To investigate multi-site CNN segmentation of 4D flow MRI for pediatric blood flow measurement.
Retrospective.
A total of 174 subjects across two sites (female: 46%; N = 38 healthy controls, N = 136 CHD patients). Participants from site 1 (N = 100), site 2 (N = 74), and both sites (N = 174) were divided into subgroups to conduct 10-fold cross validation (10% for testing, 90% for training).
3\u2009T\/1.5 T; retrospectively gated gradient recalled echo-based 4D flow MRI.
Accuracy of the 3D CNN segmentations trained on data from single site (single-site CNNs) and data across both sites (multi-site CNN) were evaluated by geometrical similarity (Dice score, human segmentation as ground truth) and net flow quantification at the ascending aorta (Qs), main pulmonary artery (Qp), and their balance (Qp\/Qs), between human observers, single-site and multi-site CNNs.
Kruskal-Wallis test, Wilcoxon rank-sum test, and Bland-Altman analysis. A P-value <0.05 was considered statistically significant.
No difference existed between single-site and multi-site CNNs for geometrical similarity in the aorta by Dice score (site 1: 0.916 vs. 0.915, P = 0.55; site 2: 0.906 vs. 0.904, P = 0.69) and for the pulmonary arteries (site 1: 0.894 vs. 0.895, P = 0.64; site 2: 0.870 vs. 0.869, P = 0.96). Qs site-1 medians were 51.0-51.3 mL\/cycle (P = 0.81) and site-2 medians were 66.7-69.4 mL\/cycle (P = 0.84). Qp site-1 medians were 46.8-48.0\u2009mL\/cycle (P = 0.97) and site-2 medians were 76.0-77.4 mL\/cycle (P = 0.98). Qp\/Qs site-1 medians were 0.87-0.88 (P = 0.97) and site-2 medians were 1.01-1.03 (P = 0.43). Bland-Altman analysis for flow quantification found equivalent performance.
Multi-site CNN-based segmentation and blood flow measurement are feasible for pediatric 4D flow MRI and maintain performance of single-site CNNs.
3 TECHNICAL EFFICACY: Stage 2.<\/p>\n\n\n\n
<\/p>\n\n\n\n
<\/p>\n\n\n\n
<\/p>\n\n\n\n
\u6458\u8981\uff1a\n\u7814\u7a76\u7528\u4e8e\u513f\u79d1\u8840\u6d41\u6d4b\u91cf\u7684 4D \u6d41 MRI \u7684\u591a\u7ad9\u70b9 CNN \u5206\u5272\u3002\n\u56de\u987e\u6027\u7684\u3002\n\u4e24\u4e2a\u5730\u70b9\u5171\u6709 174 \u540d\u53d7\u8bd5\u8005\uff08\u5973\u6027\uff1a46%\uff1bN = 38 \u540d\u5065\u5eb7\u5bf9\u7167\u8005\uff0cN = 136 \u540d\u51a0\u5fc3\u75c5\u60a3\u8005\uff09\u3002\u6765\u81ea\u7ad9\u70b9 1\uff08N = 100\uff09\u3001\u7ad9\u70b9 2\uff08N = 74\uff09\u548c\u4e24\u4e2a\u7ad9\u70b9\uff08N = 174\uff09\u7684\u53c2\u4e0e\u8005\u88ab\u5206\u6210\u5c0f\u7ec4\u8fdb\u884c 10 \u6298\u4ea4\u53c9\u9a8c\u8bc1\uff0810% \u7528\u4e8e\u6d4b\u8bd5\uff0c90% \u7528\u4e8e\u57f9\u8bad\uff09\u3002\n3\u27e9T\/1.5T\uff1b\u56de\u987e\u6027\u95e8\u63a7\u68af\u5ea6\u53ec\u56de\u57fa\u4e8e\u56de\u58f0\u7684 4D \u6d41\u52a8 MRI\u3002\n\u901a\u8fc7\u51e0\u4f55\u76f8\u4f3c\u6027\uff08Dice \u5206\u6570\uff0c\u4eba\u7c7b\u5206\u5272\u4f5c\u4e3a\u5730\u9762\u5b9e\u51b5\uff09\u548c\u51c0\u6d41\u91cf\u91cf\u5316\u8bc4\u4f30\u4e86\u5728\u5355\u4e2a\u7ad9\u70b9\uff08\u5355\u7ad9\u70b9 CNN\uff09\u548c\u8de8\u4e24\u4e2a\u7ad9\u70b9\uff08\u591a\u7ad9\u70b9 CNN\uff09\u7684\u6570\u636e\u4e0a\u8bad\u7ec3\u7684 3D CNN \u5206\u5272\u7684\u51c6\u786e\u6027\u5347\u4e3b\u52a8\u8109 (Qs)\u3001\u4e3b\u80ba\u52a8\u8109 (Qp) \u53ca\u5176\u5728\u4eba\u7c7b\u89c2\u5bdf\u8005\u3001\u5355\u7ad9\u70b9\u548c\u591a\u7ad9\u70b9 CNN \u4e4b\u95f4\u7684\u5e73\u8861 (Qp\/Qs)\u3002\nKruskal-Wallis \u68c0\u9a8c\u3001Wilcoxon \u79e9\u548c\u68c0\u9a8c\u548c Bland-Altman \u5206\u6790\u3002 P\u503c<0.05\u88ab\u8ba4\u4e3a\u5177\u6709\u7edf\u8ba1\u5b66\u610f\u4e49\u3002\n\u901a\u8fc7 Dice \u8bc4\u5206\uff08\u7ad9\u70b9 1\uff1a0.916 \u4e0e 0.915\uff0cP = 0.55\uff1b\u7ad9\u70b9 2\uff1a0.906 \u4e0e 0.904\uff0cP = 0.69\uff09\u548c\u80ba\u90e8\u51e0\u4f55\u76f8\u4f3c\u6027\uff0c\u5355\u7ad9\u70b9\u548c\u591a\u7ad9\u70b9 CNN \u4e4b\u95f4\u6ca1\u6709\u5dee\u5f02\u52a8\u8109\uff08\u7ad9\u70b9 1\uff1a0.894 \u4e0e 0.895\uff0cP = 0.64\uff1b\u7ad9\u70b9 2\uff1a0.870 \u4e0e 0.869\uff0cP = 0.96\uff09\u3002 Qs site-1 \u4e2d\u4f4d\u6570\u4e3a 51.0-51.3 mL\/\u5468\u671f (P = 0.81)\uff0csite-2 \u4e2d\u4f4d\u6570\u4e3a 66.7-69.4 mL\/\u5468\u671f (P = 0.84)\u3002 Qp site-1 \u4e2d\u4f4d\u6570\u4e3a 46.8-48.0mL\/\u5468\u671f (P = 0.97)\uff0csite-2 \u4e2d\u4f4d\u6570\u4e3a 76.0-77.4mL\/\u5468\u671f (P = 0.98)\u3002 Qp\/Qs \u7ad9\u70b9 1 \u4e2d\u4f4d\u6570\u4e3a 0.87-0.88 (P = 0.97)\uff0c\u7ad9\u70b9 2 \u4e2d\u4f4d\u6570\u4e3a 1.01-1.03 (P = 0.43)\u3002\u7528\u4e8e\u6d41\u91cf\u91cf\u5316\u7684 Bland-Altman \u5206\u6790\u53d1\u73b0\u7b49\u6548\u6027\u80fd\u3002\n\u57fa\u4e8e CNN \u7684\u591a\u7ad9\u70b9\u5206\u5272\u548c\u8840\u6d41\u6d4b\u91cf\u5bf9\u4e8e\u513f\u79d1 4D \u6d41\u52a8 MRI \u662f\u53ef\u884c\u7684\uff0c\u5e76\u4fdd\u6301\u5355\u7ad9\u70b9 CNN \u7684\u6027\u80fd\u3002\n3 \u6280\u672f\u6548\u80fd\uff1a\u7b2c 2 \u9636\u6bb5\u3002<\/pre>\n","protected":false},"excerpt":{"rendered":"To investigate multi-site CNN segmentation of 4D flow M […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":""},"categories":[1,9],"tags":[6],"jetpack_featured_media_url":"","_links":{"self":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/108"}],"collection":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=108"}],"version-history":[{"count":1,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/108\/revisions"}],"predecessor-version":[{"id":109,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/108\/revisions\/109"}],"wp:attachment":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=108"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=108"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=108"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}