Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma
Colorectal cancer is one of the most lethal cancers because of late diagnosis and challenges in the selection of therapy options. The histopathological diagnosis of colon adenocarcinoma is hindered by poor reproducibility and a lack of standard examination protocols required for appropriate treatmen...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-01-01
|
| Series: | AI |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-2688/5/1/16 |
| _version_ | 1797242328295931904 |
|---|---|
| author | Marco Leo Pierluigi Carcagnì Luca Signore Francesco Corcione Giulio Benincasa Mikko O. Laukkanen Cosimo Distante |
| author_facet | Marco Leo Pierluigi Carcagnì Luca Signore Francesco Corcione Giulio Benincasa Mikko O. Laukkanen Cosimo Distante |
| author_sort | Marco Leo |
| collection | DOAJ |
| description | Colorectal cancer is one of the most lethal cancers because of late diagnosis and challenges in the selection of therapy options. The histopathological diagnosis of colon adenocarcinoma is hindered by poor reproducibility and a lack of standard examination protocols required for appropriate treatment decisions. In the current study, using state-of-the-art approaches on benchmark datasets, we analyzed different architectures and ensembling strategies to develop the most efficient network combinations to improve binary and ternary classification. We propose an innovative two-stage pipeline approach to diagnose colon adenocarcinoma grading from histological images in a similar manner to a pathologist. The glandular regions were first segmented by a transformer architecture with subsequent classification using a convolutional neural network (CNN) ensemble, which markedly improved the learning efficiency and shortened the learning time. Moreover, we prepared and published a dataset for clinical validation of the developed artificial neural network, which suggested the discovery of novel histological phenotypic alterations in adenocarcinoma sections that could have prognostic value. Therefore, AI could markedly improve the reproducibility, efficiency, and accuracy of colon cancer diagnosis, which are required for precision medicine to personalize the treatment of cancer patients. |
| first_indexed | 2024-04-24T18:37:28Z |
| format | Article |
| id | doaj.art-3456b8c2b13f42828f9f91a62dec52f5 |
| institution | Directory Open Access Journal |
| issn | 2673-2688 |
| language | English |
| last_indexed | 2024-04-24T18:37:28Z |
| publishDate | 2024-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | AI |
| spelling | doaj.art-3456b8c2b13f42828f9f91a62dec52f52024-03-27T13:17:11ZengMDPI AGAI2673-26882024-01-015132434110.3390/ai5010016Convolutional Neural Networks in the Diagnosis of Colon AdenocarcinomaMarco Leo0Pierluigi Carcagnì1Luca Signore2Francesco Corcione3Giulio Benincasa4Mikko O. Laukkanen5Cosimo Distante6Institute of Applied Sciences and Intelligent Systems (ISASI), National Research Council (CNR) of Italy, 73100 Lecce, ItalyInstitute of Applied Sciences and Intelligent Systems (ISASI), National Research Council (CNR) of Italy, 73100 Lecce, ItalyDipartimento di Ingegneria per L’Innovazione, Università del Salento, 73100 Lecce, ItalyClinica Mediterranea, 80122 Naples, ItalyItalo Foundation, 20146 Milano, ItalyDepartment of Translational Medical Sciences, University of Naples Federico II, 80131 Naples, ItalyInstitute of Applied Sciences and Intelligent Systems (ISASI), National Research Council (CNR) of Italy, 73100 Lecce, ItalyColorectal cancer is one of the most lethal cancers because of late diagnosis and challenges in the selection of therapy options. The histopathological diagnosis of colon adenocarcinoma is hindered by poor reproducibility and a lack of standard examination protocols required for appropriate treatment decisions. In the current study, using state-of-the-art approaches on benchmark datasets, we analyzed different architectures and ensembling strategies to develop the most efficient network combinations to improve binary and ternary classification. We propose an innovative two-stage pipeline approach to diagnose colon adenocarcinoma grading from histological images in a similar manner to a pathologist. The glandular regions were first segmented by a transformer architecture with subsequent classification using a convolutional neural network (CNN) ensemble, which markedly improved the learning efficiency and shortened the learning time. Moreover, we prepared and published a dataset for clinical validation of the developed artificial neural network, which suggested the discovery of novel histological phenotypic alterations in adenocarcinoma sections that could have prognostic value. Therefore, AI could markedly improve the reproducibility, efficiency, and accuracy of colon cancer diagnosis, which are required for precision medicine to personalize the treatment of cancer patients.https://www.mdpi.com/2673-2688/5/1/16colon cancerhistological diagnosisartificial intelligencedeep learningtransformer networksdataset |
| spellingShingle | Marco Leo Pierluigi Carcagnì Luca Signore Francesco Corcione Giulio Benincasa Mikko O. Laukkanen Cosimo Distante Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma AI colon cancer histological diagnosis artificial intelligence deep learning transformer networks dataset |
| title | Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma |
| title_full | Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma |
| title_fullStr | Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma |
| title_full_unstemmed | Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma |
| title_short | Convolutional Neural Networks in the Diagnosis of Colon Adenocarcinoma |
| title_sort | convolutional neural networks in the diagnosis of colon adenocarcinoma |
| topic | colon cancer histological diagnosis artificial intelligence deep learning transformer networks dataset |
| url | https://www.mdpi.com/2673-2688/5/1/16 |
| work_keys_str_mv | AT marcoleo convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT pierluigicarcagni convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT lucasignore convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT francescocorcione convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT giuliobenincasa convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT mikkoolaukkanen convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma AT cosimodistante convolutionalneuralnetworksinthediagnosisofcolonadenocarcinoma |