Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications

Early detection of antibiotic resistance is a crucial task, especially for vulnerable patients under prolonged treatments with a single antibiotic. To solve this, machine learning approaches have been reported in the state of the art. Researchers have used MALDI-TOF MS in order to predict antibiotic...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Macaya Mejias, Vicente, Zabala-Blanco, David, López-Cortés, Xaviera A., Tirado, Felipe, Manriquez-Troncoso, José M., Ahumada-Garcia, Roberto
Formato: Articulo
Lenguaje:Inglés
Publicado: 2024
Materias:
Ciencias Informáticas
Antibiotic Resistance Prediction
MALDITOF Mass Spectrometry
Machine Learning in Medicine
Extreme Learning Machines
Weighted ELM
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/173719
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
id I19-R120-10915-173719
record_format dspace
spelling I19-R120-10915-1737192024-11-25T20:06:09Z http://sedici.unlp.edu.ar/handle/10915/173719 Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications Predicción de resistencia antibiótica mediante bases de datos de espectrometría de masas MALDI-TOF con aplicaciones de ELM Macaya Mejias, Vicente Zabala-Blanco, David López-Cortés, Xaviera A. Tirado, Felipe Manriquez-Troncoso, José M. Ahumada-Garcia, Roberto 2024-10 2024-11-25T18:08:45Z en Ciencias Informáticas Antibiotic Resistance Prediction MALDITOF Mass Spectrometry Machine Learning in Medicine Extreme Learning Machines Weighted ELM Early detection of antibiotic resistance is a crucial task, especially for vulnerable patients under prolonged treatments with a single antibiotic. To solve this, machine learning approaches have been reported in the state of the art. Researchers have used MALDI-TOF MS in order to predict antibiotic resistance and/or susceptibility in bacterial samples. Weis, et al. implemented LR, LightGBM and ANN to study the antibiotic resistance on bacterial strains of Escherichia Colt, Staphylococcus Aureus, and Klebsiella Pneumoniae. Despite promising results, the models have not achieved perfect accuracy, specifically when the classes are unbalanced. On the other hand, Extreme Learning Machine (ELM) is a training algorithm for forward propagation of single hidden layer neural networks, which converges much faster than traditional methods and offers promising performance along with less programmer intervention. In this way, this study introduced improved ELMs, including two weighted ELMs proposed by Zong, and the SMOTE technique in order to create new synthetic samples of the minority class. After heuristic optimization of ELM hiperparameters, results demonstrated 85% in accuracy and 85% in geometric mean for the classification problem in the case of weighted ELM 1 subject to the SMOTE technique of oversampling. La detección temprana de la resistencia a los antibióticos es una tarea crucial, especialmente en el caso de pacientes vulnerables sometidos a tratamientos prolongados con un único antibiótico. Para resolver este problema, se han utilizado métodos de aprendizaje automático. Los investigadores han utilizado MALDITOF MS para predecir la resistencia y/o susceptibilidad a los antibióticos en muestras bacterianas. Weis, et al. aplicaron LR, LightGBM y ANN para estudiar la resistencia a los antibióticos en cepas bacterianas de Escherichia Goli, Staphylococcus Aureus y Klebsiella Pneumoniae. A pesar de los prometedores resultados, los modelos no han logrado una precisión perfecta, concretamente cuando las clases están desequilibradas. Por otro lado, Extreme Learning Machine (ELM) es un algoritmo de entrenamiento para la propagación hacia delante de redes neuronales de una sola capa oculta, que converge mucho más rápido que los métodos tradicionales y ofrece un rendimiento prometedor junto con una menor intervención del programador. De este modo, este estudio introdujo ELMs mejorados, incluyendo dos ELMs ponderados propuestos por Zong, y la técnica SMOTE para crear nuevas muestras sintéticas de la clase minoritaria. Tras la optimización heurística de los hiperparámetros del ELM, los resultados demostraron un 85% de precisión y un 85% de media geométrica para el problema de clasificación en el caso del ELM ponderado 1 sujeto a la técnica SMOTE de sobremuestreo. Facultad de Informática Articulo Articulo http://creativecommons.org/licenses/by-nc-sa/4.0/ Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) application/pdf 88-98
institution Universidad Nacional de La Plata
institution_str I-19
repository_str R-120
collection SEDICI (UNLP)
language Inglés
topic Ciencias Informáticas
Antibiotic Resistance Prediction
MALDITOF Mass Spectrometry
Machine Learning in Medicine
Extreme Learning Machines
Weighted ELM
spellingShingle Ciencias Informáticas
Antibiotic Resistance Prediction
MALDITOF Mass Spectrometry
Machine Learning in Medicine
Extreme Learning Machines
Weighted ELM
Macaya Mejias, Vicente
Zabala-Blanco, David
López-Cortés, Xaviera A.
Tirado, Felipe
Manriquez-Troncoso, José M.
Ahumada-Garcia, Roberto
Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
topic_facet Ciencias Informáticas
Antibiotic Resistance Prediction
MALDITOF Mass Spectrometry
Machine Learning in Medicine
Extreme Learning Machines
Weighted ELM
description Early detection of antibiotic resistance is a crucial task, especially for vulnerable patients under prolonged treatments with a single antibiotic. To solve this, machine learning approaches have been reported in the state of the art. Researchers have used MALDI-TOF MS in order to predict antibiotic resistance and/or susceptibility in bacterial samples. Weis, et al. implemented LR, LightGBM and ANN to study the antibiotic resistance on bacterial strains of Escherichia Colt, Staphylococcus Aureus, and Klebsiella Pneumoniae. Despite promising results, the models have not achieved perfect accuracy, specifically when the classes are unbalanced. On the other hand, Extreme Learning Machine (ELM) is a training algorithm for forward propagation of single hidden layer neural networks, which converges much faster than traditional methods and offers promising performance along with less programmer intervention. In this way, this study introduced improved ELMs, including two weighted ELMs proposed by Zong, and the SMOTE technique in order to create new synthetic samples of the minority class. After heuristic optimization of ELM hiperparameters, results demonstrated 85% in accuracy and 85% in geometric mean for the classification problem in the case of weighted ELM 1 subject to the SMOTE technique of oversampling.
format Articulo
Articulo
author Macaya Mejias, Vicente
Zabala-Blanco, David
López-Cortés, Xaviera A.
Tirado, Felipe
Manriquez-Troncoso, José M.
Ahumada-Garcia, Roberto
author_facet Macaya Mejias, Vicente
Zabala-Blanco, David
López-Cortés, Xaviera A.
Tirado, Felipe
Manriquez-Troncoso, José M.
Ahumada-Garcia, Roberto
author_sort Macaya Mejias, Vicente
title Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
title_short Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
title_full Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
title_fullStr Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
title_full_unstemmed Predicting bacterial antibiotic resistance using MALDI-TOF mass spectrometry databases with ELM applications
title_sort predicting bacterial antibiotic resistance using maldi-tof mass spectrometry databases with elm applications
publishDate 2024
url http://sedici.unlp.edu.ar/handle/10915/173719
work_keys_str_mv AT macayamejiasvicente predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT zabalablancodavid predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT lopezcortesxavieraa predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT tiradofelipe predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT manriqueztroncosojosem predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT ahumadagarciaroberto predictingbacterialantibioticresistanceusingmalditofmassspectrometrydatabaseswithelmapplications
AT macayamejiasvicente prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
AT zabalablancodavid prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
AT lopezcortesxavieraa prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
AT tiradofelipe prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
AT manriqueztroncosojosem prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
AT ahumadagarciaroberto prediccionderesistenciaantibioticamediantebasesdedatosdeespectrometriademasasmalditofconaplicacionesdeelm
_version_ 1833157998577451008

Ejemplares similares