VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction

Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, an...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2022-06
Hauptverfasser: Desai, Arjun D, Gunel, Beliz, Ozturkler, Batu M, Beg, Harris, Vasanawala, Shreyas, Hargreaves, Brian A, Ré, Christopher, Pauly, John M, Chaudhari, Akshay S
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Desai, Arjun D
Gunel, Beliz
Ozturkler, Batu M
Beg, Harris
Vasanawala, Shreyas
Hargreaves, Brian A
Ré, Christopher
Pauly, John M
Chaudhari, Akshay S
description Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations. Our code is available at https://github.com/ad12/meddlr.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2593746319</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2593746319</sourcerecordid><originalsourceid>FETCH-proquest_journals_25937463193</originalsourceid><addsrcrecordid>eNqNjM0KwjAQhIMgKNp3WPBcqIm11pv4gx5EKVW8SYyrRjTRbCL49lbwATzNMN_M1FiTC9GNBz3OGywiuiZJwvsZT1PRZGa7Ksrpbgjry5u0onji9AsNTKSXMArnOxovvbaGYEPanGFcWU0ejXpD6aQ23_BkHRT2EMjDSCm8oZMej7AsFlCgqhbeBfV9abP6Sd4Io5-2WGc2Lcfz-OHsMyD5_dUGZyq052kusl5fdHPxX-sD4q5J2Q</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2593746319</pqid></control><display><type>article</type><title>VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction</title><source>Free E- Journals</source><creator>Desai, Arjun D ; Gunel, Beliz ; Ozturkler, Batu M ; Beg, Harris ; Vasanawala, Shreyas ; Hargreaves, Brian A ; Ré, Christopher ; Pauly, John M ; Chaudhari, Akshay S</creator><creatorcontrib>Desai, Arjun D ; Gunel, Beliz ; Ozturkler, Batu M ; Beg, Harris ; Vasanawala, Shreyas ; Hargreaves, Brian A ; Ré, Christopher ; Pauly, John M ; Chaudhari, Akshay S</creatorcontrib><description>Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations. Our code is available at https://github.com/ad12/meddlr.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Artificial neural networks ; Consistency ; Data augmentation ; Image quality ; Image reconstruction ; Inverse problems ; Magnetic resonance imaging ; Physics ; Robustness ; Signal to noise ratio ; Training</subject><ispartof>arXiv.org, 2022-06</ispartof><rights>2022. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>780,784</link.rule.ids></links><search><creatorcontrib>Desai, Arjun D</creatorcontrib><creatorcontrib>Gunel, Beliz</creatorcontrib><creatorcontrib>Ozturkler, Batu M</creatorcontrib><creatorcontrib>Beg, Harris</creatorcontrib><creatorcontrib>Vasanawala, Shreyas</creatorcontrib><creatorcontrib>Hargreaves, Brian A</creatorcontrib><creatorcontrib>Ré, Christopher</creatorcontrib><creatorcontrib>Pauly, John M</creatorcontrib><creatorcontrib>Chaudhari, Akshay S</creatorcontrib><title>VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction</title><title>arXiv.org</title><description>Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations. Our code is available at https://github.com/ad12/meddlr.</description><subject>Artificial neural networks</subject><subject>Consistency</subject><subject>Data augmentation</subject><subject>Image quality</subject><subject>Image reconstruction</subject><subject>Inverse problems</subject><subject>Magnetic resonance imaging</subject><subject>Physics</subject><subject>Robustness</subject><subject>Signal to noise ratio</subject><subject>Training</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNjM0KwjAQhIMgKNp3WPBcqIm11pv4gx5EKVW8SYyrRjTRbCL49lbwATzNMN_M1FiTC9GNBz3OGywiuiZJwvsZT1PRZGa7Ksrpbgjry5u0onji9AsNTKSXMArnOxovvbaGYEPanGFcWU0ejXpD6aQ23_BkHRT2EMjDSCm8oZMej7AsFlCgqhbeBfV9abP6Sd4Io5-2WGc2Lcfz-OHsMyD5_dUGZyq052kusl5fdHPxX-sD4q5J2Q</recordid><startdate>20220617</startdate><enddate>20220617</enddate><creator>Desai, Arjun D</creator><creator>Gunel, Beliz</creator><creator>Ozturkler, Batu M</creator><creator>Beg, Harris</creator><creator>Vasanawala, Shreyas</creator><creator>Hargreaves, Brian A</creator><creator>Ré, Christopher</creator><creator>Pauly, John M</creator><creator>Chaudhari, Akshay S</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20220617</creationdate><title>VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction</title><author>Desai, Arjun D ; Gunel, Beliz ; Ozturkler, Batu M ; Beg, Harris ; Vasanawala, Shreyas ; Hargreaves, Brian A ; Ré, Christopher ; Pauly, John M ; Chaudhari, Akshay S</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_25937463193</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Artificial neural networks</topic><topic>Consistency</topic><topic>Data augmentation</topic><topic>Image quality</topic><topic>Image reconstruction</topic><topic>Inverse problems</topic><topic>Magnetic resonance imaging</topic><topic>Physics</topic><topic>Robustness</topic><topic>Signal to noise ratio</topic><topic>Training</topic><toplevel>online_resources</toplevel><creatorcontrib>Desai, Arjun D</creatorcontrib><creatorcontrib>Gunel, Beliz</creatorcontrib><creatorcontrib>Ozturkler, Batu M</creatorcontrib><creatorcontrib>Beg, Harris</creatorcontrib><creatorcontrib>Vasanawala, Shreyas</creatorcontrib><creatorcontrib>Hargreaves, Brian A</creatorcontrib><creatorcontrib>Ré, Christopher</creatorcontrib><creatorcontrib>Pauly, John M</creatorcontrib><creatorcontrib>Chaudhari, Akshay S</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Access via ProQuest (Open Access)</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Desai, Arjun D</au><au>Gunel, Beliz</au><au>Ozturkler, Batu M</au><au>Beg, Harris</au><au>Vasanawala, Shreyas</au><au>Hargreaves, Brian A</au><au>Ré, Christopher</au><au>Pauly, John M</au><au>Chaudhari, Akshay S</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction</atitle><jtitle>arXiv.org</jtitle><date>2022-06-17</date><risdate>2022</risdate><eissn>2331-8422</eissn><abstract>Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations. Our code is available at https://github.com/ad12/meddlr.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2022-06
issn 2331-8422
language eng
recordid cdi_proquest_journals_2593746319
source Free E- Journals
subjects Artificial neural networks
Consistency
Data augmentation
Image quality
Image reconstruction
Inverse problems
Magnetic resonance imaging
Physics
Robustness
Signal to noise ratio
Training
title VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-30T21%3A21%3A37IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=VORTEX:%20Physics-Driven%20Data%20Augmentations%20Using%20Consistency%20Training%20for%20Robust%20Accelerated%20MRI%20Reconstruction&rft.jtitle=arXiv.org&rft.au=Desai,%20Arjun%20D&rft.date=2022-06-17&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2593746319%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2593746319&rft_id=info:pmid/&rfr_iscdi=true