AI supported fetal echocardiography with quality assessment

Caroline A. Taksoee-Vester; Kamil Mikolaj; Zahra Bashir; Anders N. Christensen; Olav B. Petersen; Karin Sundberg; Aasa Feragen; Morten B. S. Svendsen; Mads Nielsen; Martin G. Tolsgaard

doi:10.1038/s41598-024-56476-6

AI supported fetal echocardiography with quality assessment

Caroline A. Taksoee-Vester^*, Kamil Mikolaj, Zahra Bashir, Anders N. Christensen, Olav B. Petersen, Karin Sundberg, Aasa Feragen, Morten B. S. Svendsen, Mads Nielsen, Martin G. Tolsgaard

^*Corresponding author for this work

Research output: Contribution to journal › Journal article › Research › peer-review

43 Downloads (Orbit)

Abstract

This study aimed to develop a deep learning model to assess the quality of fetal echocardiography and to perform prospective clinical validation. The model was trained on data from the 18-22-week anomaly scan conducted in seven hospitals from 2008 to 2018. Prospective validation involved 100 patients from two hospitals. A total of 5363 images from 2551 pregnancies were used for training and validation. The model's segmentation accuracy depended on image quality measured by a quality score (QS). It achieved an overall average accuracy of 0.91 (SD 0.09) across the test set, with images having above-average QS scoring 0.97 (SD 0.03). During prospective validation of 192 images, clinicians rated 44.8% (SD 9.8) of images as equal in quality, 18.69% (SD 5.7) favoring auto-captured images and 36.51% (SD 9.0) preferring manually captured ones. Images with above average QS showed better agreement on segmentations (p

Original language	English
Article number	5809
Journal	Scientific Reports
Volume	14
Number of pages	9
ISSN	2045-2322
DOIs	https://doi.org/10.1038/s41598-024-56476-6
Publication status	Published - 2024

Access to Document

10.1038/s41598-024-56476-6Licence: CC BY

FulltextFinal published version, 1.47 MBLicence: CC BY

OpenUrl availability

Full text

Cite this

@article{aee13060932242afb24d311534acdf7f,

title = "AI supported fetal echocardiography with quality assessment",

abstract = "This study aimed to develop a deep learning model to assess the quality of fetal echocardiography and to perform prospective clinical validation. The model was trained on data from the 18-22-week anomaly scan conducted in seven hospitals from 2008 to 2018. Prospective validation involved 100 patients from two hospitals. A total of 5363 images from 2551 pregnancies were used for training and validation. The model's segmentation accuracy depended on image quality measured by a quality score (QS). It achieved an overall average accuracy of 0.91 (SD 0.09) across the test set, with images having above-average QS scoring 0.97 (SD 0.03). During prospective validation of 192 images, clinicians rated 44.8% (SD 9.8) of images as equal in quality, 18.69% (SD 5.7) favoring auto-captured images and 36.51% (SD 9.0) preferring manually captured ones. Images with above average QS showed better agreement on segmentations (p ",

author = "Taksoee-Vester, {Caroline A.} and Kamil Mikolaj and Zahra Bashir and Christensen, {Anders N.} and Petersen, {Olav B.} and Karin Sundberg and Aasa Feragen and Svendsen, {Morten B. S.} and Mads Nielsen and Tolsgaard, {Martin G.}",

year = "2024",

doi = "10.1038/s41598-024-56476-6",

language = "English",

volume = "14",

journal = "Scientific Reports",

issn = "2045-2322",

publisher = "Nature Publishing Group",

}

TY - JOUR

T1 - AI supported fetal echocardiography with quality assessment

AU - Taksoee-Vester, Caroline A.

AU - Mikolaj, Kamil

AU - Bashir, Zahra

AU - Christensen, Anders N.

AU - Petersen, Olav B.

AU - Sundberg, Karin

AU - Feragen, Aasa

AU - Svendsen, Morten B. S.

AU - Nielsen, Mads

AU - Tolsgaard, Martin G.

PY - 2024

Y1 - 2024

N2 - This study aimed to develop a deep learning model to assess the quality of fetal echocardiography and to perform prospective clinical validation. The model was trained on data from the 18-22-week anomaly scan conducted in seven hospitals from 2008 to 2018. Prospective validation involved 100 patients from two hospitals. A total of 5363 images from 2551 pregnancies were used for training and validation. The model's segmentation accuracy depended on image quality measured by a quality score (QS). It achieved an overall average accuracy of 0.91 (SD 0.09) across the test set, with images having above-average QS scoring 0.97 (SD 0.03). During prospective validation of 192 images, clinicians rated 44.8% (SD 9.8) of images as equal in quality, 18.69% (SD 5.7) favoring auto-captured images and 36.51% (SD 9.0) preferring manually captured ones. Images with above average QS showed better agreement on segmentations (p

AB - This study aimed to develop a deep learning model to assess the quality of fetal echocardiography and to perform prospective clinical validation. The model was trained on data from the 18-22-week anomaly scan conducted in seven hospitals from 2008 to 2018. Prospective validation involved 100 patients from two hospitals. A total of 5363 images from 2551 pregnancies were used for training and validation. The model's segmentation accuracy depended on image quality measured by a quality score (QS). It achieved an overall average accuracy of 0.91 (SD 0.09) across the test set, with images having above-average QS scoring 0.97 (SD 0.03). During prospective validation of 192 images, clinicians rated 44.8% (SD 9.8) of images as equal in quality, 18.69% (SD 5.7) favoring auto-captured images and 36.51% (SD 9.0) preferring manually captured ones. Images with above average QS showed better agreement on segmentations (p

U2 - 10.1038/s41598-024-56476-6

DO - 10.1038/s41598-024-56476-6

M3 - Journal article

C2 - 38461322

SN - 2045-2322

VL - 14

JO - Scientific Reports

JF - Scientific Reports

M1 - 5809

ER -

AI supported fetal echocardiography with quality assessment

Abstract

Access to Document

OpenUrl availability

Fingerprint

Cite this