On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification
The aim of this work was to test microwave brain stroke detection and classification using support vector machines (SVMs). We tested how the nature and variability of training data and system parameters impact the achieved classification accuracy. Using experimentally verified numerical models, a la...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/4/2031 |
| _version_ | 1797618287089025024 |
|---|---|
| author | Tomas Pokorny Jan Vrba Ondrej Fiser David Vrba Tomas Drizdal Marek Novak Luca Tosi Alessandro Polo Marco Salucci |
| author_facet | Tomas Pokorny Jan Vrba Ondrej Fiser David Vrba Tomas Drizdal Marek Novak Luca Tosi Alessandro Polo Marco Salucci |
| author_sort | Tomas Pokorny |
| collection | DOAJ |
| description | The aim of this work was to test microwave brain stroke detection and classification using support vector machines (SVMs). We tested how the nature and variability of training data and system parameters impact the achieved classification accuracy. Using experimentally verified numerical models, a large database of synthetic training and test data was created. The models consist of an antenna array surrounding reconfigurable geometrically and dielectrically realistic human head phantoms with virtually inserted strokes of arbitrary size, and different dielectric parameters in different positions. The generated synthetic data sets were used to test four different hypotheses, regarding the appropriate parameters of the training dataset, the appropriate frequency range and the number of frequency points, as well as the level of subject variability to reach the highest SVM classification accuracy. The results indicate that the SVM algorithm is able to detect the presence of the stroke and classify it (i.e., ischemic or hemorrhagic) even when trained with single-frequency data. Moreover, it is shown that data of subjects with smaller strokes appear to be the most suitable for training accurate SVM predictors with high generalization capabilities. Finally, the datasets created for this study are made available to the community for testing and developing their own algorithms. |
| first_indexed | 2024-03-11T08:10:56Z |
| format | Article |
| id | doaj.art-7f4178e9675442adab2b1ea629bc9f9a |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T08:10:56Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-7f4178e9675442adab2b1ea629bc9f9a2023-11-16T23:09:25ZengMDPI AGSensors1424-82202023-02-01234203110.3390/s23042031On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and ClassificationTomas Pokorny0Jan Vrba1Ondrej Fiser2David Vrba3Tomas Drizdal4Marek Novak5Luca Tosi6Alessandro Polo7Marco Salucci8Department of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicDepartment of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicDepartment of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicDepartment of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicDepartment of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicDepartment of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in Prague, 166 00 Prague, Czech RepublicELEDIA Research Center (ELEDIA@UniTN—University of Trento), DICAM—Department of Civil, Environmental, and Mechanical Engineering, Via Mesiano 77, 38123 Trento, ItalyELEDIA Research Center (ELEDIA@UniTN—University of Trento), DICAM—Department of Civil, Environmental, and Mechanical Engineering, Via Mesiano 77, 38123 Trento, ItalyELEDIA Research Center (ELEDIA@UniTN—University of Trento), DICAM—Department of Civil, Environmental, and Mechanical Engineering, Via Mesiano 77, 38123 Trento, ItalyThe aim of this work was to test microwave brain stroke detection and classification using support vector machines (SVMs). We tested how the nature and variability of training data and system parameters impact the achieved classification accuracy. Using experimentally verified numerical models, a large database of synthetic training and test data was created. The models consist of an antenna array surrounding reconfigurable geometrically and dielectrically realistic human head phantoms with virtually inserted strokes of arbitrary size, and different dielectric parameters in different positions. The generated synthetic data sets were used to test four different hypotheses, regarding the appropriate parameters of the training dataset, the appropriate frequency range and the number of frequency points, as well as the level of subject variability to reach the highest SVM classification accuracy. The results indicate that the SVM algorithm is able to detect the presence of the stroke and classify it (i.e., ischemic or hemorrhagic) even when trained with single-frequency data. Moreover, it is shown that data of subjects with smaller strokes appear to be the most suitable for training accurate SVM predictors with high generalization capabilities. Finally, the datasets created for this study are made available to the community for testing and developing their own algorithms.https://www.mdpi.com/1424-8220/23/4/2031SVMbrain strokemicrowave devicesnumerical model |
| spellingShingle | Tomas Pokorny Jan Vrba Ondrej Fiser David Vrba Tomas Drizdal Marek Novak Luca Tosi Alessandro Polo Marco Salucci On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification Sensors SVM brain stroke microwave devices numerical model |
| title | On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification |
| title_full | On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification |
| title_fullStr | On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification |
| title_full_unstemmed | On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification |
| title_short | On the Role of Training Data for SVM-Based Microwave Brain Stroke Detection and Classification |
| title_sort | on the role of training data for svm based microwave brain stroke detection and classification |
| topic | SVM brain stroke microwave devices numerical model |
| url | https://www.mdpi.com/1424-8220/23/4/2031 |
| work_keys_str_mv | AT tomaspokorny ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT janvrba ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT ondrejfiser ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT davidvrba ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT tomasdrizdal ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT mareknovak ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT lucatosi ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT alessandropolo ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification AT marcosalucci ontheroleoftrainingdataforsvmbasedmicrowavebrainstrokedetectionandclassification |