Impact of inter-reader contouring variability on textural radiomics of colorectal liver metastases.
Colorectal neoplasms
Image processing (computer-assisted)
Liver neoplasms
Radiomics
Tomography (x-ray
computed)
Journal
European radiology experimental
ISSN: 2509-9280
Titre abrégé: Eur Radiol Exp
Pays: England
ID NLM: 101721752
Informations de publication
Date de publication:
10 11 2020
10 11 2020
Historique:
received:
22
06
2020
accepted:
13
10
2020
entrez:
10
11
2020
pubmed:
11
11
2020
medline:
6
7
2021
Statut:
epublish
Résumé
Radiomics is expected to improve the management of metastatic colorectal cancer (CRC). We aimed at evaluating the impact of liver lesion contouring as a source of variability on radiomic features (RFs). After Ethics Committee approval, 70 liver metastases in 17 CRC patients were segmented on contrast-enhanced computed tomography scans by two residents and checked by experienced radiologists. RFs from grey level co-occurrence and run length matrices were extracted from three-dimensional (3D) regions of interest (ROIs) and the largest two-dimensional (2D) ROIs. Inter-reader variability was evaluated with Dice coefficient and Hausdorff distance, whilst its impact on RFs was assessed using mean relative change (MRC) and intraclass correlation coefficient (ICC). For the main lesion of each patient, one reader also segmented a circular ROI on the same image used for the 2D ROI. The best inter-reader contouring agreement was observed for 2D ROIs according to both Dice coefficient (median 0.85, interquartile range 0.78-0.89) and Hausdorff distance (0.21 mm, 0.14-0.31 mm). Comparing RF values, MRC ranged 0-752% for 2D and 0-1567% for 3D. For 24/32 RFs (75%), MRC was lower for 2D than for 3D. An ICC > 0.90 was observed for more RFs for 2D (53%) than for 3D (34%). Only 2/32 RFs (6%) showed a variability between 2D and circular ROIs higher than inter-reader variability. A 2D contouring approach may help mitigate overall inter-reader variability, albeit stable RFs can be extracted from both 3D and 2D segmentations of CRC liver metastases.
Sections du résumé
BACKGROUND
Radiomics is expected to improve the management of metastatic colorectal cancer (CRC). We aimed at evaluating the impact of liver lesion contouring as a source of variability on radiomic features (RFs).
METHODS
After Ethics Committee approval, 70 liver metastases in 17 CRC patients were segmented on contrast-enhanced computed tomography scans by two residents and checked by experienced radiologists. RFs from grey level co-occurrence and run length matrices were extracted from three-dimensional (3D) regions of interest (ROIs) and the largest two-dimensional (2D) ROIs. Inter-reader variability was evaluated with Dice coefficient and Hausdorff distance, whilst its impact on RFs was assessed using mean relative change (MRC) and intraclass correlation coefficient (ICC). For the main lesion of each patient, one reader also segmented a circular ROI on the same image used for the 2D ROI.
RESULTS
The best inter-reader contouring agreement was observed for 2D ROIs according to both Dice coefficient (median 0.85, interquartile range 0.78-0.89) and Hausdorff distance (0.21 mm, 0.14-0.31 mm). Comparing RF values, MRC ranged 0-752% for 2D and 0-1567% for 3D. For 24/32 RFs (75%), MRC was lower for 2D than for 3D. An ICC > 0.90 was observed for more RFs for 2D (53%) than for 3D (34%). Only 2/32 RFs (6%) showed a variability between 2D and circular ROIs higher than inter-reader variability.
CONCLUSIONS
A 2D contouring approach may help mitigate overall inter-reader variability, albeit stable RFs can be extracted from both 3D and 2D segmentations of CRC liver metastases.
Identifiants
pubmed: 33169295
doi: 10.1186/s41747-020-00189-8
pii: 10.1186/s41747-020-00189-8
pmc: PMC7652946
doi:
Substances chimiques
Contrast Media
0
Types de publication
Clinical Trial, Phase II
Journal Article
Multicenter Study
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
62Subventions
Organisme : Associazione Italiana per la Ricerca sul Cancro
ID : 21091
Pays : International
Références
United European Gastroenterol J. 2016 Apr;4(2):257-63
pubmed: 27087955
PLoS One. 2018 Oct 4;13(10):e0205003
pubmed: 30286184
Eur J Radiol. 2013 Jun;82(6):959-68
pubmed: 23489982
Int J Radiat Oncol Biol Phys. 2018 Nov 15;102(4):1074-1082
pubmed: 30170101
Abdom Imaging. 2015 Oct;40(7):2331-7
pubmed: 25968046
Abdom Radiol (NY). 2018 Dec;43(12):3271-3278
pubmed: 29730738
Eur Radiol. 2020 Jan;30(1):195-205
pubmed: 31392481
Sci Rep. 2019 Jan 24;9(1):614
pubmed: 30679599
Transl Oncol. 2017 Dec;10(6):886-894
pubmed: 28930698
Phys Med Biol. 2020 Sep 29;65(19):195012
pubmed: 32575082
Med Phys. 2015 Mar;42(3):1341-53
pubmed: 25735289
Radiology. 2016 Feb;278(2):563-77
pubmed: 26579733
J Radiosurg SBRT. 2014;3(2):149-163
pubmed: 29296396
Clin Transl Oncol. 2020 May;22(5):647-662
pubmed: 31359336
Ann Surg Oncol. 2017 Sep;24(9):2482-2490
pubmed: 28560599
J Chiropr Med. 2016 Jun;15(2):155-63
pubmed: 27330520
Eur J Radiol. 2017 Dec;97:76-82
pubmed: 29153371
Acta Oncol. 2018 Aug;57(8):1070-1074
pubmed: 29513054
EXCLI J. 2016 Jun 27;15:406-23
pubmed: 27540353
Theranostics. 2019 Feb 12;9(5):1303-1322
pubmed: 30867832
Radiology. 2013 Nov;269(2):451-9
pubmed: 23824993
Eur J Radiol. 2013 Feb;82(2):342-8
pubmed: 23194641
Med Phys. 2012 Oct;39(10):6332-8
pubmed: 23039669
Clin Transl Radiat Oncol. 2019 Nov 28;21:11-18
pubmed: 31886423
Insights Imaging. 2019 Mar 4;10(1):28
pubmed: 30830470
Magn Reson Imaging. 2012 Nov;30(9):1234-48
pubmed: 22898692
AJR Am J Roentgenol. 2019 Mar;212(3):497-504
pubmed: 30620678
Eur J Radiol. 2018 May;102:15-21
pubmed: 29685529
Sci Rep. 2016 Mar 24;6:23428
pubmed: 27009765
Gut. 2020 Mar;69(3):531-539
pubmed: 31101691
Int J Radiat Oncol Biol Phys. 2018 Nov 15;102(4):1143-1158
pubmed: 30170872
Phys Med. 2017 Jun;38:122-139
pubmed: 28595812
AJR Am J Roentgenol. 2019 Aug;213(2):377-383
pubmed: 31063427
Quant Imaging Med Surg. 2019 Mar;9(3):453-464
pubmed: 31032192
Ann Oncol. 2016 Aug;27(8):1386-422
pubmed: 27380959
Nat Rev Clin Oncol. 2017 Dec;14(12):749-762
pubmed: 28975929
Transl Cancer Res. 2016 Aug;5(4):410-423
pubmed: 30687593