Tumor Segmentation and Feature Extraction from Whole-Body FDG-PET/CT Using Cascaded 2D and 3D Convolutional Neural Netwo
- PDF / 787,057 Bytes
- 7 Pages / 595.276 x 790.866 pts Page_size
- 21 Downloads / 204 Views
ORIGINAL PAPER
Tumor Segmentation and Feature Extraction from Whole-Body FDG-PET/CT Using Cascaded 2D and 3D Convolutional Neural Networks Skander Jemaa 1 & Jill Fredrickson 1 & Richard A. D. Carano 1 & Tina Nielsen 2 & Alex de Crespigny 1 & Thomas Bengtsson 1
# The Author(s) 2020
Abstract 18 F-Fluorodeoxyglucose-positron emission tomography (FDG-PET) is commonly used in clinical practice and clinical drug development to identify and quantify metabolically active tumors. Manual or computer-assisted tumor segmentation in FDGPET images is a common way to assess tumor burden, such approaches are both labor intensive and may suffer from high interreader variability. We propose an end-to-end method leveraging 2D and 3D convolutional neural networks to rapidly identify and segment tumors and to extract metabolic information in eyes to thighs (whole body) FDG-PET/CT scans. The developed architecture is computationally efficient and devised to accommodate the size of whole-body scans, the extreme imbalance between tumor burden and the volume of healthy tissue, and the heterogeneous nature of the input images. Our dataset consists of a total of 3664 eyes to thighs FDG-PET/CT scans, from multi-site clinical trials in patients with non-Hodgkin’s lymphoma (NHL) and advanced non-small cell lung cancer (NSCLC). Tumors were segmented and reviewed by board-certified radiologists. We report a mean 3D Dice score of 88.6% on an NHL hold-out set of 1124 scans and a 93% sensitivity on 274 NSCLC hold-out scans. The method is a potential tool for radiologists to rapidly assess eyes to thighs FDG-avid tumor burden. Keywords FDG-PET . Deep learning . Lymphoma . DLBCL . NHL . Tumor segmentation
Introduction 18
F-fluorodeoxyglucose positron-emission tomography (FDG-PET) is a widely used imaging modality in oncology, where radiolabeled glucose is intravenously administered and is rapidly taken up by metabolically active tumors. This imaging technology provides a means to visualize and quantify metabolically active tumor burden in patients, and FDG-PET has been applied to a wide range of cancer types, with differing degrees of FDG uptake. Some tumors (e.g., prostate cancer) exhibit relatively low FDG uptake and, thus, may not be detectable by FDG-PET, whereas many other tumor types (e.g., non-small cell lung cancer, non-Hodgkin’s lymphoma) demonstrate high FDG uptake, making them highly visible in FDG-PET images [1]. FDG-PET has been found to be
* Skander Jemaa [email protected] 1
Genentech, Inc., South San Francisco, CA, USA
2
F. Hoffman-La Roche Ltd., Basel, Switzerland
superior to anatomical imaging modalities (cf., MRI, CT, US) for detection of these FDG-avid tumors [1] and FDGPET tumor burden metrics [2] have been shown to be prognostic of clinical outcome [1, 3], Moreover, FDG-PET imaging may provide an early indicator of therapeutic efficacy and is an established modality in the assessment of response to treatment in patients with malignant lymphomas [4–6]. Analysis and interpretation of FDG-PET images is performed by trained ra
Data Loading...
