1
|
Shen H, He P, Ren Y, Huang Z, Li S, Wang G, Cong M, Luo D, Shao D, Lee EYP, Cui R, Huo L, Qin J, Liu J, Hu Z, Liu Z, Zhang N. A deep learning model based on the attention mechanism for automatic segmentation of abdominal muscle and fat for body composition assessment. Quant Imaging Med Surg 2023; 13:1384-1398. [PMID: 36915346 PMCID: PMC10006126 DOI: 10.21037/qims-22-330] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/06/2022] [Accepted: 11/27/2022] [Indexed: 02/12/2023]
Abstract
Background Quantitative muscle and fat data obtained through body composition analysis are expected to be a new stable biomarker for the early and accurate prediction of treatment-related toxicity, treatment response, and prognosis in patients with lung cancer. The use of these biomarkers can enable the adjustment of individualized treatment regimens in a timely manner, which is critical to further improving patient prognosis and quality of life. We aimed to develop a deep learning model based on attention for fully automated segmentation of the abdomen from computed tomography (CT) to quantify body composition. Methods A fully automatic segmentation deep learning model was designed based on the attention mechanism and using U-Net as the framework. Subcutaneous fat, skeletal muscle, and visceral fat were manually segmented by two experts to serve as ground truth labels. The performance of the model was evaluated using Dice similarity coefficients (DSCs) and Hausdorff distance at 95th percentile (HD95). Results The mean DSC for subcutaneous fat and skeletal muscle were high for both the enhanced CT test set (0.93±0.06 and 0.96±0.02, respectively) and the plain CT test set (0.90±0.09 and 0.95±0.01, respectively). Nevertheless, the model did not perform well in the segmentation performance of visceral fat, especially for the enhanced CT test set. The mean DSC for the enhanced CT test set was 0.87±0.11, while the mean DSC for the plain CT test set was 0.92±0.03. We discuss the reasons for this result. Conclusions This work demonstrates a method for the automatic outlining of subcutaneous fat, skeletal muscle, and visceral fat areas at L3.
Collapse
Affiliation(s)
- Hao Shen
- Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China.,University of Chinese Academy of Sciences, Beijing, China
| | - Pin He
- Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China
| | - Ya Ren
- Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China
| | - Zhengyong Huang
- Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China.,University of Chinese Academy of Sciences, Beijing, China
| | - Shuluan Li
- Department of Medical Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China
| | - Guoshuai Wang
- Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China.,University of Chinese Academy of Sciences, Beijing, China
| | - Minghua Cong
- Department of Comprehensive Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Dehong Luo
- Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China
| | - Dan Shao
- Department of Nuclear Medicine, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Guangzhou, China
| | - Elaine Yuen-Phin Lee
- Department of Diagnostic Radiology, Clinical School of Medicine, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong, China
| | - Ruixue Cui
- Nuclear Medicine Department, State Key Laboratory of Complex Severe and Rare Diseases, Center for Rare Diseases Research, Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China
| | - Li Huo
- Nuclear Medicine Department, State Key Laboratory of Complex Severe and Rare Diseases, Center for Rare Diseases Research, Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China
| | - Jing Qin
- Centre for Smart Health, School of Nursing, The Hong Kong Polytechnic University, Hong Kong, China
| | - Jun Liu
- Department of Radiology, The Second Xiangya Hospital, Central South University, Changsha, China
| | - Zhanli Hu
- Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Zhou Liu
- Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen, China
| | - Na Zhang
- Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| |
Collapse
|
2
|
Alavi DH, Sakinis T, Henriksen HB, Beichmann B, Fløtten A, Blomhoff R, Lauritzen PM. Body composition assessment by artificial intelligence from routine computed tomography scans in colorectal cancer: Introducing BodySegAI. JCSM CLINICAL REPORTS 2022. [DOI: 10.1002/crt2.53] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/06/2022] Open
Affiliation(s)
- Dena Helene Alavi
- Department of Nutrition, Institute of Basic Medical Sciences University of Oslo Oslo Norway
| | - Tomas Sakinis
- Medical Division, Radiology and Nuclear Medicine, Neuroimaging Research Group Oslo University Hospital Oslo Norway
| | - Hege Berg Henriksen
- Department of Nutrition, Institute of Basic Medical Sciences University of Oslo Oslo Norway
| | - Benedicte Beichmann
- Department of Nutrition, Institute of Basic Medical Sciences University of Oslo Oslo Norway
| | - Ann‐Monica Fløtten
- Division of Radiology and Nuclear Medicine Oslo University Hospital Oslo Norway
| | - Rune Blomhoff
- Department of Nutrition, Institute of Basic Medical Sciences University of Oslo Oslo Norway
- Department of Clinical Service, Division of Cancer Medicine Oslo University Hospital Oslo Norway
| | | |
Collapse
|
3
|
Jones CK, Wang G, Yedavalli V, Sair H. Direct quantification of epistemic and aleatoric uncertainty in 3D U-net segmentation. J Med Imaging (Bellingham) 2022; 9:034002. [DOI: 10.1117/1.jmi.9.3.034002] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/21/2021] [Accepted: 05/18/2022] [Indexed: 11/14/2022] Open
Affiliation(s)
- Craig K. Jones
- Johns Hopkins University, Malone Center for Engineering in Healthcare, Baltimore, Maryland
| | - Guoqing Wang
- Johns Hopkins Bloomberg School of Public Health, Department of Biostatistics, Baltimore, Maryland
| | - Vivek Yedavalli
- Johns Hopkins Hospital, Russell H. Morgan Department of Radiology and Radiological Sciences, Baltimo
| | - Haris Sair
- Johns Hopkins University, Malone Center for Engineering in Healthcare, Baltimore, Maryland
| |
Collapse
|
4
|
McSweeney DM, Henderson EG, van Herk M, Weaver J, Bromiley PA, Green A, McWilliam A. Transfer learning for data-efficient abdominal muscle segmentation with convolutional neural networks. Med Phys 2022; 49:3107-3120. [PMID: 35170063 PMCID: PMC9313817 DOI: 10.1002/mp.15533] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/19/2021] [Revised: 02/03/2022] [Accepted: 02/03/2022] [Indexed: 11/20/2022] Open
Abstract
Background Skeletal muscle segmentation is an important procedure for assessing sarcopenia, an emerging imaging biomarker of patient frailty. Data annotation remains the bottleneck for training deep learning auto‐segmentation models. Purpose There is a need to define methodologies for applying models to different domains (e.g., anatomical regions or imaging modalities) without dramatically increasing data annotation. Methods To address this problem, we empirically evaluate the generalizability of various source tasks for transfer learning: natural image classification, natural image segmentation, unsupervised image reconstruction, and self‐supervised jigsaw solving. Axial CT slices at L3 were extracted from PET‐CT scans for 204 oesophago‐gastric cancer patients and the skeletal muscle manually delineated by an expert. Features were transferred and segmentation models trained on subsets (n=5,10,25,50,75,100,125) of the manually annotated training set. Four‐fold cross‐validation was performed to evaluate model generalizability. Human‐level performance was established by performing an inter‐observer study consisting of ten trained radiographers. Results We find that accurate segmentation models can be trained on a fraction of the data required by current approaches. The Dice similarity coefficient and root mean square distance‐to‐agreement were calculated for each prediction and used to assess model performance. Models pre‐trained on a segmentation task and fine‐tuned on 10 images produce delineations that are comparable to those from trained observers and extract reliable measures of muscle health. Conclusions Appropriate transfer learning can generate convolutional neural networks for abdominal muscle segmentation that achieve human‐level performance while decreasing the required data by an order of magnitude, compared to previous methods (n=160→10). This work enables the development of future models for assessing skeletal muscle at other anatomical sites where large annotated data sets are scarce and clinical needs are yet to be addressed.
Collapse
Affiliation(s)
- Dónal M McSweeney
- Division of Cancer Sciences, University of Manchester, Manchester, M13 9PL, UK.,Radiotherapy Related Research, The Christie Foundation Trust, Manchester, M20 4BX, UK
| | - Edward G Henderson
- Division of Cancer Sciences, University of Manchester, Manchester, M13 9PL, UK.,Radiotherapy Related Research, The Christie Foundation Trust, Manchester, M20 4BX, UK
| | - Marcel van Herk
- Division of Cancer Sciences, University of Manchester, Manchester, M13 9PL, UK.,Radiotherapy Related Research, The Christie Foundation Trust, Manchester, M20 4BX, UK
| | - Jamie Weaver
- Department of Medical Oncology, The Christie Hospital NHS Foundation Trust, Manchester, M20 4BX, UK
| | - Paul A Bromiley
- Division of Informatics, Imaging and Data Sciences, University of Manchester, Manchester, M13 9PL, UK
| | - Andrew Green
- Division of Cancer Sciences, University of Manchester, Manchester, M13 9PL, UK.,Radiotherapy Related Research, The Christie Foundation Trust, Manchester, M20 4BX, UK
| | - Alan McWilliam
- Division of Cancer Sciences, University of Manchester, Manchester, M13 9PL, UK.,Radiotherapy Related Research, The Christie Foundation Trust, Manchester, M20 4BX, UK
| |
Collapse
|
5
|
Lim WH, Park CM. Validation for measurements of skeletal muscle areas using low-dose chest computed tomography. Sci Rep 2022; 12:463. [PMID: 35013501 PMCID: PMC8748601 DOI: 10.1038/s41598-021-04492-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/14/2021] [Accepted: 12/23/2021] [Indexed: 11/21/2022] Open
Abstract
Various methods were suggested to measure skeletal muscle areas (SMAs) using chest low-dose computed tomography (chest LDCT) as a substitute for SMA at 3rd lumbar vertebra level (L3-SMA). In this study, four SMAs (L1-SMA, T12-erector spinae muscle areas, chest wall muscle area at carina level, pectoralis muscle area at aortic arch level) were segmented semi-automatically in 780 individuals taking concurrent chest and abdomen LDCT for healthcare screening. Four SMAs were compared to L3-SMA and annual changes were calculated from individuals with multiple examinations (n = 101). Skeletal muscle index (SMI; SMA/height2) cut-off for sarcopenia was determined by lower 5th percentile of young individuals (age ≤ 40 years). L1-SMA showed the greatest correlation to L3-SMA (men, R2 = 0.7920; women, R2 = 0.7396), and the smallest annual changes (0.3300 ± 4.7365%) among four SMAs. L1-SMI cut-offs for determining sarcopenia were 39.2cm2/m2 in men, and 27.5cm2/m2 in women. Forty-six men (9.5%) and ten women (3.4%) were found to have sarcopenia using L1-SMI cut-offs. In conclusion, L1-SMA could be a reasonable substitute for L3-SMA in chest LDCT. Suggested L1-SMI cut-offs for sarcopenia were 39.2cm2/m2 for men and 27.5cm2/m2 for women in Asian.
Collapse
Affiliation(s)
- Woo Hyeon Lim
- Department of Radiology, Namwon Medical Center, Namwon-si, Jeollabuk-do, Korea.,Department of Radiology, Seoul National University College of Medicine, 101 Daehak-ro, Chongno-gu, Seoul, 03080, Korea
| | - Chang Min Park
- Department of Radiology, Seoul National University College of Medicine, 101 Daehak-ro, Chongno-gu, Seoul, 03080, Korea. .,Institute of Radiation Medicine, Seoul National University Medical Research Center, Seoul, Korea. .,Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea. .,Institute of Medical and Biological Engineering, Medical Research Center, Seoul National University, Seoul, Korea.
| |
Collapse
|
6
|
Deep learning-based quantification of temporalis muscle has prognostic value in patients with glioblastoma. Br J Cancer 2021; 126:196-203. [PMID: 34848854 PMCID: PMC8770629 DOI: 10.1038/s41416-021-01590-9] [Citation(s) in RCA: 12] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/07/2021] [Revised: 08/25/2021] [Accepted: 10/06/2021] [Indexed: 01/19/2023] Open
Abstract
Background Glioblastoma is the commonest malignant brain tumour. Sarcopenia is associated with worse cancer survival, but manually quantifying muscle on imaging is time-consuming. We present a deep learning-based system for quantification of temporalis muscle, a surrogate for skeletal muscle mass, and assess its prognostic value in glioblastoma. Methods A neural network for temporalis segmentation was trained with 366 MRI head images from 132 patients from 4 different glioblastoma data sets and used to quantify muscle cross-sectional area (CSA). Association between temporalis CSA and survival was determined in 96 glioblastoma patients from internal and external data sets. Results The model achieved high segmentation accuracy (Dice coefficient 0.893). Median age was 55 and 58 years and 75.6 and 64.7% were males in the in-house and TCGA-GBM data sets, respectively. CSA was an independently significant predictor for survival in both the in-house and TCGA-GBM data sets (HR 0.464, 95% CI 0.218–0.988, p = 0.046; HR 0.466, 95% CI 0.235–0.925, p = 0.029, respectively). Conclusions Temporalis CSA is a prognostic marker in patients with glioblastoma, rapidly and accurately assessable with deep learning. We are the first to show that a head/neck muscle-derived sarcopenia metric generated using deep learning is associated with oncological outcomes and one of the first to show deep learning-based muscle quantification has prognostic value in cancer.
Collapse
|
7
|
Jullien M, Tessoulin B, Ghesquières H, Oberic L, Morschhauser F, Tilly H, Ribrag V, Lamy T, Thieblemont C, Villemagne B, Gressin R, Bouabdallah K, Haioun C, Damaj G, Fornecker LM, Schiano De Colella JM, Feugier P, Hermine O, Cartron G, Bonnet C, André M, Bailly C, Casasnovas RO, Le Gouill S. Deep-Learning Assessed Muscular Hypodensity Independently Predicts Mortality in DLBCL Patients Younger Than 60 Years. Cancers (Basel) 2021; 13:cancers13184503. [PMID: 34572728 PMCID: PMC8466314 DOI: 10.3390/cancers13184503] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/18/2021] [Revised: 08/27/2021] [Accepted: 09/03/2021] [Indexed: 01/06/2023] Open
Abstract
Simple Summary Cachexia is a major cause of mortality in cancer patients and is characterized by a continuous skeletal muscle loss. Muscle depletion assessed by computed tomography (CT) is a predictive marker in solid tumors but has never been assessed in non-Hodgkin’s lymphoma. Despite software improvements, its measurement remains highly time-consuming and cannot be performed in clinical practice. We report the development of a CT segmentation algorithm based on convolutional neural networks. It automates the extraction of anthropometric data from pretherapeutic CT to assess precise body composition of young diffuse large B cell lymphoma (DLBCL) patients at the time of diagnosis. In this population, muscle hypodensity appears to be an independent risk factor for mortality, and can be estimated at diagnosis with this new tool. Abstract Background. Muscle depletion (MD) assessed by computed tomography (CT) has been shown to be a predictive marker in solid tumors, but has not been assessed in non-Hodgkin’s lymphomas. Despite software improvements, MD measurement remains highly time-consuming and cannot be used in clinical practice. Methods. This study reports the development of a Deep-Learning automatic segmentation algorithm (DLASA) to measure MD, and investigate its predictive value in a cohort of 656 diffuse large B cell lymphoma (DLBCL) patients included in the GAINED phase III prospective trial (NCT01659099). Results. After training on a series of 190 patients, the DLASA achieved a Dice coefficient of 0.97 ± 0.03. In the cohort, the median skeletal muscle index was 50.2 cm2/m2 and median muscle attenuation (MA) was 36.1 Hounsfield units (HU). No impact of sarcopenia was found on either progression free survival (PFS) or overall survival (OS). Muscular hypodensity, defined as MA below the tenth percentile according to sex, was associated with a lower OS and PFS, respectively (HR = 2.80 (95% CI 1.58–4.95), p < 0.001, and HR = 2.22 (95% CI 1.43–3.45), p < 0.001). Muscular hypodensity appears to be an independent risk factor for mortality in DLBCL and because of DLASA can be estimated in routine practice.
Collapse
Affiliation(s)
- Maxime Jullien
- Department of Hematology, Nantes University Hospital, INSERM CRCINA Nantes-Angers, NeXT Université de Nantes, 44000 Nantes, France; (M.J.); (B.T.)
| | - Benoit Tessoulin
- Department of Hematology, Nantes University Hospital, INSERM CRCINA Nantes-Angers, NeXT Université de Nantes, 44000 Nantes, France; (M.J.); (B.T.)
| | - Hervé Ghesquières
- Department of Hematology, Hospices Civils de Lyon, Centre Hospitalier Lyon-Sud, Claude Bernard Lyon-1 University, 69310 Pierre Bénite, France;
| | - Lucie Oberic
- Department of Hematology, IUC Toulouse Oncopole, 31000 Toulouse, France;
| | - Franck Morschhauser
- Department of Hematology, Univ. Lille, CHU Lille, EA 7365-GRITA-Groupe de Recherche sur les Formes Injectables et les Technologies Associées, 59000 Lille, France;
| | - Hervé Tilly
- Department of Hematology, Centre H. Becquerel, 76000 Rouen, France;
| | - Vincent Ribrag
- Department of Hematology, Gustave Roussy, Université Paris-Saclay, 94800 Villejuif, France;
| | - Thierry Lamy
- Department of Hematology, University Hospital of Rennes, 35000 Rennes, France;
| | - Catherine Thieblemont
- Department of Hematology, APHP, Hopital Saint Louis, Université Paris Diderot, 75011 Paris, France;
| | - Bruno Villemagne
- Department of Hematology, Hopital Departemental de Vendée, 85000 La Roche sur Yon, France;
| | - Rémy Gressin
- Department of Hematology, CHU Grenoble, 38000 Grenoble, France;
| | - Kamal Bouabdallah
- Department of Hematology, University Hospital of Bordeaux, F-33000 Bordeaux, France;
| | - Corinne Haioun
- Lymphoïd Malignancies Unit, Hôpital Henri Mondor, AP-HP, 94000 Créteil, France;
| | - Gandhi Damaj
- Department of Hematology, Institut D’hématologie de Basse Normandie, 14000 Caen, France;
| | - Luc-Matthieu Fornecker
- Department of Hematology, Institut de Cancérologie Strasbourg Europe (ICANS), University Hospital of Strasbourg, 67000 Strasbourg, France;
| | | | - Pierre Feugier
- Department of Hematology, University Hospital of Nancy, 54000 Nancy, France;
| | - Olivier Hermine
- Department of Hematology, Hopital Necker, F-75015 Paris, France;
| | - Guillaume Cartron
- Department of Clinical Hematology, University Hospital of Montpellier, UMR-CNRS 5535, 34000 Montpellier, France;
| | - Christophe Bonnet
- Department of Hematology, CHU Liege, Liege University, 4000 Liege, Belgium;
| | - Marc André
- Department of Hematology, CHU UCL Namur, Université Catholique de Louvain, 5000 Namur, Belgium;
| | - Clément Bailly
- Department of Nuclear Medicine, University Hospital of Nantes, 44000 Nantes, France;
| | - René-Olivier Casasnovas
- Department of Hematology, University Hospital F. Mitterrand and Inserm UMR 1231, 21000 Dijon, France;
| | - Steven Le Gouill
- Department of Hematology, Nantes University Hospital, INSERM CRCINA Nantes-Angers, NeXT Université de Nantes, 44000 Nantes, France; (M.J.); (B.T.)
- Correspondence: ; Tel.: +33-(0)1-44-32-41-00
| |
Collapse
|