A Multicenter, Scan-Rescan, Human and Machine Learning CMR Study to Test Generalizability and Precision in Imaging Biomarker Analysis.
dc.contributor.author | Bhuva, A | en_US |
dc.contributor.author | Bai, W | en_US |
dc.contributor.author | Lau, C | en_US |
dc.contributor.author | Davies, R | en_US |
dc.contributor.author | Ye, Y | en_US |
dc.contributor.author | Bulluck, H | en_US |
dc.contributor.author | McAlindon, E | en_US |
dc.contributor.author | Culotta, V | en_US |
dc.contributor.author | Swoboda, P | en_US |
dc.contributor.author | Captur, G | en_US |
dc.contributor.author | Treibel, T | en_US |
dc.contributor.author | Augusto, J | en_US |
dc.contributor.author | Knott, K | en_US |
dc.contributor.author | Seraphim, A | en_US |
dc.contributor.author | Cole, G | en_US |
dc.contributor.author | Petersen, S | en_US |
dc.contributor.author | Edwards, N | en_US |
dc.contributor.author | Greenwood, J | en_US |
dc.contributor.author | Bucciarelli-Ducci, C | en_US |
dc.contributor.author | Hughes, A | en_US |
dc.contributor.author | Rueckert, D | en_US |
dc.contributor.author | Moon, J | en_US |
dc.contributor.author | Manisty, C | en_US |
dc.date.accessioned | 2020-02-10T08:56:57Z | |
dc.date.available | 2019-07-25 | en_US |
dc.date.issued | 2019-10 | en_US |
dc.identifier.uri | https://qmro.qmul.ac.uk/xmlui/handle/123456789/62698 | |
dc.description.abstract | BACKGROUND: Automated analysis of cardiac structure and function using machine learning (ML) has great potential, but is currently hindered by poor generalizability. Comparison is traditionally against clinicians as a reference, ignoring inherent human inter- and intraobserver error, and ensuring that ML cannot demonstrate superiority. Measuring precision (scan:rescan reproducibility) addresses this. We compared precision of ML and humans using a multicenter, multi-disease, scan:rescan cardiovascular magnetic resonance data set. METHODS: One hundred ten patients (5 disease categories, 5 institutions, 2 scanner manufacturers, and 2 field strengths) underwent scan:rescan cardiovascular magnetic resonance (96% within one week). After identification of the most precise human technique, left ventricular chamber volumes, mass, and ejection fraction were measured by an expert, a trained junior clinician, and a fully automated convolutional neural network trained on 599 independent multicenter disease cases. Scan:rescan coefficient of variation and 1000 bootstrapped 95% CIs were calculated and compared using mixed linear effects models. RESULTS: Clinicians can be confident in detecting a 9% change in left ventricular ejection fraction, with greater than half of coefficient of variation attributable to intraobserver variation. Expert, trained junior, and automated scan:rescan precision were similar (for left ventricular ejection fraction, coefficient of variation 6.1 [5.2%-7.1%], P=0.2581; 8.3 [5.6%-10.3%], P=0.3653; 8.8 [6.1%-11.1%], P=0.8620). Automated analysis was 186× faster than humans (0.07 versus 13 minutes). CONCLUSIONS: Automated ML analysis is faster with similar precision to the most precise human techniques, even when challenged with real-world scan:rescan data. Assessment of multicenter, multi-vendor, multi-field strength scan:rescan data (available at www.thevolumesresource.com) permits a generalizable assessment of ML precision and may facilitate direct translation of ML to clinical practice. | en_US |
dc.format.extent | e009214 - ? | en_US |
dc.language | eng | en_US |
dc.language.iso | en | en_US |
dc.relation.ispartof | Circ Cardiovasc Imaging | en_US |
dc.rights | Creative Commons Attribution Non-Commercial-NoDerivs License | |
dc.rights.uri | https://creativecommons.org/licenses/by-nc-nd/4.0/ | |
dc.subject | artificial intelligence | en_US |
dc.subject | image processing | en_US |
dc.subject | left ventricular remodeling | en_US |
dc.subject | magnetic resonance imaging, cine | en_US |
dc.subject | ventricular function | en_US |
dc.title | A Multicenter, Scan-Rescan, Human and Machine Learning CMR Study to Test Generalizability and Precision in Imaging Biomarker Analysis. | en_US |
dc.type | Article | |
dc.rights.holder | © 2019 The Authors. | |
dc.identifier.doi | 10.1161/CIRCIMAGING.119.009214 | en_US |
pubs.author-url | https://www.ncbi.nlm.nih.gov/pubmed/31547689 | en_US |
pubs.issue | 10 | en_US |
pubs.notes | Not known | en_US |
pubs.publication-status | Published | en_US |
pubs.volume | 12 | en_US |
rioxxterms.funder | Default funder | en_US |
rioxxterms.identifier.project | Default project | en_US |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |
This item appears in the following Collection(s)
-
NIHR Advanced Imaging [386]