Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral

Teledermatology has developed rapidly in recent years and is nowadays an essential tool for early diagnosis. In this work, we aim to improve existing Teledermatology processes for skin lesion diagnosis by developing a deep learning approach for risk prioritization with a dataset of retrospective data from referral requests of the Portuguese National Health System. Given the high complexity of this task, we propose a new prioritization pipeline guided and inspired by domain knowledge. We explored automatic lesion segmentation and tested different learning schemes, namely hierarchical classification and curriculum learning approaches, optionally including additional patient metadata. The final priority level prediction can then be obtained by combining predicted diagnosis and a baseline priority level accounting for explicit expert knowledge. In both the differential diagnosis and prioritization branches, lesion segmentation with 30% tolerance for contextual information was shown to improve classification when compared with a flat baseline model trained on original images; furthermore, the addition of patient information was not beneficial for most experiments. Curriculum learning delivered better results than a flat or hierarchical approach. The combination of diagnosis information and a knowledge map, created in collaboration with dermatologists, together with the priority achieved interesting results (best macro F1 of 43.93% for a validated test set), paving the way for new data-centric and knowledge-driven approaches.

Reliability Analysis of Intelligent Manufacturing Systems Based on Improved FMEA Combined With Machine Learning

Along with the booming of intelligent manufacturing, the reliability management of intelligent manufacturing systems appears increasingly important. Failure mode and effects analysis (FMEA) is a prospective reliability management instrument extensively utilized to manage failure modes of systems, products, processes, and services in various industries. However, the conventional FMEA method has been criticized for its inherent limitations. Therefore, this paper devises a method based on improved FMEA model combined with machine learning for complex systems and applies it to the reliability management of intelligent manufacturing systems. The structured network of failure modes is constructed based on the knowledge graph for the intelligent manufacturing systems. The grey relation analysis (GRA) is applied to determine the risk prioritization of failure modes, hereafter the clustering analysis is employed to extract the features of failure modes. The results show that the proposed method can more accurately reflect the coupling relationship between the failure modes compared with the conventional FMEA method. This research provides significant support for the reliability and risk management of complex systems such as intelligent manufacturing systems.

A Concept of Risk Prioritization in FMEA Analysis for Fluid Power Systems

FMEA analysis is a tool of quality improvement that has been widely used for decades. Its classical version prioritizes risk of failure by risk priority number (RPN). The RPN is a product of severity (S), occurrence (O), and detection (D), where all of the factors have equal levels of significance. This assumption is one of the most commonly criticized drawbacks, as it has given unreasonable results for real-world applications. The RPN can produce equal values for combinations of risk factors with different risk implications. Another issue is that of the uncertainties and subjectivities of information employed in FMEA analysis that may arise from lack of knowledge, experience, and employed linguistic terms. Many alternatives of risk assessment methods have been proposed to overcome the weaknesses of classical FMEA risk management in which we can distinguish methods of modification of RPN numbers of employing new tools. In this study, we propose a modification of the traditional RPN number. The main difference is that severity and occurrence are valued based on subfactors. The detection number remained unchanged. Additionally, the proposed method prioritizes risk in terms of implied risk to the systems by implementing functional failures (effects of potential failures). A typical fluid power system was used to illustrate the application of this method. The method showed the correct failure classification, which meets the industrial experience and other research results of failures of fluid power systems.

A definition and categorization system for advanced materials : the foundation for risk-informed environmental health and safety testing

Novel materials with unique or enhanced properties relative to conventional materials are being developed at an increasing rate. These materials are often referred to as advanced materials (AdMs) and they enable technological innovations that can benefit society. Despite their benefits, however, the unique characteristics of many AdMs, including many nanomaterials, are poorly understood and may pose environmental safety and occupational health (ESOH) risks that are not readily determined by traditional risk assessment methods. To assess these risks while keeping up with the pace of development, technology developers and risk assessors frequently employ risk-screening methods that depend on a clear definition for the materials that are to be assessed (e.g., engineered nanomaterial) as well as a method for binning materials into categories for ESOH risk prioritization. In this study, we aim to establish a practitioner-driven definition for AdMs and a practitioner-validated framework for categorizing AdMs into conceptual groupings based on material characteristics. The definition and categorization framework established here serve as a first step in determining if and when there is a need for specific ESOH and regulatory screening for an AdM as well as the type and extent of risk-related information that should be collected or generated for AdMs and AdM-enabled technologies.

Risk Assessing The Risks: Deliberations On Risk Prioritization

One of the dilemmas facing the quality risk management function is with a series of completed risk assessments and a series of multiple outcomes that require addressing, in the context of limited resources. When faced with multiple risks, how are these to be prioritized?

A Z-number integrated weighted MULTIMO-ORA method for risk prioritization in FMEA

Current risk prioritization approaches for FMEA models are insufficient to cope with risk analysis problem in which the self-confidence of expert’s judgment and the deviation among risk evaluation information are considered, simultaneously. Therefore, to remedy this limitation, this paper reports an extended risk prioritization approach by integrating the MULTIMOORA approach, Z-numbers and power weighted average (PWA) operator. Firstly, the Z-numbers with triangular fuzzy numbers are applied to reflect the self-confidence and uncertainty of expert’s judgment. Next, the PWA operator for Z-numbers (Z-PWA) with similarity measure is proposed to obtain the group risk evaluation matrix by considering the influence of the deviation among risk evaluation information. Then, an extended version of MULTIMOORA method with developed entropy method is presented to calculate risk priority ranking order of each failure. Finally, the equipment failures in a certain oil and gas plant is utilized to test the extended risk prioritization approach for FMEA model. After that, the sensitivity and comparison studies are led to illustrate the availability and reliability of the proposed risk prioritization approach for FMEA based risk analysis problem.

SEPIA: simulation-based evaluation of prioritization algorithms

Background The ability to prioritize people living with HIV (PLWH) by risk of future transmissions could aid public health officials in optimizing epidemiological intervention. While methods exist to perform such prioritization based on molecular data, their effectiveness and accuracy are poorly understood, and it is unclear how one can directly compare the accuracy of different methods. We introduce SEPIA (Simulation-based Evaluation of PrIoritization Algorithms), a novel simulation-based framework for determining the effectiveness of prioritization algorithms. SEPIA expands upon prior related work by defining novel metrics of effectiveness with which to compare prioritization techniques, as well as by creating a simulation-based tool with which to perform such effectiveness comparisons. Under several metrics of effectiveness that we propose, we compare two existing prioritization approaches: one phylogenetic (ProACT) and one distance-based (growth of HIV-TRACE transmission clusters). Results Using all proposed metrics, ProACT consistently slightly outperformed the transmission cluster growth approach. However, both methods consistently performed just marginally better than random, suggesting that there is significant room for improvement in prioritization tools. Conclusion We hope that, by providing ways to quantify the effectiveness of prioritization methods in simulation, SEPIA will aid researchers in developing novel risk prioritization tools for PLWH.

Risk Prioritization for Failure Modes in Mining Railcars

Railway transportation systems are generally used to transport minerals from large-scale mines. Any failure in the railcar components may cause delays or even catastrophic derailment accidents. Failure mode and effect analysis (FMEA) is an effective tool for the risk assessment of mechanical systems. This method is an appropriate approach to identify the critical failure modes and provide proper control measures to reduce the level of risk. This research aims to propose an approach to identify and prioritize the failure modes based on their importance degree. To achieve this, the analytical hierarchy process (AHP) is used along with the FMEA. To compensate for the scarcities of the conventional FMEA in using the linguistic variables, the proposed approach is developed under the fuzzy environment. The proposed approach was applied in a case study, a rolling stock operated in an iron ore mine located in Sweden. The results of this study are helpful to identify not only the most important failure modes but also the most serious and critical ones.