Artificial intelligence-aided CT segmentation for body composition analysis: a validation study
Tóm tắt
Body composition is associated with survival outcome in oncological patients, but it is not routinely calculated. Manual segmentation of subcutaneous adipose tissue (SAT) and muscle is time-consuming and therefore limited to a single CT slice. Our goal was to develop an artificial-intelligence (AI)-based method for automated quantification of three-dimensional SAT and muscle volumes from CT images. Ethical approvals from Gothenburg and Lund Universities were obtained. Convolutional neural networks were trained to segment SAT and muscle using manual segmentations on CT images from a training group of 50 patients. The method was applied to a separate test group of 74 cancer patients, who had two CT studies each with a median interval between the studies of 3 days. Manual segmentations in a single CT slice were used for comparison. The accuracy was measured as overlap between the automated and manual segmentations. The accuracy of the AI method was 0.96 for SAT and 0.94 for muscle. The average differences in volumes were significantly lower than the corresponding differences in areas in a single CT slice: 1.8% versus 5.0% (p < 0.001) for SAT and 1.9% versus 3.9% (p < 0.001) for muscle. The 95% confidence intervals for predicted volumes in an individual subject from the corresponding single CT slice areas were in the order of ± 20%. The AI-based tool for quantification of SAT and muscle volumes showed high accuracy and reproducibility and provided a body composition analysis that is more relevant than manual analysis of a single CT slice.
Tài liệu tham khảo
Brown JC, Cespedes Feliciano EM, Caan BJ (2018) The evolution of body composition in oncology-epidemiology, clinical trials, and the future of patient care: facts and numbers. J Cachexia Sarcopenia Muscle 9:1200–1208 doi: 10.1002%2Fjcsm.12379
Kamarajah SK, Bundred J, Tan BH (2019) Body composition assessment and sarcopenia in patients with gastric cancer: a systematic review and metaanalysis. Gastric Cancer 22:10–22. https://doi.org/10.1007/s10120-018-0882-2
Baracos VE, Reiman T, Mourtzakis M, Gioulbasanis I, Antoun S (2010) Body composition in patients with non− small cell lung cancer: a contemporary view of cancer cachexia with the use of computed tomography image analysis. Am J Clin Nutr 91:1133S–1137S. https://doi.org/10.3945/ajcn.2010.28608C
Hopkins JJ, Skubleny D, Bigam DL, Baracos VE, Eurich DT, Sawyer MB (2018) Barriers to the interpretation of body composition in colorectal cancer: a review of the methodological inconsistency and complexity of the CT-defined body habitus. Ann Surg Oncol 25:1381–1394. https://doi.org/10.1245/s10434-018-6395-8
Trestini I, Carbognin L, Monteverdi S et al (2018) Clinical implication of changes in body composition and weight in patients with early-stage and metastatic breast cancer. Crit Rev Oncol Hematol 129:54–66. https://doi.org/10.1016/j.critrevonc.2018.06.011
Peng YC, Wu CH, Tien YW, Lu TP, Wang YH, Chen BB (2020) Preoperative sarcopenia is associated with poor overall survival in pancreatic cancer patients following pancreaticoduodenectomy. Eur Radiol https://doi.org/10.1007/s00330-020-07294-7
Lin TY, Peng CH, Hung SC, Tarng DC (2018) Body composition is associated with clinical outcomes in patients with non–dialysis-dependent chronic kidney disease. Kidney Int 93:733–740. https://doi.org/10.1016/j.kint.2017.08.025
Cruz-Jentoft AJ, Bahat G, Bauer J et al (2019) Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing 48:16–31. https://doi.org/10.1093/ageing/afy169
Albano D, Messina C, Vitale J, Sconfienza LM (2020) Imaging of sarcopenia: old evidence and new insights. Eur Radiol 30:2199–2208. https://doi.org/10.1007/s00330-019-06573-2
Marhold M, Topakian T, Unseld M (2020) Sarcopenia in cancer—a focus on elderly cancer patients. memo. https://doi.org/10.1007/s12254-020-00637-6
Shen W, Punyanitya M, Wang Z et al (2004) Visceral adipose tissue: relations between single-slice areas and total volume. Am J Clin Nutr 80:271–278. https://doi.org/10.1093/ajcn/80.2.271
Weston AD, Korfiatis P, Kline TL et al (2019) Automated abdominal segmentation of CT scans for body composition analysis using deep learning. Radiology 290:669–679. https://doi.org/10.1148/radiol.2018181432
Lee H, Troschel FM, Tajmir S et al (2017) Pixel-level deep segmentation: artificial intelligence quantifies muscle on computed tomography for body morphometric analysis. J Digit Imaging 30:487–498. https://doi.org/10.1007/s10278-017-9988-z
Paris MT, Tandon P, Heyland DK et al (2020) Automated body composition analysis of clinically acquired computed tomography scans using neural networks. Clin Nutr 39:3049–3055. https://doi.org/10.1016/j.clnu.2020.01.008
Dabiri S, Popuri K, Feliciano EMC, Caan BJ, Baracos VE, Beg MF (2019) Muscle segmentation in axial computed tomography (CT) images at the lumbar (L3) and thoracic (T4) levels for body composition analysis. Comput Med Imaging Graph 75:47–55. https://doi.org/10.1016/j.compmedimag.2019.04.007
Bridge CP, Rosenthal M, Wright B et al (2018) Fully-automated analysis of body composition from CT in cancer patients using convolutional neural networks. Radiology 290:204–213. https://doi.org/10.1007/978-3-030-01201-4_22
Kjölhede H, Ahlgren G, Almquist H et al (2012) Combined 18F-fluorocholine and 18F-fluoride positron emission tomography/computed tomography imaging for staging of high-risk prostate cancer. BJU international 110:1501–1506. https://doi.org/10.1111/j.1464-410X.2012.11123.x
Trägårdh E, Borrelli P, Kaboteh R et al (2020) RECOMIA—a cloud-based platform for artificial intelligence research in nuclear medicine and radiology. EJNMMI physics 7:1–12. https://doi.org/10.1186/s40658-020-00316-9
Takahashi N, Sugimoto M, Psutka SP, Chen B, Moynagh MR, Carter RE (2017) Validation study of a new semi-automated software program for CT body composition analysis. Abdom Radiol (NY) 42:2369–2375. https://doi.org/10.1007/s00261-017-1123-6
Tegels JJ, Van Vugt JL, Reisinger KW et al (2015) Sarcopenia is highly prevalent in patients undergoing surgery for gastric cancer but not associated with worse outcomes. J Surg Oncol 112:403–407. https://doi.org/10.1002/jso.24015
Feliciano EMC, Kroenke CH, Meyerhardt JA et al (2017) Association of systemic inflammation and sarcopenia with survival in nonmetastatic colorectal cancer: results from the C SCANS study. JAMA Oncol 3:e172319–e172319. https://doi.org/10.1001/jamaoncol.2017.2319
Lee S, Janssen I, Ross R (2004) Interindividual variation in abdominal subcutaneous and visceral adipose tissue: influence of measurement site. J Appl Physiol (1985)97:948–954. https://doi.org/10.1152/japplphysiol.01200.2003
Greenfield JR, Samaras K, Chisholm DJ, Campbell LV (2002) Regional intra-subject variability in abdominal adiposity limits usefulness of computed tomography. Obes Res. 10:260–265. https://doi.org/10.1038/oby.2002.35
Decazes P, Metivier D, Rouquette A, Talbot J-N, Kerrou K (2016) A method to improve the semiquantification of 18F-FDG uptake: reliability of the estimated lean body mass using the conventional, low-dose CT from PET/CT. J Nucl Med 57:753–758. https://doi.org/10.2967/jnumed.115.164913
Decazes P, Tonnelet D, Vera P, Gardin I (2019) Anthropometer3D: automatic multi-slice segmentation software for the measurement of anthropometric parameters from CT of PET/CT. J Digit Imaging 32:241–250. https://doi.org/10.1007/s10278-019-00178-3
Kullberg J, Hedström A, Brandberg J et al (2017) Automated analysis of liver fat, muscle and adipose tissue distribution from CT suitable for large-scale studies. Sci Rep 7:10425. https://doi.org/10.1038/s41598-017-08925-8
Kullberg J, Ahlström H, Johansson L, Frimmel H (2007) Automated and reproducible segmentation of visceral and subcutaneous adipose tissue from abdominal MRI. Int. Int J Obes 31:1806–1817. https://doi.org/10.1038/sj.ijo.0803671
Makrogiannis S, Caturegli G, Davatzikos C, Ferrucci L (2013) Computer-aided assessment of regional abdominal fat with food residue removal in CT. Acad Radiol 20:1413–1421. https://doi.org/10.1016/j.acra.2013.08.007