Predicting standardized uptake value of brown adipose tissue from CT scans using convolutional neural networks

The standard method for identifying active Brown Adipose Tissue (BAT) is [<sup>18</sup>F]-Fluorodeoxyglucose ([<sup>18</sup>F]-FDG) PET/CT imaging, which is costly and exposes patients to radiation, making it impractical for population studies. These issues can be addressed with computational methods that predict [<sup>18</sup>F]-FDG uptake by BAT from CT; earlier population studies pave the way for developing such methods by showing some correlation between the Hounsfield Unit (HU) of BAT in CT and the corresponding [<sup>18</sup>F]-FDG uptake in PET. In this study, we propose training convolutional neural networks (CNNs) to predict [<sup>18</sup>F]-FDG uptake by BAT from unenhanced CT scans in the restricted regions that are likely to contain BAT. Using the Attention U-Net architecture, we perform experiments on datasets from four different cohorts, the largest study to date. We segment BAT regions using predicted [<sup>18</sup>F]-FDG uptake values, achieving 23% to 40% better accuracy than conventional CT thresholding. Additionally, BAT volumes computed from the segmentations distinguish the subjects with and without active BAT with an AUC of 0.8, compared to 0.6 for CT thresholding. These findings suggest CNNs can facilitate large-scale imaging studies more efficiently and cost-effectively using only CT.

Authors

Institutions

Publication Details

Journal
Nature Communications
Published
2024-09-27
DOI
https://doi.org/10.1038/s41467-024-52622-w
Citations
2
Primary Topic
Adipose Tissue and Metabolism
Type
article
Field-Weighted Citation Impact
0.65

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Predicting standardized uptake value of brown adipose tissue from CT scans using convolutional neural networks

Thomas Sartoretti, Hebert Alberto Vargas, Anton S. Becker, Jonatan R. Ruiz et al.
2 citations
Nature Communications
Adipose Tissue and Metabolism
0.65
article

Predicting standardized uptake value of brown adipose tissue from CT scans using convolutional neural networks

Thomas Sartoretti, Hebert Alberto Vargas, Anton S. Becker, Jonatan R. Ruiz, Claudia Irene Maushart, Christian Wolfrum, Matthias Johannes Betz, Borja Martínez‐Téllez, Damian Wild, Moritz Schwyzer, Ender Konukoğlu, Irene A. Burger, Alin Chirindel, Bart Deplancke, Vincent Gardeux, Ertunç Erdil, Nicola Zamboni
article en
2 citations

Abstract

The standard method for identifying active Brown Adipose Tissue (BAT) is [<sup>18</sup>F]-Fluorodeoxyglucose ([<sup>18</sup>F]-FDG) PET/CT imaging, which is costly and exposes patients to radiation, making it impractical for population studies. These issues can be addressed with computational methods that predict [<sup>18</sup>F]-FDG uptake by BAT from CT; earlier population studies pave the way for developing such methods by showing some correlation between the Hounsfield Unit (HU) of BAT in CT and the corresponding [<sup>18</sup>F]-FDG uptake in PET. In this study, we propose training convolutional neural networks (CNNs) to predict [<sup>18</sup>F]-FDG uptake by BAT from unenhanced CT scans in the restricted regions that are likely to contain BAT. Using the Attention U-Net architecture, we perform experiments on datasets from four different cohorts, the largest study to date. We segment BAT regions using predicted [<sup>18</sup>F]-FDG uptake values, achieving 23% to 40% better accuracy than conventional CT thresholding. Additionally, BAT volumes computed from the segmentations distinguish the subjects with and without active BAT with an AUC of 0.8, compared to 0.6 for CT thresholding. These findings suggest CNNs can facilitate large-scale imaging studies more efficiently and cost-effectively using only CT.

Nature CommunicationsVol. 15(1)
SIB Swiss Institute of Bioinformatics (CH), Memorial Sloan Kettering Cancer Center (US), Universidad de Granada (ES), University of Basel (CH), University of Zurich (CH), Leiden University Medical Center (NL), Instituto de Salud Carlos III (ES), University Hospital of Basel (CH), ETH Zurich (CH), University Hospital of Zurich (CH), Instituto de Investigación Biosanitaria de Granada (ES), Centro de Investigación Biomédica en Red (ES), École Polytechnique Fédérale de Lausanne (CH), University of Almería (ES), New York University (US)
National Science Foundation, Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung, Eidgenössische Technische Hochschule Zürich, Bentham-Moxon Trust, Agencia Estatal de Investigación, European Social Fund
Zero hunger
Openalex Percentile: Top 29%
Adipose Tissue and Metabolism
0.65
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.