%PDF-1.3
%
1 0 obj
<<
/Keywords(Muscles, Neural networks \(computer\), Body composition, Subcutaneous fat, Tomography \(x-ray, computed\), )
/Creator(Arbortext Advanced Print Publisher 9.1.440/W Unicode)
/ModDate(D:20240328151123+00'00')
/CreationDate(D:20210310094230+08'00')
/Producer(PDFsharp 1.50.5147 \(www.pdfsharp.com\) \(Original: Acrobat Distiller 10.0.0 \(Windows\); modified using iText 5.3.5 2000-2012 1T3XT BVBA \(SPRINGER SBM; licensed version\)\))
/Subject(Background: 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. Methods: 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. Results: 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%. Conclusions: 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.)
/Author(Borrelli, P., Kaboteh, R., Enqvist, O. et al)
/Title(Artificial intelligence-aided CT segmentation for body composition analysis: a validation study)
>>
endobj
2 0 obj
<<
/Subtype/XML
/Type/Metadata
/Length 20010
>>
stream
Body composition,Muscles,Neural networks (computer),Subcutaneous fat,Tomography (x-ray,computed)
Acrobat Distiller 10.0.0 (Windows); modified using iText® 5.3.5 ©2000-2012 1T3XT BVBA (SPRINGER SBM; licensed version)
application/pdf
10.1186/s41747-021-00210-8
European Radiology Experimental
Eur Radiol Exp, doi:10.1186/s41747-021-00210-8
Body composition
Muscles
Neural networks (computer)
Subcutaneous fat
Tomography (x-ray
computed)
Artificial intelligence-aided CT segmentation for body composition analysis: a validation study
Pablo Borrelli
Reza Kaboteh
Olof Enqvist
Johannes Ulén
Elin Trägårdh
Henrik Kjölhede
Lars Edenbrandt
2021-03-11T01:37:31+01:00
2021-03-11T01:37:31+01:00
2021-03-10T09:42:30+08:00
Arbortext Advanced Print Publisher 9.1.440/W Unicode
uuid:1ca826f7-5404-4b13-b6d8-99425b993f4a
uuid:df26773a-0999-4215-9530-a2ca08dfce36
default
1
converted
uuid:1ca826f7-5404-4b13-b6d8-99425b993f4a
converted to PDF/A-2b
pdfToolbox
2021-03-11T03:17:30+08:00
2
B
Pablo Borrelli
http://orcid.org/0000-0002-1665-7088
Springer Nature ORCID Schema
http://springernature.com/ns/xmpExtensions/2.0/
sn
authorInfo
Bag AuthorInformation
external
Author information: contains the name of each author and his/her ORCID (ORCiD: Open Researcher and Contributor ID). An ORCID is a persistent identifier (a non-proprietary alphanumeric code) to uniquely identify scientific and other academic authors.
editorInfo
Bag EditorInformation
external
Editor information: contains the name of each editor and his/her ORCID identifier.
seriesEditorInfo
Bag SeriesEditorInformation
external
Series editor information: contains the name of each series editor and his/her ORCID identifier.
AuthorInformation
http://springernature.com/ns/xmpExtensions/2.0/authorInfo/
author
Specifies the types of author information: name and ORCID of an author.
name
Text
Gives the name of an author.
orcid
URI
Gives the ORCID of an author.
EditorInformation
http://springernature.com/ns/xmpExtensions/2.0/editorInfo/
editor
Specifies the types of editor information: name and ORCID of an editor.
name
Text
Gives the name of an editor.
orcid
URI
Gives the ORCID of an editor.
SeriesEditorInformation
http://springernature.com/ns/xmpExtensions/2.0/seriesEditorInfo/
seriesEditor
Specifies the types of series editor information: name and ORCID of a series editor.
name
Text
Gives the name of a series editor.
orcid
URI
Gives the ORCID of a series editor.
http://ns.adobe.com/pdf/1.3/
pdf
Adobe PDF Schema
internal
A name object indicating whether the document has been modified to include trapping information
Trapped
Text
http://ns.adobe.com/pdfx/1.3/
pdfx
Adobe Document Info PDF eXtension Schema
internal
ID of PDF/X standard
GTS_PDFXVersion
Text
internal
Conformance level of PDF/X standard
GTS_PDFXConformance
Text
internal
Company creating the PDF
Company
Text
internal
Date when document was last modified
SourceModified
Text
http://ns.adobe.com/xap/1.0/mm/
xmpMM
XMP Media Management Schema
internal
UUID based identifier for specific incarnation of a document
InstanceID
URI
internal
The common identifier for all versions and renditions of a document.
OriginalDocumentID
URI
http://www.aiim.org/pdfa/ns/id/
pdfaid
PDF/A ID Schema
internal
Part of PDF/A standard
part
Integer
internal
Amendment of PDF/A standard
amd
Text
internal
Conformance level of PDF/A standard
conformance
Text
endstream
endobj
3 0 obj
<<
/Type/Encoding
/Differences[1/C15 188/onequarter 254/thorn]
>>
endobj
4 0 obj
<<
/D[5 0 R/FitR 0 484 596 484]
>>
endobj
5 0 obj
<<
/Contents 276 0 R
/Thumb 311 0 R
/Type/Page
/Resources
<<
/ColorSpace
<<
/Cs6 10 0 R
>>
/ExtGState
<<
/GS5 190 0 R
/GS3 11 0 R
>>
/ProcSet[/PDF/Text/ImageB/ImageC/ImageI]
/Shading
<<
/Sh1 202 0 R
>>
/Font
<<
/F10 54 0 R
/F7 312 0 R
/F11 13 0 R
/F8 14 0 R
/F9 15 0 R
/F1 16 0 R
/F2 17 0 R
/F3 80 0 R
/F4 18 0 R
/F5 81 0 R
/F13 12 0 R
/F6 115 0 R
>>
/Properties
<<
/MC7 182 0 R
/MC8 196 0 R
/MC5 167 0 R
/MC6 174 0 R
/MC3 155 0 R
/MC4 161 0 R
/MC1 143 0 R
/MC2 149 0 R
>>
>>
/Annots[313 0 R 314 0 R 315 0 R 316 0 R 317 0 R 318 0 R 319 0 R 320 0 R 321 0 R 322 0 R]
/CropBox[0 0 595.276 790.866]
/Parent 31 0 R
/Rotate 0
/MediaBox[0 0 595.276 790.866]
/Group
<<
/CS/DeviceRGB
/S/Transparency
>>
>>
endobj
6 0 obj
<<
/Subtype/Form
/Filter/FlateDecode
/Type/XObject
/Resources
<<
>>
/BBox[0 0 9.638 9.80804]
/Length 15
>>
stream
HR(T0 u
endstream
endobj
7 0 obj
<<
/Contents 8 0 R
/Thumb 9 0 R
/Type/Page
/Resources
<<
/ColorSpace
<<
/Cs6 10 0 R
>>
/ExtGState
<<
/GS2 11 0 R
>>
/ProcSet[/PDF/Text/ImageB/ImageC/ImageI]
/Font
<<
/F10 12 0 R
/F11 13 0 R
/F8 14 0 R
/F9 15 0 R
/F1 16 0 R
/F2 17 0 R
/F4 18 0 R
/F13 19 0 R
>>
>>
/Annots[20 0 R 21 0 R 22 0 R 23 0 R 24 0 R 25 0 R 26 0 R 27 0 R 28 0 R 29 0 R 30 0 R]
/CropBox[0 0 595.276 790.866]
/Parent 31 0 R
/Rotate 0
/MediaBox[0 0 595.276 790.866]
/Group
<<
/CS/DeviceRGB
/S/Transparency
>>
>>
endobj
8 0 obj
<<
/Length 7513
/Filter/FlateDecode
>>
stream
x\vH+αhէYkcPd,f=L%Z@|DFܸqy{q;yk=iWCݷcvߤ4\`\wo~^\oP\OoU^E
ɟ_eE]deXͿoUʒTo.5_^+lۡ;S[e1%V>r[vtyUlҬٕI$C=]խ.ƬM럿吭"+*,wZo:l<$WYER
5\namRuߍͦpyd@Ո߇NM=Nv]3U7b?6ܬ-`[Owg'~|tN.o
λװZO/s٪ڧ}.|֖+;Z/Z.9[m}32_^eI:
`itz{\0R-U`W`l2uhb8#Nek31qZhu1y
,2}}IjwuVм}>}
M}}ݶdNz7U8vh W剩Y+҈[|xO
Bzv7ѐ6|ndp4w}{Wvp#*ߏx{_
ӀMKnщ0ͺjyЂ2i>gx:ف.`+5-Wȗ43u_<^/ն*S<+{79
ju۫
Lܤ鑌9w]w~9|PCմ⧥LqƉH~14̒Q;h=]WLX؊bN Fs¶ImN;&1wG+āqtuqmpڲiu|5ԇ/q+#s
rtMc3h+#}L!`0j
C@k7rcj ps_N /S C`# /0Z06!@O;|