Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization
The prevailing approach for three-dimensional (3D) medical image segmentation is to use convolutional networks. Recently, deep learning methods have achieved human-level performance in several important applied problems, such as volumetry for lung-cancer diagnosis or delineation for radiation therap...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-02-01
|
| Series: | Journal of Imaging |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-433X/7/2/35 |
| _version_ | 1797396750753857536 |
|---|---|
| author | Boris Shirokikh Alexey Shevtsov Alexandra Dalechina Egor Krivov Valery Kostjuchenko Andrey Golanov Victor Gombolevskiy Sergey Morozov Mikhail Belyaev |
| author_facet | Boris Shirokikh Alexey Shevtsov Alexandra Dalechina Egor Krivov Valery Kostjuchenko Andrey Golanov Victor Gombolevskiy Sergey Morozov Mikhail Belyaev |
| author_sort | Boris Shirokikh |
| collection | DOAJ |
| description | The prevailing approach for three-dimensional (3D) medical image segmentation is to use convolutional networks. Recently, deep learning methods have achieved human-level performance in several important applied problems, such as volumetry for lung-cancer diagnosis or delineation for radiation therapy planning. However, state-of-the-art architectures, such as U-Net and DeepMedic, are computationally heavy and require workstations accelerated with graphics processing units for fast inference. However, scarce research has been conducted concerning enabling fast central processing unit computations for such networks. Our paper fills this gap. We propose a new segmentation method with a human-like technique to segment a 3D study. First, we analyze the image at a small scale to identify areas of interest and then process only relevant feature-map patches. Our method not only reduces the inference time from 10 min to 15 s but also preserves state-of-the-art segmentation quality, as we illustrate in the set of experiments with two large datasets. |
| first_indexed | 2024-03-09T00:56:09Z |
| format | Article |
| id | doaj.art-f9d1a620d9b64861a3b933029d834a89 |
| institution | Directory Open Access Journal |
| issn | 2313-433X |
| language | English |
| last_indexed | 2024-03-09T00:56:09Z |
| publishDate | 2021-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Imaging |
| spelling | doaj.art-f9d1a620d9b64861a3b933029d834a892023-12-11T16:55:37ZengMDPI AGJournal of Imaging2313-433X2021-02-01723510.3390/jimaging7020035Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target LocalizationBoris Shirokikh0Alexey Shevtsov1Alexandra Dalechina2Egor Krivov3Valery Kostjuchenko4Andrey Golanov5Victor Gombolevskiy6Sergey Morozov7Mikhail Belyaev8Center for Neurobiology and Brain Restoration, Skolkovo Institute of Science and Technology, 121205 Moscow, RussiaCenter for Neurobiology and Brain Restoration, Skolkovo Institute of Science and Technology, 121205 Moscow, RussiaMoscow Gamma-Knife Center, 125047 Moscow, RussiaSector of Data Analysis for Neuroscience, Kharkevich Institute for Information Transmission Problems, 127051 Moscow, RussiaMoscow Gamma-Knife Center, 125047 Moscow, RussiaDepartment of Radiosurgery and Radiation, Burdenko Neurosurgery Institute, 125047 Moscow, RussiaMedical Research Department, Research and Practical Clinical Center of Diagnostics and Telemedicine Technologies of the Department of Health Care of Moscow, 127051 Moscow, RussiaMedical Research Department, Research and Practical Clinical Center of Diagnostics and Telemedicine Technologies of the Department of Health Care of Moscow, 127051 Moscow, RussiaCenter for Neurobiology and Brain Restoration, Skolkovo Institute of Science and Technology, 121205 Moscow, RussiaThe prevailing approach for three-dimensional (3D) medical image segmentation is to use convolutional networks. Recently, deep learning methods have achieved human-level performance in several important applied problems, such as volumetry for lung-cancer diagnosis or delineation for radiation therapy planning. However, state-of-the-art architectures, such as U-Net and DeepMedic, are computationally heavy and require workstations accelerated with graphics processing units for fast inference. However, scarce research has been conducted concerning enabling fast central processing unit computations for such networks. Our paper fills this gap. We propose a new segmentation method with a human-like technique to segment a 3D study. First, we analyze the image at a small scale to identify areas of interest and then process only relevant feature-map patches. Our method not only reduces the inference time from 10 min to 15 s but also preserves state-of-the-art segmentation quality, as we illustrate in the set of experiments with two large datasets.https://www.mdpi.com/2313-433X/7/2/35deep learningmedical image segmentationcomputed tomography (CT)magnetic resonance imaging (MRI) |
| spellingShingle | Boris Shirokikh Alexey Shevtsov Alexandra Dalechina Egor Krivov Valery Kostjuchenko Andrey Golanov Victor Gombolevskiy Sergey Morozov Mikhail Belyaev Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization Journal of Imaging deep learning medical image segmentation computed tomography (CT) magnetic resonance imaging (MRI) |
| title | Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization |
| title_full | Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization |
| title_fullStr | Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization |
| title_full_unstemmed | Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization |
| title_short | Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization |
| title_sort | accelerating 3d medical image segmentation by adaptive small scale target localization |
| topic | deep learning medical image segmentation computed tomography (CT) magnetic resonance imaging (MRI) |
| url | https://www.mdpi.com/2313-433X/7/2/35 |
| work_keys_str_mv | AT borisshirokikh accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT alexeyshevtsov accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT alexandradalechina accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT egorkrivov accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT valerykostjuchenko accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT andreygolanov accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT victorgombolevskiy accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT sergeymorozov accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization AT mikhailbelyaev accelerating3dmedicalimagesegmentationbyadaptivesmallscaletargetlocalization |