
DOCUMENT
Background Identify and establish consensus regarding potential prognostic factors for the development of chronic pain after a first episode of idiopathic, non-traumatic neck pain. Design This study used two consensus group methods: a modified Nominal Group (m-NGT) and a Delphi Technique. Methods The goal of the m-NGT was to obtain and categorize a list of potential modifiable prognostic factors. These factors were presented to a multidisciplinary panel in a two-round Delphi survey, which was conducted between November 2018 and January 2020. The participants were asked whether factors identified are of prognostic value, whether these factors are modifiable, and how to measure these factors in clinical practice. Consensus was a priori defined as 70% agreement among participants. Results Eighty-four factors were identified and grouped into seven categories during the expert meeting using the modified NGT. A workgroup reduced the list to 47 factors and grouped them into 12 categories. Of these factors, 26 were found to be potentially prognostic for chronification of neck pain (> 70% agreement). Twenty-one out of these 26 factors were found to be potentially modifiable by physiotherapists based on a two-round Delphi survey. Conclusion Based on an expert meeting (m-NGT) and a two-round Delphi survey, our study documents consensus (> 70%) on 26 prognostic factors. Twenty-one out of these 26 factors were found to be modifiable, and most factors were psychological in nature.
DOCUMENT
Objective To develop and internally validate a prognostic model to predict chronic pain after a new episode of acute or subacute non-specific idiopathic, non-traumatic neck pain in patients presenting to physiotherapy primary care, emphasising modifiable biomedical, psychological and social factors. Design A prospective cohort study with a 6-month follow-up between January 2020 and March 2023. Setting 30 physiotherapy primary care practices. Participants Patients with a new presentation of non-specific idiopathic, non-traumatic neck pain, with a duration lasting no longer than 12 weeks from onset. Baseline measures Candidate prognostic variables collected from participants included age and sex, neck pain symptoms, work-related factors, general factors, psychological and behavioural factors and the remaining factors: therapeutic relation and healthcare provider attitude. Outcome measures Pain intensity at 6 weeks, 3 months and 6 months on a Numeric Pain Rating Scale (NPRS) after inclusion. An NPRS score of ≥3 at each time point was used to define chronic neck pain. Results 62 (10%) of the 603 participants developed chronic neck pain. The prognostic factors in the final model were sex, pain intensity, reported pain in different body regions, headache since and before the neck pain, posture during work, employment status, illness beliefs about pain identity and recovery, treatment beliefs, distress and self-efficacy. The model demonstrated an optimism-corrected area under the curve of 0.83 and a corrected R2 of 0.24. Calibration was deemed acceptable to good, as indicated by the calibration curve. The Hosmer–Lemeshow test yielded a p-value of 0.7167, indicating a good model fit. Conclusion This model has the potential to obtain a valid prognosis for developing chronic pain after a new episode of acute and subacute non-specific idiopathic, non-traumatic neck pain. It includes mostly potentially modifiable factors for physiotherapy practice. External validation of this model is recommended.
LINK
Background: Prognosis of acute idiopathic neck pain is poor. An overview of modifiable and non-modifiable prognostic factors for the development of chronic musculoskeletal neck pain after an episode of idiopathic, non-traumatic neck pain is needed. Objective: Identify prognostic factors for pain intensity and perceived non-recovery at three, six and 12 months after a first episode of idiopathic, non-traumatic neck pain. Study design: Systematic review METHODS: Systematic literature search up to October 21, 2017 for prospective prognostic studies with main outcomes perceived non-recovery and pain intensity. The QUIPS was used for quality assessment. Results: Out of 2737 screened articles six prospective studies with high-risk-of-bias were identified, analyzing 47 and 43 factors for the outcome variables 'pain intensity' and 'perceived non-recovery', respectively. Based on univariate- and multivariate analyses we found moderate evidence for 'age> 40 years' and 'concomitant back pain' to be prognostic for 'pain intensity'. For the outcome 'perceived non-recovery' at 12 months, we found moderate evidence for both 'a previous period of neck pain' and 'accompanying headache' as prognostic variables for persistent pain, based on univariate analysis. No prognostic factor was found which was retained in more than one multivariate analysis for the outcome variable 'perceived non-recovery'. However, the quality of the evidence for these prognostic factors was low to very low. Conclusion: This review identifies prognostic factors for neck pain, of which only a few are modifiable. Further research is needed before drawing definite conclusions about the prognostic value of these factors.
LINK
BACKGROUND: Prediction models and prognostic scores have been increasingly popular in both clinical practice and clinical research settings, for example to aid in risk-based decision making or control for confounding. In many medical fields, a large number of prognostic scores are available, but practitioners may find it difficult to choose between them due to lack of external validation as well as lack of comparisons between them.METHODS: Borrowing methodology from network meta-analysis, we describe an approach to Multiple Score Comparison meta-analysis (MSC) which permits concurrent external validation and comparisons of prognostic scores using individual patient data (IPD) arising from a large-scale international collaboration. We describe the challenges in adapting network meta-analysis to the MSC setting, for instance the need to explicitly include correlations between the scores on a cohort level, and how to deal with many multi-score studies. We propose first using IPD to make cohort-level aggregate discrimination or calibration scores, comparing all to a common comparator. Then, standard network meta-analysis techniques can be applied, taking care to consider correlation structures in cohorts with multiple scores. Transitivity, consistency and heterogeneity are also examined.RESULTS: We provide a clinical application, comparing prognostic scores for 3-year mortality in patients with chronic obstructive pulmonary disease using data from a large-scale collaborative initiative. We focus on the discriminative properties of the prognostic scores. Our results show clear differences in performance, with ADO and eBODE showing higher discrimination with respect to mortality than other considered scores. The assumptions of transitivity and local and global consistency were not violated. Heterogeneity was small.CONCLUSIONS: We applied a network meta-analytic methodology to externally validate and concurrently compare the prognostic properties of clinical scores. Our large-scale external validation indicates that the scores with the best discriminative properties to predict 3 year mortality in patients with COPD are ADO and eBODE.
DOCUMENT
BACKGROUND: The SpO2/FiO2 is a useful oxygenation parameter with prognostic capacity in patients with ARDS. We investigated the prognostic capacity of SpO2/FiO2 for mortality in patients with ARDS due to COVID-19.METHODS: This was a post-hoc analysis of a national multicenter cohort study in invasively ventilated patients with ARDS due to COVID-19. The primary endpoint was 28-day mortality.RESULTS: In 869 invasively ventilated patients, 28-day mortality was 30.1%. The SpO2/FiO2 on day 1 had no prognostic value. The SpO2/FiO2 on day 2 and day 3 had prognostic capacity for death, with the best cut-offs being 179 and 199, respectively. Both SpO2/FiO2 on day 2 (OR, 0.66 [95%-CI 0.46-0.96]) and on day 3 (OR, 0.70 [95%-CI 0.51-0.96]) were associated with 28-day mortality in a model corrected for age, pH, lactate levels and kidney dysfunction (AUROC 0.78 [0.76-0.79]). The measured PaO2/FiO2 and the PaO2/FiO2 calculated from SpO2/FiO2 were strongly correlated (Spearman's r = 0.79).CONCLUSIONS: In this cohort of patients with ARDS due to COVID-19, the SpO2/FiO2 on day 2 and day 3 are independently associated with and have prognostic capacity for 28-day mortality. The SpO2/FiO2 is a useful metric for risk stratification in invasively ventilated COVID-19 patients.
MULTIFILE
