Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans

This paper considers automatic clinical workflow description of full-length routine fetal anomaly ultrasound scans using deep learning approaches for spatio-temporal video analysis. Multiple architectures consisting of 2D and 2D + t CNN, LSTM, and convolutional LSTM are investigated and compared. Th...

Full description

Bibliographic Details
Main Authors: Sharma, H, Droste, R, Chatelain, P, Drukker, L, Papageorghiou, A, Noble, J
Format: Conference item
Published: IEEE 2019
_version_ 1826268019287916544
author Sharma, H
Droste, R
Chatelain, P
Drukker, L
Papageorghiou, A
Noble, J
author_facet Sharma, H
Droste, R
Chatelain, P
Drukker, L
Papageorghiou, A
Noble, J
author_sort Sharma, H
collection OXFORD
description This paper considers automatic clinical workflow description of full-length routine fetal anomaly ultrasound scans using deep learning approaches for spatio-temporal video analysis. Multiple architectures consisting of 2D and 2D + t CNN, LSTM, and convolutional LSTM are investigated and compared. The contributions of short-term and long-term temporal changes are studied, and a multi-stream framework analysis is found to achieve the best top-l accuracy =0.77 and top-3 accuracy =0.94. Automated partitioning and characterisation on unlabelled full-length video scans show high correlation (ρ=0.95, p=0.0004) with workflow statistics of manually labelled videos, suggesting practicality of proposed methods.
first_indexed 2024-03-06T21:03:10Z
format Conference item
id oxford-uuid:3b8ec67e-c443-40ac-a2ef-e8da14fb5f62
institution University of Oxford
last_indexed 2024-03-06T21:03:10Z
publishDate 2019
publisher IEEE
record_format dspace
spelling oxford-uuid:3b8ec67e-c443-40ac-a2ef-e8da14fb5f622022-03-26T14:08:20ZSpatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scansConference itemhttp://purl.org/coar/resource_type/c_5794uuid:3b8ec67e-c443-40ac-a2ef-e8da14fb5f62Symplectic Elements at OxfordIEEE2019Sharma, HDroste, RChatelain, PDrukker, LPapageorghiou, ANoble, JThis paper considers automatic clinical workflow description of full-length routine fetal anomaly ultrasound scans using deep learning approaches for spatio-temporal video analysis. Multiple architectures consisting of 2D and 2D + t CNN, LSTM, and convolutional LSTM are investigated and compared. The contributions of short-term and long-term temporal changes are studied, and a multi-stream framework analysis is found to achieve the best top-l accuracy =0.77 and top-3 accuracy =0.94. Automated partitioning and characterisation on unlabelled full-length video scans show high correlation (ρ=0.95, p=0.0004) with workflow statistics of manually labelled videos, suggesting practicality of proposed methods.
spellingShingle Sharma, H
Droste, R
Chatelain, P
Drukker, L
Papageorghiou, A
Noble, J
Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title_full Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title_fullStr Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title_full_unstemmed Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title_short Spatio-temporal partitioning and description of full-length routine fetal anomaly ultrasound scans
title_sort spatio temporal partitioning and description of full length routine fetal anomaly ultrasound scans
work_keys_str_mv AT sharmah spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans
AT droster spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans
AT chatelainp spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans
AT drukkerl spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans
AT papageorghioua spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans
AT noblej spatiotemporalpartitioninganddescriptionoffulllengthroutinefetalanomalyultrasoundscans