Coil Sketching for computationally-efficient MR iterative reconstruction

Purpose: Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for 3D non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during rec...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Oscanoa, Julio A, Ong, Frank, Iyer, Siddharth S, Li, Zhitao, Sandino, Christopher M, Ozturkler, Batu, Ennis, Daniel B, Pilanci, Mert, Vasanawala, Shreyas S
Format: Artikel
Sprache:eng
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Oscanoa, Julio A
Ong, Frank
Iyer, Siddharth S
Li, Zhitao
Sandino, Christopher M
Ozturkler, Batu
Ennis, Daniel B
Pilanci, Mert
Vasanawala, Shreyas S
description Purpose: Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for 3D non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during reconstruction as in coil compression. While effective for Cartesian imaging, coil compression inherently loses signal energy, producing shading artifacts that compromise image quality for 3D non-Cartesian imaging. We propose coil sketching, a general and versatile method for computationally-efficient iterative MR image reconstruction. Theory and Methods: We based our method on randomized sketching algorithms, a type of large-scale optimization algorithms well established in the fields of machine learning and big data analysis. We adapt the sketching theory to the MRI reconstruction problem via a structured sketching matrix that, similar to coil compression, considers high-energy virtual coils obtained from principal component analysis. But, unlike coil compression, it also considers random linear combinations of the remaining low-energy coils, effectively leveraging information from all coils. Results: First, we performed ablation experiments to validate the sketching matrix design on both Cartesian and non-Cartesian datasets. The resulting design yielded both improved computational efficiency and preserved signal-to-noise ratio (SNR) as measured by the inverse g-factor. Then, we verified the efficacy of our approach on high-dimensional non-Cartesian 3D cones datasets, where coil sketching yielded up to three-fold faster reconstructions with equivalent image quality. Conclusion: Coil sketching is a general and versatile reconstruction framework for computationally fast and memory-efficient reconstruction.
doi_str_mv 10.48550/arxiv.2305.06482
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2305_06482</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2305_06482</sourcerecordid><originalsourceid>FETCH-LOGICAL-a672-7fd32bce75f7e2d955a2363dcd959f475ae810897a8b1200109988850b1ac5d53</originalsourceid><addsrcrecordid>eNotj8FOwzAQRH3hgAofwAn_QIJjZ2P7iCKgSEWVoPdo46yp1TSuXLeif09bOM1I8zTSY-yhEmVtAMQTpp9wLKUSUIqmNvKWzdsYRv61oezWYfrmPibu4nZ3yJhDnHAcTwV5H1ygKfOPTx4ypfN0JJ7IxWmf08FdyDt243Hc0_1_ztjq9WXVzovF8u29fV4U2GhZaD8o2TvS4DXJwQKgVI0a3LlaX2tAMpUwVqPpKylEJaw1xoDoK3QwgJqxx7_bq0q3S2GL6dRdlLqrkvoFMIxHKg</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Coil Sketching for computationally-efficient MR iterative reconstruction</title><source>arXiv.org</source><creator>Oscanoa, Julio A ; Ong, Frank ; Iyer, Siddharth S ; Li, Zhitao ; Sandino, Christopher M ; Ozturkler, Batu ; Ennis, Daniel B ; Pilanci, Mert ; Vasanawala, Shreyas S</creator><creatorcontrib>Oscanoa, Julio A ; Ong, Frank ; Iyer, Siddharth S ; Li, Zhitao ; Sandino, Christopher M ; Ozturkler, Batu ; Ennis, Daniel B ; Pilanci, Mert ; Vasanawala, Shreyas S</creatorcontrib><description>Purpose: Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for 3D non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during reconstruction as in coil compression. While effective for Cartesian imaging, coil compression inherently loses signal energy, producing shading artifacts that compromise image quality for 3D non-Cartesian imaging. We propose coil sketching, a general and versatile method for computationally-efficient iterative MR image reconstruction. Theory and Methods: We based our method on randomized sketching algorithms, a type of large-scale optimization algorithms well established in the fields of machine learning and big data analysis. We adapt the sketching theory to the MRI reconstruction problem via a structured sketching matrix that, similar to coil compression, considers high-energy virtual coils obtained from principal component analysis. But, unlike coil compression, it also considers random linear combinations of the remaining low-energy coils, effectively leveraging information from all coils. Results: First, we performed ablation experiments to validate the sketching matrix design on both Cartesian and non-Cartesian datasets. The resulting design yielded both improved computational efficiency and preserved signal-to-noise ratio (SNR) as measured by the inverse g-factor. Then, we verified the efficacy of our approach on high-dimensional non-Cartesian 3D cones datasets, where coil sketching yielded up to three-fold faster reconstructions with equivalent image quality. Conclusion: Coil sketching is a general and versatile reconstruction framework for computationally fast and memory-efficient reconstruction.</description><identifier>DOI: 10.48550/arxiv.2305.06482</identifier><language>eng</language><creationdate>2023-05</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2305.06482$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2305.06482$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Oscanoa, Julio A</creatorcontrib><creatorcontrib>Ong, Frank</creatorcontrib><creatorcontrib>Iyer, Siddharth S</creatorcontrib><creatorcontrib>Li, Zhitao</creatorcontrib><creatorcontrib>Sandino, Christopher M</creatorcontrib><creatorcontrib>Ozturkler, Batu</creatorcontrib><creatorcontrib>Ennis, Daniel B</creatorcontrib><creatorcontrib>Pilanci, Mert</creatorcontrib><creatorcontrib>Vasanawala, Shreyas S</creatorcontrib><title>Coil Sketching for computationally-efficient MR iterative reconstruction</title><description>Purpose: Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for 3D non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during reconstruction as in coil compression. While effective for Cartesian imaging, coil compression inherently loses signal energy, producing shading artifacts that compromise image quality for 3D non-Cartesian imaging. We propose coil sketching, a general and versatile method for computationally-efficient iterative MR image reconstruction. Theory and Methods: We based our method on randomized sketching algorithms, a type of large-scale optimization algorithms well established in the fields of machine learning and big data analysis. We adapt the sketching theory to the MRI reconstruction problem via a structured sketching matrix that, similar to coil compression, considers high-energy virtual coils obtained from principal component analysis. But, unlike coil compression, it also considers random linear combinations of the remaining low-energy coils, effectively leveraging information from all coils. Results: First, we performed ablation experiments to validate the sketching matrix design on both Cartesian and non-Cartesian datasets. The resulting design yielded both improved computational efficiency and preserved signal-to-noise ratio (SNR) as measured by the inverse g-factor. Then, we verified the efficacy of our approach on high-dimensional non-Cartesian 3D cones datasets, where coil sketching yielded up to three-fold faster reconstructions with equivalent image quality. Conclusion: Coil sketching is a general and versatile reconstruction framework for computationally fast and memory-efficient reconstruction.</description><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj8FOwzAQRH3hgAofwAn_QIJjZ2P7iCKgSEWVoPdo46yp1TSuXLeif09bOM1I8zTSY-yhEmVtAMQTpp9wLKUSUIqmNvKWzdsYRv61oezWYfrmPibu4nZ3yJhDnHAcTwV5H1ygKfOPTx4ypfN0JJ7IxWmf08FdyDt243Hc0_1_ztjq9WXVzovF8u29fV4U2GhZaD8o2TvS4DXJwQKgVI0a3LlaX2tAMpUwVqPpKylEJaw1xoDoK3QwgJqxx7_bq0q3S2GL6dRdlLqrkvoFMIxHKg</recordid><startdate>20230510</startdate><enddate>20230510</enddate><creator>Oscanoa, Julio A</creator><creator>Ong, Frank</creator><creator>Iyer, Siddharth S</creator><creator>Li, Zhitao</creator><creator>Sandino, Christopher M</creator><creator>Ozturkler, Batu</creator><creator>Ennis, Daniel B</creator><creator>Pilanci, Mert</creator><creator>Vasanawala, Shreyas S</creator><scope>GOX</scope></search><sort><creationdate>20230510</creationdate><title>Coil Sketching for computationally-efficient MR iterative reconstruction</title><author>Oscanoa, Julio A ; Ong, Frank ; Iyer, Siddharth S ; Li, Zhitao ; Sandino, Christopher M ; Ozturkler, Batu ; Ennis, Daniel B ; Pilanci, Mert ; Vasanawala, Shreyas S</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a672-7fd32bce75f7e2d955a2363dcd959f475ae810897a8b1200109988850b1ac5d53</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><toplevel>online_resources</toplevel><creatorcontrib>Oscanoa, Julio A</creatorcontrib><creatorcontrib>Ong, Frank</creatorcontrib><creatorcontrib>Iyer, Siddharth S</creatorcontrib><creatorcontrib>Li, Zhitao</creatorcontrib><creatorcontrib>Sandino, Christopher M</creatorcontrib><creatorcontrib>Ozturkler, Batu</creatorcontrib><creatorcontrib>Ennis, Daniel B</creatorcontrib><creatorcontrib>Pilanci, Mert</creatorcontrib><creatorcontrib>Vasanawala, Shreyas S</creatorcontrib><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Oscanoa, Julio A</au><au>Ong, Frank</au><au>Iyer, Siddharth S</au><au>Li, Zhitao</au><au>Sandino, Christopher M</au><au>Ozturkler, Batu</au><au>Ennis, Daniel B</au><au>Pilanci, Mert</au><au>Vasanawala, Shreyas S</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Coil Sketching for computationally-efficient MR iterative reconstruction</atitle><date>2023-05-10</date><risdate>2023</risdate><abstract>Purpose: Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for 3D non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during reconstruction as in coil compression. While effective for Cartesian imaging, coil compression inherently loses signal energy, producing shading artifacts that compromise image quality for 3D non-Cartesian imaging. We propose coil sketching, a general and versatile method for computationally-efficient iterative MR image reconstruction. Theory and Methods: We based our method on randomized sketching algorithms, a type of large-scale optimization algorithms well established in the fields of machine learning and big data analysis. We adapt the sketching theory to the MRI reconstruction problem via a structured sketching matrix that, similar to coil compression, considers high-energy virtual coils obtained from principal component analysis. But, unlike coil compression, it also considers random linear combinations of the remaining low-energy coils, effectively leveraging information from all coils. Results: First, we performed ablation experiments to validate the sketching matrix design on both Cartesian and non-Cartesian datasets. The resulting design yielded both improved computational efficiency and preserved signal-to-noise ratio (SNR) as measured by the inverse g-factor. Then, we verified the efficacy of our approach on high-dimensional non-Cartesian 3D cones datasets, where coil sketching yielded up to three-fold faster reconstructions with equivalent image quality. Conclusion: Coil sketching is a general and versatile reconstruction framework for computationally fast and memory-efficient reconstruction.</abstract><doi>10.48550/arxiv.2305.06482</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2305.06482
ispartof
issn
language eng
recordid cdi_arxiv_primary_2305_06482
source arXiv.org
title Coil Sketching for computationally-efficient MR iterative reconstruction
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T07%3A23%3A51IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Coil%20Sketching%20for%20computationally-efficient%20MR%20iterative%20reconstruction&rft.au=Oscanoa,%20Julio%20A&rft.date=2023-05-10&rft_id=info:doi/10.48550/arxiv.2305.06482&rft_dat=%3Carxiv_GOX%3E2305_06482%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true