Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification
Hyperspectral imaging techniques (HSI) do not require contact with patients and are non-ionizing as well as non-invasive. As a consequence, they have been extensively applied in the medical field. HSI is being combined with machine learning (ML) processes to obtain models to assist in diagnosis. In...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/11/3827 |
| _version_ | 1827690949554208768 |
|---|---|
| author | Gemma Urbanos Alberto Martín Guillermo Vázquez Marta Villanueva Manuel Villa Luis Jimenez-Roldan Miguel Chavarrías Alfonso Lagares Eduardo Juárez César Sanz |
| author_facet | Gemma Urbanos Alberto Martín Guillermo Vázquez Marta Villanueva Manuel Villa Luis Jimenez-Roldan Miguel Chavarrías Alfonso Lagares Eduardo Juárez César Sanz |
| author_sort | Gemma Urbanos |
| collection | DOAJ |
| description | Hyperspectral imaging techniques (HSI) do not require contact with patients and are non-ionizing as well as non-invasive. As a consequence, they have been extensively applied in the medical field. HSI is being combined with machine learning (ML) processes to obtain models to assist in diagnosis. In particular, the combination of these techniques has proven to be a reliable aid in the differentiation of healthy and tumor tissue during brain tumor surgery. ML algorithms such as support vector machine (SVM), random forest (RF) and convolutional neural networks (CNN) are used to make predictions and provide in-vivo visualizations that may assist neurosurgeons in being more precise, hence reducing damages to healthy tissue. In this work, thirteen in-vivo hyperspectral images from twelve different patients with high-grade gliomas (grade III and IV) have been selected to train SVM, RF and CNN classifiers. Five different classes have been defined during the experiments: healthy tissue, tumor, venous blood vessel, arterial blood vessel and dura mater. Overall accuracy (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>O</mi><mi>A</mi><mi>C</mi><mi>C</mi></mrow></semantics></math></inline-formula>) results vary from 60% to 95% depending on the training conditions. Finally, as far as the contribution of each band to the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>O</mi><mi>A</mi><mi>C</mi><mi>C</mi></mrow></semantics></math></inline-formula> is concerned, the results obtained in this work are 3.81 times greater than those reported in the literature. |
| first_indexed | 2024-03-10T10:50:08Z |
| format | Article |
| id | doaj.art-c3c6b2de9ed649da9ca64abfa3e8dbc1 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T10:50:08Z |
| publishDate | 2021-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-c3c6b2de9ed649da9ca64abfa3e8dbc12023-11-21T22:20:59ZengMDPI AGSensors1424-82202021-05-012111382710.3390/s21113827Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer ClassificationGemma Urbanos0Alberto Martín1Guillermo Vázquez2Marta Villanueva3Manuel Villa4Luis Jimenez-Roldan5Miguel Chavarrías6Alfonso Lagares7Eduardo Juárez8César Sanz9Research Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainInstituto de Investigación Sanitaria Hospital 12 de Octubre (Imas12), 28041 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainInstituto de Investigación Sanitaria Hospital 12 de Octubre (Imas12), 28041 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainResearch Center on Software Technologies and Multimedia Systems (CITSEM), Campus Sur UPM, Universidad Politécnica de Madrid (UPM), 28031 Madrid, SpainHyperspectral imaging techniques (HSI) do not require contact with patients and are non-ionizing as well as non-invasive. As a consequence, they have been extensively applied in the medical field. HSI is being combined with machine learning (ML) processes to obtain models to assist in diagnosis. In particular, the combination of these techniques has proven to be a reliable aid in the differentiation of healthy and tumor tissue during brain tumor surgery. ML algorithms such as support vector machine (SVM), random forest (RF) and convolutional neural networks (CNN) are used to make predictions and provide in-vivo visualizations that may assist neurosurgeons in being more precise, hence reducing damages to healthy tissue. In this work, thirteen in-vivo hyperspectral images from twelve different patients with high-grade gliomas (grade III and IV) have been selected to train SVM, RF and CNN classifiers. Five different classes have been defined during the experiments: healthy tissue, tumor, venous blood vessel, arterial blood vessel and dura mater. Overall accuracy (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>O</mi><mi>A</mi><mi>C</mi><mi>C</mi></mrow></semantics></math></inline-formula>) results vary from 60% to 95% depending on the training conditions. Finally, as far as the contribution of each band to the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>O</mi><mi>A</mi><mi>C</mi><mi>C</mi></mrow></semantics></math></inline-formula> is concerned, the results obtained in this work are 3.81 times greater than those reported in the literature.https://www.mdpi.com/1424-8220/21/11/3827hyperspectral imagingmachine learningclassificationsupport vector machinerandom forestconvolutional neural network |
| spellingShingle | Gemma Urbanos Alberto Martín Guillermo Vázquez Marta Villanueva Manuel Villa Luis Jimenez-Roldan Miguel Chavarrías Alfonso Lagares Eduardo Juárez César Sanz Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification Sensors hyperspectral imaging machine learning classification support vector machine random forest convolutional neural network |
| title | Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification |
| title_full | Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification |
| title_fullStr | Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification |
| title_full_unstemmed | Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification |
| title_short | Supervised Machine Learning Methods and Hyperspectral Imaging Techniques Jointly Applied for Brain Cancer Classification |
| title_sort | supervised machine learning methods and hyperspectral imaging techniques jointly applied for brain cancer classification |
| topic | hyperspectral imaging machine learning classification support vector machine random forest convolutional neural network |
| url | https://www.mdpi.com/1424-8220/21/11/3827 |
| work_keys_str_mv | AT gemmaurbanos supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT albertomartin supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT guillermovazquez supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT martavillanueva supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT manuelvilla supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT luisjimenezroldan supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT miguelchavarrias supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT alfonsolagares supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT eduardojuarez supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification AT cesarsanz supervisedmachinelearningmethodsandhyperspectralimagingtechniquesjointlyappliedforbraincancerclassification |