Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images
In this paper, we propose an approach based on ensemble learning to classify histology tissues stained with hematoxylin and eosin. The proposal was applied to representative images of colorectal cancer, oral epithelial dysplasia, non-Hodgkin’s lymphoma, and liver tissues (the classification of gende...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-01-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/14/3/1084 |
| _version_ | 1827354923983962112 |
|---|---|
| author | Jaqueline J. Tenguam Leonardo H. da Costa Longo Guilherme F. Roberto Thaína A. A. Tosta Paulo R. de Faria Adriano M. Loyola Sérgio V. Cardoso Adriano B. Silva Marcelo Z. do Nascimento Leandro A. Neves |
| author_facet | Jaqueline J. Tenguam Leonardo H. da Costa Longo Guilherme F. Roberto Thaína A. A. Tosta Paulo R. de Faria Adriano M. Loyola Sérgio V. Cardoso Adriano B. Silva Marcelo Z. do Nascimento Leandro A. Neves |
| author_sort | Jaqueline J. Tenguam |
| collection | DOAJ |
| description | In this paper, we propose an approach based on ensemble learning to classify histology tissues stained with hematoxylin and eosin. The proposal was applied to representative images of colorectal cancer, oral epithelial dysplasia, non-Hodgkin’s lymphoma, and liver tissues (the classification of gender and age from liver tissue samples). The ensemble learning considered multiple combinations of techniques that are commonly used to develop computer-aided diagnosis methods in medical imaging. The feature extraction was defined with different descriptors, exploring the deep learning and handcrafted methods. The deep-learned features were obtained using five different convolutional neural network architectures. The handcrafted features were representatives of fractal techniques (multidimensional and multiscale approaches), Haralick descriptors, and local binary patterns. A two-stage feature selection process (ranking with metaheuristics) was defined to obtain the main combinations of descriptors and, consequently, techniques. Each combination was tested through a rigorous ensemble process, exploring heterogeneous classifiers, such as Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, and Naive Bayes. The ensemble learning presented here provided accuracy rates from 90.72% to 100.00% and offered relevant information about the combinations of techniques in multiple histological images and the main features present in the top-performing solutions, using smaller sets of descriptors (limited to a maximum of 53), which involved each ensemble process and solutions that have not yet been explored. The developed methodology, i.e., making the knowledge of each ensemble learning comprehensible to specialists, complements the main contributions of this study to supporting the development of computer-aided diagnosis systems for histological images. |
| first_indexed | 2024-03-08T04:00:40Z |
| format | Article |
| id | doaj.art-8a07eeff21964a2294c7cedaaa84c888 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-08T04:00:40Z |
| publishDate | 2024-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-8a07eeff21964a2294c7cedaaa84c8882024-02-09T15:07:39ZengMDPI AGApplied Sciences2076-34172024-01-01143108410.3390/app14031084Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological ImagesJaqueline J. Tenguam0Leonardo H. da Costa Longo1Guilherme F. Roberto2Thaína A. A. Tosta3Paulo R. de Faria4Adriano M. Loyola5Sérgio V. Cardoso6Adriano B. Silva7Marcelo Z. do Nascimento8Leandro A. Neves9Department of Computer Science and Statistics (DCCE), São Paulo State University (UNESP), Rua Cristóvão Colombo, 2265, São José do Rio Preto 15054-000, São Paulo, BrazilDepartment of Computer Science and Statistics (DCCE), São Paulo State University (UNESP), Rua Cristóvão Colombo, 2265, São José do Rio Preto 15054-000, São Paulo, BrazilDepartment of Informatics Engineering, Faculty of Engineering, University of Porto, Dr. Roberto Frias, sn, 4200-465 Porto, PortugalScience and Technology Institute, Federal University of São Paulo (UNIFESP), Avenida Cesare Mansueto Giulio Lattes, 1201, São José dos Campos 12247-014, São Paulo, BrazilDepartment of Histology and Morphology, Institute of Biomedical Science, Federal University of Uberlândia (UFU), Av. Amazonas, S/N, Uberlândia 38405-320, Minas Gerais, BrazilArea of Oral Pathology, School of Dentistry, Federal University of Uberlândia (UFU), R. Ceará—Umuarama, Uberlândia 38402-018, Minas Gerais, BrazilArea of Oral Pathology, School of Dentistry, Federal University of Uberlândia (UFU), R. Ceará—Umuarama, Uberlândia 38402-018, Minas Gerais, BrazilFaculty of Computer Science (FACOM), Federal University of Uberlândia (UFU), Avenida João Naves de Ávila 2121, Bl.B, Uberlândia 38400-902, Minas Gerais, BrazilFaculty of Computer Science (FACOM), Federal University of Uberlândia (UFU), Avenida João Naves de Ávila 2121, Bl.B, Uberlândia 38400-902, Minas Gerais, BrazilDepartment of Computer Science and Statistics (DCCE), São Paulo State University (UNESP), Rua Cristóvão Colombo, 2265, São José do Rio Preto 15054-000, São Paulo, BrazilIn this paper, we propose an approach based on ensemble learning to classify histology tissues stained with hematoxylin and eosin. The proposal was applied to representative images of colorectal cancer, oral epithelial dysplasia, non-Hodgkin’s lymphoma, and liver tissues (the classification of gender and age from liver tissue samples). The ensemble learning considered multiple combinations of techniques that are commonly used to develop computer-aided diagnosis methods in medical imaging. The feature extraction was defined with different descriptors, exploring the deep learning and handcrafted methods. The deep-learned features were obtained using five different convolutional neural network architectures. The handcrafted features were representatives of fractal techniques (multidimensional and multiscale approaches), Haralick descriptors, and local binary patterns. A two-stage feature selection process (ranking with metaheuristics) was defined to obtain the main combinations of descriptors and, consequently, techniques. Each combination was tested through a rigorous ensemble process, exploring heterogeneous classifiers, such as Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, and Naive Bayes. The ensemble learning presented here provided accuracy rates from 90.72% to 100.00% and offered relevant information about the combinations of techniques in multiple histological images and the main features present in the top-performing solutions, using smaller sets of descriptors (limited to a maximum of 53), which involved each ensemble process and solutions that have not yet been explored. The developed methodology, i.e., making the knowledge of each ensemble learning comprehensible to specialists, complements the main contributions of this study to supporting the development of computer-aided diagnosis systems for histological images.https://www.mdpi.com/2076-3417/14/3/1084ensemble learninghandcrafted featuresdeep-learned featurestwo-stage feature selection methodhistological images |
| spellingShingle | Jaqueline J. Tenguam Leonardo H. da Costa Longo Guilherme F. Roberto Thaína A. A. Tosta Paulo R. de Faria Adriano M. Loyola Sérgio V. Cardoso Adriano B. Silva Marcelo Z. do Nascimento Leandro A. Neves Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images Applied Sciences ensemble learning handcrafted features deep-learned features two-stage feature selection method histological images |
| title | Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images |
| title_full | Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images |
| title_fullStr | Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images |
| title_full_unstemmed | Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images |
| title_short | Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&E Histological Images |
| title_sort | ensemble learning based solutions an approach for evaluating multiple features in the context of h e histological images |
| topic | ensemble learning handcrafted features deep-learned features two-stage feature selection method histological images |
| url | https://www.mdpi.com/2076-3417/14/3/1084 |
| work_keys_str_mv | AT jaquelinejtenguam ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT leonardohdacostalongo ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT guilhermefroberto ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT thainaaatosta ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT paulordefaria ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT adrianomloyola ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT sergiovcardoso ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT adrianobsilva ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT marcelozdonascimento ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages AT leandroaneves ensemblelearningbasedsolutionsanapproachforevaluatingmultiplefeaturesinthecontextofhehistologicalimages |