Allan Danilo de Lima1, Agnaldo J Lopes2, Jorge Luis Machado do Amaral3, Pedro Lopes de Melo4. 1. Electronic Engineering Post-Graduation Program, State University of Rio de Janeiro, Rio de Janeiro, Brazil. 2. Pulmonary Function Laboratory, Faculty of Medical Sciences, State University of Rio de Janeiro, Rio de Janeiro, Brazil. 3. Department of Electronics and Telecommunications Engineering, Rio de Janeiro State University, Rio de Janeiro, Brazil. 4. Biomedical Instrumentation Laboratory, Institute of Biology Roberto Alcantara Gomes and Laboratory of Clinical and Experimental Research in Vascular Biology (BioVasc), Rio de Janeiro State University, Rio de Janeiro, Brazil. plopes@uerj.br.
Abstract
BACKGROUND: In this work, we developed many machine learning classifiers to assist in diagnosing respiratory changes associated with sarcoidosis, based on results from the Forced Oscillation Technique (FOT), a non-invasive method used to assess pulmonary mechanics. In addition to accurate results, there is a particular interest in their interpretability and explainability, so we used Genetic Programming since the classification is made with intelligible expressions and we also evaluate the feature importance in different experiments to find the more discriminative features. METHODOLOGY/PRINCIPAL FINDINGS: We used genetic programming in its traditional tree form and a grammar-based form. To check if interpretable results are competitive, we compared their performance to K-Nearest Neighbors, Support Vector Machine, AdaBoost, Random Forest, LightGBM, XGBoost, Decision Trees and Logistic Regressor. We also performed experiments with fuzzy features and tested a feature selection technique to bring even more interpretability. The data used to feed the classifiers come from the FOT exams in 72 individuals, of which 25 were healthy, and 47 were diagnosed with sarcoidosis. Among the latter, 24 showed normal conditions by spirometry, and 23 showed respiratory changes. The results achieved high accuracy (AUC > 0.90) in two analyses performed (controls vs. individuals with sarcoidosis and normal spirometry and controls vs. individuals with sarcoidosis and altered spirometry). Genetic Programming and Grammatical Evolution were particularly beneficial because they provide intelligible expressions to make the classification. The observation of which features were selected most frequently also brought explainability to the study of sarcoidosis. CONCLUSIONS: The proposed system may provide decision support for clinicians when they are struggling to give a confirmed clinical diagnosis. Clinicians may reference the prediction results and make better decisions, improving the productivity of pulmonary function services by AI-assisted workflow.
BACKGROUND: In this work, we developed many machine learning classifiers to assist in diagnosing respiratory changes associated with sarcoidosis, based on results from the Forced Oscillation Technique (FOT), a non-invasive method used to assess pulmonary mechanics. In addition to accurate results, there is a particular interest in their interpretability and explainability, so we used Genetic Programming since the classification is made with intelligible expressions and we also evaluate the feature importance in different experiments to find the more discriminative features. METHODOLOGY/PRINCIPAL FINDINGS: We used genetic programming in its traditional tree form and a grammar-based form. To check if interpretable results are competitive, we compared their performance to K-Nearest Neighbors, Support Vector Machine, AdaBoost, Random Forest, LightGBM, XGBoost, Decision Trees and Logistic Regressor. We also performed experiments with fuzzy features and tested a feature selection technique to bring even more interpretability. The data used to feed the classifiers come from the FOT exams in 72 individuals, of which 25 were healthy, and 47 were diagnosed with sarcoidosis. Among the latter, 24 showed normal conditions by spirometry, and 23 showed respiratory changes. The results achieved high accuracy (AUC > 0.90) in two analyses performed (controls vs. individuals with sarcoidosis and normal spirometry and controls vs. individuals with sarcoidosis and altered spirometry). Genetic Programming and Grammatical Evolution were particularly beneficial because they provide intelligible expressions to make the classification. The observation of which features were selected most frequently also brought explainability to the study of sarcoidosis. CONCLUSIONS: The proposed system may provide decision support for clinicians when they are struggling to give a confirmed clinical diagnosis. Clinicians may reference the prediction results and make better decisions, improving the productivity of pulmonary function services by AI-assisted workflow.
Authors: Paolo Spagnolo; Giulio Rossi; Rocco Trisolini; Nicola Sverzellati; Robert P Baughman; Athol U Wells Journal: Lancet Respir Med Date: 2018-04-03 Impact factor: 30.700
Authors: Jorge L M Amaral; Agnaldo J Lopes; Alvaro C D Faria; Pedro L Melo Journal: Comput Methods Programs Biomed Date: 2014-11-22 Impact factor: 5.428
Authors: Marko Topalovic; Nilakash Das; Pierre-Régis Burgel; Marc Daenen; Eric Derom; Christel Haenebalcke; Rob Janssen; Huib A M Kerstjens; Giuseppe Liistro; Renaud Louis; Vincent Ninane; Christophe Pison; Marc Schlesser; Piet Vercauter; Claus F Vogelmeier; Emiel Wouters; Jokke Wynants; Wim Janssens Journal: Eur Respir J Date: 2019-04-11 Impact factor: 16.671
Authors: Jorge L M Amaral; Agnaldo J Lopes; Juliana Veiga; Alvaro C D Faria; Pedro L Melo Journal: Comput Methods Programs Biomed Date: 2017-03-28 Impact factor: 5.428
Authors: Ane Johannessen; Sverre Lehmann; Ernst R Omenaas; Geir Egil Eide; Per S Bakke; Amund Gulsvik Journal: Am J Respir Crit Care Med Date: 2006-03-23 Impact factor: 21.405
Authors: Domingos S M Andrade; Luigi Maciel Ribeiro; Agnaldo J Lopes; Jorge L M Amaral; Pedro L Melo Journal: Biomed Eng Online Date: 2021-03-25 Impact factor: 2.819