Explainable deep learning model to predict invasive bacterial infection in febrile young infants: A retrospective study.

BACKGROUND: Machine learning models have demonstrated superior performance in predicting invasive bacterial infection (IBI) in febrile infants compared to commonly used risk stratification criteria in recent studies. However, the black-box nature of these models can make them difficult to apply in clinical practice. In this study, we developed and validated an explainable deep learning model that can predict IBI in febrile infants ≤ 60 days of age visiting the emergency department. METHODS: We conducted a retrospective study of febrile infants aged ≤ 60 days who presented to the pediatric emergency department of a medical center in Taiwan between January 1, 2011 and December 31, 2019. Patients with uncertain test results and complex chronic health conditions were excluded. IBI was defined as the growth of a pathogen in the blood or cerebrospinal fluid. We used a deep neural network to develop a predictive model for IBI and compared its performance to the IBI score and step-by-step approach. The SHapley Additive Explanations (SHAP) technique was used to explain the model's predictions at different levels. RESULTS: Our study included 1847 patients, 53 (2.7%) of whom had IBI. The deep learning model performed similarly to the IBI score and step-by-step approach in terms of sensitivity and negative predictive value, but provided better specificity (54%), positive predictive value (5%), and area under the receiver-operating characteristic curve (0.87). SHapley Additive exPlanations identified five influential predictive variables (absolute neutrophil count, body temperature, heart rate, age, and C-reactive protein). CONCLUSION: We have developed an explainable deep learning model that can predict IBI in febrile infants aged 0-60 days. The model not only performs better than previous scoring systems, but also provides insight into how it arrives at its predictions through individual features and cases.

Product characteristics from the torrefaction of oil palm fiber pellets in inert and oxidative atmospheres.

The aim of this work was to study the characteristics of solid and liquid products from the torrefaction of oil palm fiber pellets (OPFP) in inert and oxidative environments. The torrefaction temperature and O2 concentration in the carrier gas were in the ranges of 275-350°C and 0-10 vol%, respectively, while the torrefaction duration was 30 min. The oxidative torrefaction of OPFP at 275°C drastically intensified the HHV of the biomass when compared to the non-oxidative torrefaction. OPFP torrefied at 300°C is recommended to upgrade the biomass, irrespective of the atmosphere. The HHV of condensed liquid was between 10.1 and 13.2 MJ kg(-)(1), and was promoted to 23.2-28.7 MJ kg(-)(1) following dewatering. This accounts for 92-139% improvement in the calorific value of the liquid. This reveals that the recovery of condensed liquid with dewatering is able to enhance the energy efficiency of a torrefaction system.

Biomass torrefaction characteristics in inert and oxidative atmospheres at various superficial velocities.

The reaction characteristics of four biomass materials (i.e. oil palm fiber, coconut fiber, eucalyptus, and Cryptomeria japonica) with non-oxidative and oxidative torrefaction at various superficial velocities are investigated where nitrogen and air are used as carrier gases. Three torrefaction temperatures of 250, 300, and 350 °C are considered. At a given temperature, the solid yield of biomass is not affected by N2 superficial velocity, revealing that the thermal degradation is controlled by heat and mass transfer in biomass. Increasing air superficial velocity decreases the solid yield, especially in oil palm fiber and coconut fiber, implying that the torrefaction reaction of biomass is dominated by surface oxidation. There exists an upper limit of air superficial velocity in the decrement of solid yield, suggesting that beyond this limit the thermal degradation of biomass is no longer governed by surface oxidation, but rather is controlled by internal mass transport.