Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks
Hyperspectral imaging (HSI) technology has demonstrated potential to provide useful information about the chemical composition of tissue and its morphological features in a single image modality. Deep learning (DL) techniques have demonstrated the ability of automatic feature extraction from data fo...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/7/1911 |
| _version_ | 1797571967123980288 |
|---|---|
| author | Samuel Ortega Martin Halicek Himar Fabelo Rafael Camacho María de la Luz Plaza Fred Godtliebsen Gustavo M. Callicó Baowei Fei |
| author_facet | Samuel Ortega Martin Halicek Himar Fabelo Rafael Camacho María de la Luz Plaza Fred Godtliebsen Gustavo M. Callicó Baowei Fei |
| author_sort | Samuel Ortega |
| collection | DOAJ |
| description | Hyperspectral imaging (HSI) technology has demonstrated potential to provide useful information about the chemical composition of tissue and its morphological features in a single image modality. Deep learning (DL) techniques have demonstrated the ability of automatic feature extraction from data for a successful classification. In this study, we exploit HSI and DL for the automatic differentiation of glioblastoma (GB) and non-tumor tissue on hematoxylin and eosin (H&E) stained histological slides of human brain tissue. GB detection is a challenging application, showing high heterogeneity in the cellular morphology across different patients. We employed an HSI microscope, with a spectral range from 400 to 1000 nm, to collect 517 HS cubes from 13 GB patients using 20× magnification. Using a convolutional neural network (CNN), we were able to automatically detect GB within the pathological slides, achieving average sensitivity and specificity values of 88% and 77%, respectively, representing an improvement of 7% and 8% respectively, as compared to the results obtained using RGB (red, green, and blue) images. This study demonstrates that the combination of hyperspectral microscopic imaging and deep learning is a promising tool for future computational pathologies. |
| first_indexed | 2024-03-10T20:48:00Z |
| format | Article |
| id | doaj.art-8d9bf8926ed44b0692266a34a4de9e2e |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T20:48:00Z |
| publishDate | 2020-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-8d9bf8926ed44b0692266a34a4de9e2e2023-11-19T20:08:18ZengMDPI AGSensors1424-82202020-03-01207191110.3390/s20071911Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural NetworksSamuel Ortega0Martin Halicek1Himar Fabelo2Rafael Camacho3María de la Luz Plaza4Fred Godtliebsen5Gustavo M. Callicó6Baowei Fei7Quantitative Bioimaging Laboratory, Department of Bioengineering, The University of Texas at Dallas, Richardson, TX 75080, USAQuantitative Bioimaging Laboratory, Department of Bioengineering, The University of Texas at Dallas, Richardson, TX 75080, USAInstitute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria (ULPGC), 35017 Las Palmas de Gran Canaria, SpainDepartment of Pathological Anatomy, University Hospital Doctor Negrin of Gran Canaria, Barranco de la Ballena s/n, 35010 Las Palmas de Gran Canaria, SpainDepartment of Pathological Anatomy, University Hospital Doctor Negrin of Gran Canaria, Barranco de la Ballena s/n, 35010 Las Palmas de Gran Canaria, SpainDepartment of Mathematics and Statistics, UiT The Artic, University of Norway, Hansine Hansens veg 18, 9019 Tromsø, NorwayInstitute for Applied Microelectronics (IUMA), University of Las Palmas de Gran Canaria (ULPGC), 35017 Las Palmas de Gran Canaria, SpainQuantitative Bioimaging Laboratory, Department of Bioengineering, The University of Texas at Dallas, Richardson, TX 75080, USAHyperspectral imaging (HSI) technology has demonstrated potential to provide useful information about the chemical composition of tissue and its morphological features in a single image modality. Deep learning (DL) techniques have demonstrated the ability of automatic feature extraction from data for a successful classification. In this study, we exploit HSI and DL for the automatic differentiation of glioblastoma (GB) and non-tumor tissue on hematoxylin and eosin (H&E) stained histological slides of human brain tissue. GB detection is a challenging application, showing high heterogeneity in the cellular morphology across different patients. We employed an HSI microscope, with a spectral range from 400 to 1000 nm, to collect 517 HS cubes from 13 GB patients using 20× magnification. Using a convolutional neural network (CNN), we were able to automatically detect GB within the pathological slides, achieving average sensitivity and specificity values of 88% and 77%, respectively, representing an improvement of 7% and 8% respectively, as compared to the results obtained using RGB (red, green, and blue) images. This study demonstrates that the combination of hyperspectral microscopic imaging and deep learning is a promising tool for future computational pathologies.https://www.mdpi.com/1424-8220/20/7/1911hyperspectral imagingmedical optics and biotechnologyoptical pathologyconvolutional neural networkstissue diagnosticstissue characterization |
| spellingShingle | Samuel Ortega Martin Halicek Himar Fabelo Rafael Camacho María de la Luz Plaza Fred Godtliebsen Gustavo M. Callicó Baowei Fei Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks Sensors hyperspectral imaging medical optics and biotechnology optical pathology convolutional neural networks tissue diagnostics tissue characterization |
| title | Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks |
| title_full | Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks |
| title_fullStr | Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks |
| title_full_unstemmed | Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks |
| title_short | Hyperspectral Imaging for the Detection of Glioblastoma Tumor Cells in H&E Slides Using Convolutional Neural Networks |
| title_sort | hyperspectral imaging for the detection of glioblastoma tumor cells in h e slides using convolutional neural networks |
| topic | hyperspectral imaging medical optics and biotechnology optical pathology convolutional neural networks tissue diagnostics tissue characterization |
| url | https://www.mdpi.com/1424-8220/20/7/1911 |
| work_keys_str_mv | AT samuelortega hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT martinhalicek hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT himarfabelo hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT rafaelcamacho hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT mariadelaluzplaza hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT fredgodtliebsen hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT gustavomcallico hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks AT baoweifei hyperspectralimagingforthedetectionofglioblastomatumorcellsinheslidesusingconvolutionalneuralnetworks |