Enhancing sepsis management through machine learning techniques: A review / Mejora en el manejo de sepsis mediante técnicas de aprendizaje automático: una revisión

Sepsis is a major public health problem and a leading cause of death in the world, where delay in the beginning of treatment, along with clinical guidelines non-adherence have been proved to be associated with higher mortality. Machine Learning is increasingly being adopted in developing innovative Clinical Decision Support Systems in many areas of medicine, showing a great potential for automatic prediction of diverse patient conditions, as well as assistance in clinical decision making. In this context, this work conducts a narrative review to provide an overview of how specific Machine Learning techniques can be used to improve sepsis management, discussing the main tasks addressed, the most popular methods and techniques, as well as the obtained results, in terms of both intelligent system accuracy and clinical outcomes improvement (AU)

La sepsis representa un problema de salud pública de primer orden y es una de las principales causas de muerte a nivel mundial. El retraso en el inicio del tratamiento, junto con la no adherencia a las guías de práctica clínica se asocian a una mayor mortalidad. El aprendizaje automático o machine learning están siendo empleados en el desarrollo de sistemas de apoyo a la decisión clínica, innovadores en muchas áreas de la medicina, mostrando un gran potencial para la predicción de diversas condiciones del paciente, así como en la asistencia durante el proceso de toma de decisiones médicas. En este sentido, este trabajo lleva a cabo una revisión narrativa para proporcionar una visión general de cómo las técnicas de machine learning pueden ser empleadas para mejorar el manejo de la sepsis, discutiendo las principales tareas que tratan de resolver, los métodos y las técnicas más empleados, así como los resultados obtenidos, tanto en términos de precisión de los sistemas inteligentes, como en la mejora de los resultados clínicos (AU)

Enhancing sepsis management through machine learning techniques: A review.

Sepsis is a major public health problem and a leading cause of death in the world, where delay in the beginning of treatment, along with clinical guidelines non-adherence have been proved to be associated with higher mortality. Machine Learning is increasingly being adopted in developing innovative Clinical Decision Support Systems in many areas of medicine, showing a great potential for automatic prediction of diverse patient conditions, as well as assistance in clinical decision making. In this context, this work conducts a narrative review to provide an overview of how specific Machine Learning techniques can be used to improve sepsis management, discussing the main tasks addressed, the most popular methods and techniques, as well as the obtained results, in terms of both intelligent system accuracy and clinical outcomes improvement.

Enhancing sepsis management through machine learning techniques: A review.

Sepsis is a major public health problem and a leading cause of death in the world, where delay in the beginning of treatment, along with clinical guidelines non-adherence have been proved to be associated with higher mortality. Machine Learning is increasingly being adopted in developing innovative Clinical Decision Support Systems in many areas of medicine, showing a great potential for automatic prediction of diverse patient conditions, as well as assistance in clinical decision making. In this context, this work conducts a narrative review to provide an overview of how specific Machine Learning techniques can be used to improve sepsis management, discussing the main tasks addressed, the most popular methods and techniques, as well as the obtained results, in terms of both intelligent system accuracy and clinical outcomes improvement.

High performance computing for three-dimensional agent-based molecular models.

Agent-based simulations are increasingly popular in exploring and understanding cellular systems, but the natural complexity of these systems and the desire to grasp different modelling levels demand cost-effective simulation strategies and tools. In this context, the present paper introduces novel sequential and distributed approaches for the three-dimensional agent-based simulation of individual molecules in cellular events. These approaches are able to describe the dimensions and position of the molecules with high accuracy and thus, study the critical effect of spatial distribution on cellular events. Moreover, two of the approaches allow multi-thread high performance simulations, distributing the three-dimensional model in a platform independent and computationally efficient way. Evaluation addressed the reproduction of molecular scenarios and different scalability aspects of agent creation and agent interaction. The three approaches simulate common biophysical and biochemical laws faithfully. The distributed approaches show improved performance when dealing with large agent populations while the sequential approach is better suited for small to medium size agent populations. Overall, the main new contribution of the approaches is the ability to simulate three-dimensional agent-based models at the molecular level with reduced implementation effort and moderate-level computational capacity. Since these approaches have a generic design, they have the major potential of being used in any event-driven agent-based tool.

Mass-Up: an all-in-one open software application for MALDI-TOF mass spectrometry knowledge discovery.

BACKGROUND: Mass spectrometry is one of the most important techniques in the field of proteomics. MALDI-TOF mass spectrometry has become popular during the last decade due to its high speed and sensitivity for detecting proteins and peptides. MALDI-TOF-MS can be also used in combination with Machine Learning techniques and statistical methods for knowledge discovery. Although there are many software libraries and tools that can be combined for these kind of analysis, there is still a need for all-in-one solutions with graphical user-friendly interfaces and avoiding the need of programming skills. RESULTS: Mass-Up, an open software multiplatform application for MALDI-TOF-MS knowledge discovery is herein presented. Mass-Up software allows data preprocessing, as well as subsequent analysis including (i) biomarker discovery, (ii) clustering, (iii) biclustering, (iv) three-dimensional PCA visualization and (v) classification of large sets of spectra data. CONCLUSIONS: Mass-Up brings knowledge discovery within reach of MALDI-TOF-MS researchers. Mass-Up is distributed under license GPLv3 and it is open and free to all users at http://sing.ei.uvigo.es/mass-up.

Speeding up the screening of steroids in urine: development of a user-friendly library.

This work presents a novel database search engine - MLibrary - designed to assist the user in the detection and identification of androgenic anabolic steroids (AAS) and its metabolites by matrix assisted laser desorption/ionization (MALDI) and mass spectrometry-based strategies. The detection of the AAS in the samples was accomplished by searching (i) the mass spectrometric (MS) spectra against the library developed to identify possible positives and (ii) by comparison of the tandem mass spectrometric (MS/MS) spectra produced after fragmentation of the possible positives with a complete set of spectra that have previously been assigned to the software. The urinary screening for anabolic agents plays a major role in anti-doping laboratories as they represent the most abused drug class in sports. With the help of the MLibrary software application, the use of MALDI techniques for doping control is simplified and the time for evaluation and interpretation of the results is reduced. To do so, the search engine takes as input several MALDI-TOF-MS and MALDI-TOF-MS/MS spectra. It aids the researcher in an automatic mode by identifying possible positives in a single MS analysis and then confirming their presence in tandem MS analysis by comparing the experimental tandem mass spectrometric data with the database. Furthermore, the search engine can, potentially, be further expanded to other compounds in addition to AASs. The applicability of the MLibrary tool is shown through the analysis of spiked urine samples.

BioAnnote: a software platform for annotating biomedical documents with application in medical learning environments.

Automatic term annotation from biomedical documents and external information linking are becoming a necessary prerequisite in modern computer-aided medical learning systems. In this context, this paper presents BioAnnote, a flexible and extensible open-source platform for automatically annotating biomedical resources. Apart from other valuable features, the software platform includes (i) a rich client enabling users to annotate multiple documents in a user friendly environment, (ii) an extensible and embeddable annotation meta-server allowing for the annotation of documents with local or remote vocabularies and (iii) a simple client/server protocol which facilitates the use of our meta-server from any other third-party application. In addition, BioAnnote implements a powerful scripting engine able to perform advanced batch annotations.

Using variable precision rough set for selection and classification of biological knowledge integrated in DNA gene expression.

DNA microarrays have contributed to the exponential growth of genomic and experimental data in the last decade. This large amount of gene expression data has been used by researchers seeking diagnosis of diseases like cancer using machine learning methods. In turn, explicit biological knowledge about gene functions has also grown tremendously over the last decade. This work integrates explicit biological knowledge, provided as gene sets, into the classication process by means of Variable Precision Rough Set Theory (VPRS). The proposed model is able to highlight which part of the provided biological knowledge has been important for classification. This paper presents a novel model for microarray data classification which is able to incorporate prior biological knowledge in the form of gene sets. Based on this knowledge, we transform the input microarray data into supergenes, and then we apply rough set theory to select the most promising supergenes and to derive a set of easy interpretable classification rules. The proposed model is evaluated over three breast cancer microarrays datasets obtaining successful results compared to classical classification techniques. The experimental results shows that there are not significant differences between our model and classical techniques but it is able to provide a biological-interpretable explanation of how it classifies new samples.

Indirect ultrasonication for protein quantification and peptide mass mapping through mass spectrometry-based techniques.

We report in this work a fast protocol for protein quantification and for peptide mass mapping that rely on (18)O isotopic labeling through the decoupling procedure. It is demonstrated that the purity and source of trypsin do not compromise the labeling degree and efficiency of the decoupled labeling reaction, and that the pH of the labeling reaction is a critical factor to obtain a significant (18)O double labeling. We also show that the same calibration curve can be used for MALDI protein quantification during several days maintaining a reasonable accuracy, thus simplifying the handling of the quantification process. In addition we demonstrate that (18)O isotopic labeling through the decoupling procedure can be successfully used to elaborate peptide mass maps. BSA was successfully quantified using the same calibration curve in different days and plasma from a freshwater fish, Cyprinus carpio, was used to elaborate the peptide mass maps.

AIBench: a rapid application development framework for translational research in biomedicine.

Applied research in both biomedical discovery and translational medicine today often requires the rapid development of fully featured applications containing both advanced and specific functionalities, for real use in practice. In this context, new tools are demanded that allow for efficient generation, deployment and reutilization of such biomedical applications as well as their associated functionalities. In this context this paper presents AIBench, an open-source Java desktop application framework for scientific software development with the goal of providing support to both fundamental and applied research in the domain of translational biomedicine. AIBench incorporates a powerful plug-in engine, a flexible scripting platform and takes advantage of Java annotations, reflection and various design principles in order to make it easy to use, lightweight and non-intrusive. By following a basic input-processing-output life cycle, it is possible to fully develop multiplatform applications using only three types of concepts: operations, data-types and views. The framework automatically provides functionalities that are present in a typical scientific application including user parameter definition, logging facilities, multi-threading execution, experiment repeatability and user interface workflow management, among others. The proposed framework architecture defines a reusable component model which also allows assembling new applications by the reuse of libraries from past projects or third-party software.