A web-based microbiological hazard identification tool for infant foods

An integrated approach to identify and assess Microbiological Hazards (MHs) and mitigate risks in infant food chains is crucial to ensure safe foods for infants and young children. A systematic procedure was developed to identify MHs in specific infant foods. This includes five major steps: 1) relevant hazard-food pairing, 2) process inactivation efficiency, 3) recontamination possibility after processing, 4) MHs growth opportunity, and 5) MHs-food association level. These steps were integrated into an online tool called the Microbiological Hazards IDentification (MiID) decision support system (DSS), targeting food companies, governmental agencies and academia users, and is accessible at https://foodmicrobiologywur.shinyapps.io/Microbial_hazards_ID/. The MiID DSS was validated in four case studies, focussing on infant formula, fruit puree, cereal-based meals, and fresh fruits, each representing distinct products and processing characteristics. The results obtained through the application of the MiID DSS, compared with identification by food safety experts, consistently identified the top MHs in these food products. This process affirms its effectiveness in systematic hazard identification. The introduction of the MiID DSS helps to structure the first steps in HACCP (hazard analysis) and in risk assessment (hazard identification) to follow a structured and well-documented procedure, balancing the risk of overlooking relevant MHs or including too many irrelevant MHs. It is a valuable addition to risk analysis/assessment in infant food chains and has the potential for future extension. This includes the incorporation of newly acquired data related to infant foods via a semi-publicly hosted platform, or it can be adapted for hazard identification in general food products using a similar framework.

Evaluation of deep learning-based depression detection using medical claims data

Human accuracy in diagnosing psychiatric disorders is still low. Even though digitizing health care leads to more and more data, the successful adoption of AI-based digital decision support (DDSS) is rare. One reason is that AI algorithms are often not evaluated based on large, real-world data. This research shows the potential of using deep learning on the medical claims data of 812,853 people between 2018 and 2022, with 26,973,943 ICD-10-coded diseases, to predict depression (F32 and F33 ICD-10 codes). The dataset used represents almost the entire adult population of Estonia. Based on these data, to show the critical importance of the underlying temporal properties of the data for the detection of depression, we evaluate the performance of non-sequential models (LR, FNN), sequential models (LSTM, CNN-LSTM) and the sequential model with a decay factor (GRU-Δt, GRU-decay). Furthermore, since explainability is necessary for the medical domain, we combine a self-attention model with the GRU decay and evaluate its performance. We named this combination Att-GRU-decay. After extensive empirical experimentation, our model (Att-GRU-decay), with an AUC score of 0.990, an AUPRC score of 0.974, a specificity of 0.999 and a sensitivity of 0.944, proved to be the most accurate. The results of our novel Att-GRU-decay model outperform the current state of the art, demonstrating the potential usefulness of deep learning algorithms for DDSS development. We further expand this by describing a possible application scenario of the proposed algorithm for depression screening in a general practitioner (GP) setting-not only to decrease healthcare costs, but also to improve the quality of care and ultimately decrease people's suffering.

Robust design of a logistics system using FePIA procedure and analysis of trade-offs between CO2 emissions and net present value

The problems of flexible planning of the design of logistics systems for the collection of food products such as raw milk can result in a decrease in the performance of critical indicators for their performance. This paper proposes a new efficient methodology for robustly designing a first-mile logistics system for storing and refrigerating milk as a perishable product considering decisions related to open facilities and the flow of products, including sustainability indices. The proposed approach is modeled as a bi-objective problem by considering the minimization of greenhouse gas emissions (CO2) produced by milk transportation canteens and the maximization of the system configuration's net present value (NPV). We have analyzed and determined the most robust configuration for the first time and explained the robustness-NPV and robustness-CO2 relationships. The proposed mathematical model is solved by the Epsilon constraints method, and the robustness is calculated considering an extension of the FePIA methodology for multiobjective problems. A novel contribution is a balance in the possible future values generated by the company related to its cash flows and the generation of CO2 emissions when using a motorized transport frequently used in the shipment of raw milk considering a new important aspect such as the volume of product transported and the slope of the path between the production farm and the storage cooling tanks.

Decision support systems (DSS) for wastewater treatment plants - A review of the state of the art

The use of decision support systems (DSS) allows integrating all the issues related with sustainable development in view of providing a useful support to solve multi-scenario problems. In this work an extensive review on the DSSs applied to wastewater treatment plants (WWTPs) is presented. The main aim of the work is to provide an updated compendium on DSSs in view of supporting researchers and engineers on the selection of the most suitable method to address their management/operation/design problems. Results showed that DSSs were mostly used as a comprehensive tool that is capable of integrating several data and a multi-criteria perspective in order to provide more reliable results. Only one energy-focused DSS was found in literature, while DSSs based on quality and operational issues are very often applied to site-specific conditions. Finally, it would be important to encourage the development of more user-friendly DSSs to increase general interest and usability.

Response planning for accidental oil spills in Persian Gulf: A decision support system (DSS) based on consequence modeling

Different causes lead to accidental oil spills from fixed and mobile sources in the marine environment. Therefore, it is essential to have a systematic plan for mitigating oil spill consequences. In this research, a general DSS is proposed for passive and active response planning in Persian Gulf, before and after a spill. The DSS is based on NOAA's advanced oil spill model (GNOME), which is now linked with credible met-ocean datasets of CMEMS and ECMWF. The developed open-source tool converts the results of the Lagrangian oil spill model to quantitative parameters such as mean concentration and time of impact of oil. Using them, two new parameters, emergency response priority number (ERPN) and risk index (RI), are defined and used for response planning. The tool was tested in both deterministic and probabilistic modes, and found to be useful for evaluation of emergency response drills and risk-based prioritization of coastal areas.

Decision support for using mobile Rapid DNA analysis at the crime scene

Mobile Rapid DNA technology is close to being incorporated into crime scene investigations, with the potential to identify a perpetrator within hours. However, the use of these techniques entails the risk of losing the sample and potential evidence, because the device not only consumes the inserted sample, it is also is less sensitive than traditional technologies used in forensic laboratories. Scene of Crime Officers (SoCOs) therefore will face a 'time/success rate trade-off' issue when making a decision to apply this technology. In this study we designed and experimentally tested a Decision Support System (DSS) for the use of Rapid DNA technologies based on Rational Decision Theory (RDT). In a vignette study, where SoCOs had to decide on the use of a Rapid DNA analysis device, participating SoCOs were assigned to either the control group (making decisions under standard conditions), the Success Rate (SR) group (making decisions with additional information on DNA success rates of traces), or the DSS group (making decisions supported by introduction to RDT, including information on DNA success rates of traces). This study provides positive evidence that a systematic approach for decision-making on using Rapid DNA analysis assists SoCOs in the decision to use the rapid device. The results demonstrated that participants using a DSS made different and more transparent decisions on the use of Rapid DNA analysis when different case characteristics were explicitly considered. In the DSS group the decision to apply Rapid DNA analysis was influenced by the factors "time pressure" and "trace characteristics" like DNA success rates. In the SR group, the decisions depended solely on the trace characteristics and in the control group the decisions did not show any systematic differences on crime type or trace characteristic. Guiding complex decisions on the use of Rapid DNA analyses with a DSS could be an important step towards the use of these devices at the crime scene.

Approach Towards an Evidence-Oriented Knowledge and Data Acquisition for the Optimization of Interdisciplinary Care in Dentistry and General Medicine

Associations between dental and chronic-systemic diseases were observed frequently in medical research, however the findings of this research have so far found little relevance in everyday clinical treatment. Major problems are the assessment of evidence for correlations between such diseases and how to integrate current medical knowledge into the intersectoral care of dentists and general practitioners. On the example of dental and chronic-systemic diseases, the Dent@Prevent project develops an interdisciplinary decision support system (DSS), which provides the specialists with information relevant for the treatment of such cases. To provide the physicians with relevant medical knowledge, a mixed-methods approach is developed to acquire the knowledge in an evidence-oriented way. This procedure includes a literature review, routine data analyses, focus groups of dentists and general practitioners as well as the identification and integration of applicable guidelines and Patient Reported Measures (PRMs) into the treatment process. The developed mixed methods approach for an evidence-oriented knowledge acquisition indicates to be applicable and supportable for interdisciplinary projects. It can raise the systematic quality of the knowledge-acquisition process and can be applicable for an evidence-based system development. Further research is necessary to assess the impact on patient care and to evaluate possible applicability in other interdisciplinary areas.

mHealth App for iOS to Help in Diagnostic Decision in Ophthalmology to Primary Care Physicians

Decision support systems (DSS) are increasingly demanded due that diagnosis is one of the main activities that physicians accomplish every day. This fact seems critical when primary care physicians deal with uncommon problems belonging to specialized areas. The main objective of this paper is the development and user evaluation of a mobile DSS for iOS named OphthalDSS. This app has as purpose helping in anterior segment ocular diseases' diagnosis, besides offering educative content about ophthalmic diseases to users. For the deployment of this work, firstly it has been used the Apple IDE, Xcode, to develop the OphthalDSS mobile application using Objective-C as programming language. The core of the decision support system implemented by OphthalDSS is a decision tree developed by expert ophthalmologists. In order to evaluate the Quality of Experience (QoE) of primary care physicians after having tried the OphthalDSS app, a written inquiry based on the Likert scale was used. A total of 50 physicians answered to it, after trying the app during 1 month in their medical consultation. OphthalDSS is capable of helping to make diagnoses of diseases related to the anterior segment of the eye. Other features of OphthalDSS are a guide of each disease and an educational section. A 70% of the physicians answered in the survey that OphthalDSS performs in the way that they expected, and a 95% assures their trust in the reliability of the clinical information. Moreover, a 75% of them think that the decision system has a proper performance. Most of the primary care physicians agree with that OphthalDSS does the function that they expected, it is a user-friendly and the contents and structure are adequate. We can conclude that OphthalDSS is a practical tool but physicians require extra content that makes it a really useful one.

Reliability and validity of DS-ADHD: A decision support system on attention deficit hyperactivity disorders

BACKGROUND AND OBJECTIVES

The purpose of this study is to examine the reliability of the clinical use of the self-built decision support system, diagnosis-supported attention deficit hyperactivity disorder (DS-ADHD), in an effort to develop the DS-ADHD system, by probing into the development of indicating patterns of past screening support systems for ADHD.

METHODS

The study collected data based on 107 subjects, who were divided into two groups, non-ADHD and ADHD, based on the doctor's determination, using the DSM-IV diagnostic standards. The two groups then underwent Test of Variables of Attention (TOVA) and DS-ADHD testing. The survey and testing results underwent one-way ANOVA and split-half method statistical analysis, in order to further understand whether there were any differences between the DS-ADHD and the identification tools used in today's clinical trials.

RESULTS

The results of the study are as follows: 1) The ROC area between the TOVA and the clinical identification rate is 0.787 (95% confidence interval: 0.701-0.872); 2) The ROC area between the DS-ADHD and the clinical identification rate is 0.867 (95% confidence interval: 0.801-0.933).

CONCLUSIONS

The study results show that DS-ADHD has the characteristics of screening for ADHD, based on its reliability and validity. It does not display any statistical differences when compared with TOVA systems that are currently on the market. However, the system is more effective and the accuracy rate is better than TOVA. It is a good tool to screen ADHD not only in Chinese children, but also in western country.

Operator decision support system for integrated wastewater management including wastewater treatment plants and receiving water bodies

An operator decision support system (ODSS) is proposed to support operators of wastewater treatment plants (WWTPs) in making appropriate decisions. This system accounts for water quality (WQ) variations in WWTP influent and effluent and in the receiving water body (RWB). The proposed system is comprised of two diagnosis modules, three prediction modules, and a scenario-based supporting module (SSM). In the diagnosis modules, the WQs of the influent and effluent WWTP and of the RWB are assessed via multivariate analysis. Three prediction modules based on the k-nearest neighbors (k-NN) method, activated sludge model no. 2d (ASM2d) model, and QUAL2E model are used to forecast WQs for 3 days in advance. To compare various operating alternatives, SSM is applied to test various predetermined operating conditions in terms of overall oxygen transfer coefficient (Kla), waste sludge flow rate (Qw), return sludge flow rate (Qr), and internal recycle flow rate (Qir). In the case of unacceptable total phosphorus (TP), SSM provides appropriate information for the chemical treatment. The constructed ODSS was tested using data collected from Geumho River, which was the RWB, and S WWTP in Daegu City, South Korea. The results demonstrate the capability of the proposed ODSS to provide WWTP operators with more objective qualitative and quantitative assessments of WWTP and RWB WQs. Moreover, the current study shows that ODSS, using data collected from the study area, can be used to identify operational alternatives through SSM at an integrated urban wastewater management level.

'Rapid Learning health care in oncology' - an approach towards decision support systems enabling customised radiotherapy'

PURPOSE

An overview of the Rapid Learning methodology, its results, and the potential impact on radiotherapy.

MATERIAL AND RESULTS

Rapid Learning methodology is divided into four phases. In the data phase, diverse data are collected about past patients, treatments used, and outcomes. Innovative information technologies that support semantic interoperability enable distributed learning and data sharing without additional burden on health care professionals and without the need for data to leave the hospital. In the knowledge phase, prediction models are developed for new data and treatment outcomes by applying machine learning methods to data. In the application phase, this knowledge is applied in clinical practice via novel decision support systems or via extensions of existing models such as Tumour Control Probability models. In the evaluation phase, the predictability of treatment outcomes allows the new knowledge to be evaluated by comparing predicted and actual outcomes.

CONCLUSION

Personalised or tailored cancer therapy ensures not only that patients receive an optimal treatment, but also that the right resources are being used for the right patients. Rapid Learning approaches combined with evidence based medicine are expected to improve the predictability of outcome and radiotherapy is the ideal field to study the value of Rapid Learning. The next step will be to include patient preferences in the decision making.