Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?
Background Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model. Methods Current screening...
Gespeichert in:
| Veröffentlicht in: | Prenatal diagnosis 2024-06, Vol.44 (6-7), p.717-724 |
|---|---|
| Hauptverfasser: | , , , , , , , , , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 724 |
|---|---|
| container_issue | 6-7 |
| container_start_page | 717 |
| container_title | Prenatal diagnosis |
| container_volume | 44 |
| creator | Day, Thomas G. Budd, Samuel Tan, Jeremy Matthew, Jacqueline Skelton, Emily Jowett, Victoria Lloyd, David Gomez, Alberto Hajnal, Jo V. Razavi, Reza Kainz, Bernhard Simpson, John M. |
| description | Background
Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model.
Methods
Current screening programme performance was calculated from local and national sources. AI models were trained using four‐chamber ultrasound views of the fetal heart, using a ResNet classifier.
Results
Estimated current fetal screening programme sensitivity and specificity for HLHS were 94.3% and 99.985%, respectively. Depending on calibration, AI models to detect HLHS were either highly sensitive (sensitivity 100%, specificity 94.0%) or highly specific (sensitivity 93.3%, specificity 100%). Our analysis suggests that our highly sensitive model would generate 45,134 screen positive results for a gain of 14 additional HLHS cases. Our highly specific model would be associated with two fewer detected HLHS cases, and 118 fewer false positives.
Conclusion
If used independently, our AI model performance is slightly worse than the performance level of the current screening programme in detecting HLHS, and this performance is likely to deteriorate further when used prospectively. This demonstrates that collaboration between humans and AI will be key for effective future clinical use.
Key points
What is already known on this topic?
Artificial intelligence (AI) can be used to interpret medical images and make diagnoses, including detecting fetal congenital heart disease (CHD) by ultrasound.
The sensitivity of the current English screening programme for fetal cardiac malformations is publicly available, but specificity is not reported.
What this study adds?
The current screening programme in our region is operating at a very high specificity for fetal hypoplastic left heart syndrome (HLHS).
Using a curated retrospective dataset, it is possible to train AI models to detect HLHS with a performance approaching that of the current screening programme.
Current AI models do not have high enough specificity to be used independently for screening for fetal CHD, meaning that human‐AI interaction when performing or interpreting ultrasound will be important to select cases for specialist referral. |
| doi_str_mv | 10.1002/pd.6445 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2870998132</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2870998132</sourcerecordid><originalsourceid>FETCH-LOGICAL-c3785-375c127e81dda6a5fab79a23b970dcdb60e41c8add104ece4586259f8918a2113</originalsourceid><addsrcrecordid>eNp1kV1LHDEUhkNp0a0t_Qcl0AsLsjYfM5OMNyK2akGoF-11OJucWSMzyTSZQfbn-E_NdtULoVc5hIfnfLyEfOLsmDMmvo3uuKmq-g1ZcNaqJRNCviULxkstdc33yfuc7wqoRav2yL5USjVMVwvycJMwwAQ9dR7WIWafaezo7WaMYw958pb22E30FiFNNG-CS3FAGgOd-ylBjnNwdM4-rGkBfOetLy4fJux7v8Zg8YRexXvqImY6YupiGqD8UhuHERLSKVKgdk5ljOK3CTFsZWOK6wTDgKcfyLsO-owfn94D8ufix-_zq-X1r8uf52fXSyuVrpdS1ZYLhZo7Bw3UHaxUC0KuWsWcdauGYcWtBuc4q9BiVetG1G2nW65BcC4PyNedt7T-O2OezOCzLWtAwDhnI7Ribau5FAX98gq9i3MKZTojWaNrJculC3W4o2yKOSfszJj8AGljODPb1MzozDa1Qn5-8s2rAd0L9xxTAY52wL3vcfM_j7n5_k_3COONowE</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3068573028</pqid></control><display><type>article</type><title>Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?</title><source>MEDLINE</source><source>Wiley Online Library Journals Frontfile Complete</source><creator>Day, Thomas G. ; Budd, Samuel ; Tan, Jeremy ; Matthew, Jacqueline ; Skelton, Emily ; Jowett, Victoria ; Lloyd, David ; Gomez, Alberto ; Hajnal, Jo V. ; Razavi, Reza ; Kainz, Bernhard ; Simpson, John M.</creator><creatorcontrib>Day, Thomas G. ; Budd, Samuel ; Tan, Jeremy ; Matthew, Jacqueline ; Skelton, Emily ; Jowett, Victoria ; Lloyd, David ; Gomez, Alberto ; Hajnal, Jo V. ; Razavi, Reza ; Kainz, Bernhard ; Simpson, John M.</creatorcontrib><description>Background
Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model.
Methods
Current screening programme performance was calculated from local and national sources. AI models were trained using four‐chamber ultrasound views of the fetal heart, using a ResNet classifier.
Results
Estimated current fetal screening programme sensitivity and specificity for HLHS were 94.3% and 99.985%, respectively. Depending on calibration, AI models to detect HLHS were either highly sensitive (sensitivity 100%, specificity 94.0%) or highly specific (sensitivity 93.3%, specificity 100%). Our analysis suggests that our highly sensitive model would generate 45,134 screen positive results for a gain of 14 additional HLHS cases. Our highly specific model would be associated with two fewer detected HLHS cases, and 118 fewer false positives.
Conclusion
If used independently, our AI model performance is slightly worse than the performance level of the current screening programme in detecting HLHS, and this performance is likely to deteriorate further when used prospectively. This demonstrates that collaboration between humans and AI will be key for effective future clinical use.
Key points
What is already known on this topic?
Artificial intelligence (AI) can be used to interpret medical images and make diagnoses, including detecting fetal congenital heart disease (CHD) by ultrasound.
The sensitivity of the current English screening programme for fetal cardiac malformations is publicly available, but specificity is not reported.
What this study adds?
The current screening programme in our region is operating at a very high specificity for fetal hypoplastic left heart syndrome (HLHS).
Using a curated retrospective dataset, it is possible to train AI models to detect HLHS with a performance approaching that of the current screening programme.
Current AI models do not have high enough specificity to be used independently for screening for fetal CHD, meaning that human‐AI interaction when performing or interpreting ultrasound will be important to select cases for specialist referral.</description><identifier>ISSN: 0197-3851</identifier><identifier>ISSN: 1097-0223</identifier><identifier>EISSN: 1097-0223</identifier><identifier>DOI: 10.1002/pd.6445</identifier><identifier>PMID: 37776084</identifier><language>eng</language><publisher>England: Wiley Subscription Services, Inc</publisher><subject>Artificial Intelligence ; Cardiovascular diseases ; Female ; Fetuses ; Heart ; Heart diseases ; Humans ; Hypoplastic Left Heart Syndrome - diagnostic imaging ; Medical diagnosis ; Pregnancy ; Prenatal diagnosis ; Sensitivity and Specificity ; Ultrasonic imaging ; Ultrasonography, Prenatal - methods ; Ultrasonography, Prenatal - standards ; Ultrasound</subject><ispartof>Prenatal diagnosis, 2024-06, Vol.44 (6-7), p.717-724</ispartof><rights>2023 The Authors. Prenatal Diagnosis published by John Wiley & Sons Ltd.</rights><rights>2023. This article is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c3785-375c127e81dda6a5fab79a23b970dcdb60e41c8add104ece4586259f8918a2113</citedby><cites>FETCH-LOGICAL-c3785-375c127e81dda6a5fab79a23b970dcdb60e41c8add104ece4586259f8918a2113</cites><orcidid>0000-0001-8391-7903</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://onlinelibrary.wiley.com/doi/pdf/10.1002%2Fpd.6445$$EPDF$$P50$$Gwiley$$Hfree_for_read</linktopdf><linktohtml>$$Uhttps://onlinelibrary.wiley.com/doi/full/10.1002%2Fpd.6445$$EHTML$$P50$$Gwiley$$Hfree_for_read</linktohtml><link.rule.ids>314,776,780,1411,27901,27902,45550,45551</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/37776084$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Day, Thomas G.</creatorcontrib><creatorcontrib>Budd, Samuel</creatorcontrib><creatorcontrib>Tan, Jeremy</creatorcontrib><creatorcontrib>Matthew, Jacqueline</creatorcontrib><creatorcontrib>Skelton, Emily</creatorcontrib><creatorcontrib>Jowett, Victoria</creatorcontrib><creatorcontrib>Lloyd, David</creatorcontrib><creatorcontrib>Gomez, Alberto</creatorcontrib><creatorcontrib>Hajnal, Jo V.</creatorcontrib><creatorcontrib>Razavi, Reza</creatorcontrib><creatorcontrib>Kainz, Bernhard</creatorcontrib><creatorcontrib>Simpson, John M.</creatorcontrib><title>Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?</title><title>Prenatal diagnosis</title><addtitle>Prenat Diagn</addtitle><description>Background
Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model.
Methods
Current screening programme performance was calculated from local and national sources. AI models were trained using four‐chamber ultrasound views of the fetal heart, using a ResNet classifier.
Results
Estimated current fetal screening programme sensitivity and specificity for HLHS were 94.3% and 99.985%, respectively. Depending on calibration, AI models to detect HLHS were either highly sensitive (sensitivity 100%, specificity 94.0%) or highly specific (sensitivity 93.3%, specificity 100%). Our analysis suggests that our highly sensitive model would generate 45,134 screen positive results for a gain of 14 additional HLHS cases. Our highly specific model would be associated with two fewer detected HLHS cases, and 118 fewer false positives.
Conclusion
If used independently, our AI model performance is slightly worse than the performance level of the current screening programme in detecting HLHS, and this performance is likely to deteriorate further when used prospectively. This demonstrates that collaboration between humans and AI will be key for effective future clinical use.
Key points
What is already known on this topic?
Artificial intelligence (AI) can be used to interpret medical images and make diagnoses, including detecting fetal congenital heart disease (CHD) by ultrasound.
The sensitivity of the current English screening programme for fetal cardiac malformations is publicly available, but specificity is not reported.
What this study adds?
The current screening programme in our region is operating at a very high specificity for fetal hypoplastic left heart syndrome (HLHS).
Using a curated retrospective dataset, it is possible to train AI models to detect HLHS with a performance approaching that of the current screening programme.
Current AI models do not have high enough specificity to be used independently for screening for fetal CHD, meaning that human‐AI interaction when performing or interpreting ultrasound will be important to select cases for specialist referral.</description><subject>Artificial Intelligence</subject><subject>Cardiovascular diseases</subject><subject>Female</subject><subject>Fetuses</subject><subject>Heart</subject><subject>Heart diseases</subject><subject>Humans</subject><subject>Hypoplastic Left Heart Syndrome - diagnostic imaging</subject><subject>Medical diagnosis</subject><subject>Pregnancy</subject><subject>Prenatal diagnosis</subject><subject>Sensitivity and Specificity</subject><subject>Ultrasonic imaging</subject><subject>Ultrasonography, Prenatal - methods</subject><subject>Ultrasonography, Prenatal - standards</subject><subject>Ultrasound</subject><issn>0197-3851</issn><issn>1097-0223</issn><issn>1097-0223</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>24P</sourceid><sourceid>EIF</sourceid><recordid>eNp1kV1LHDEUhkNp0a0t_Qcl0AsLsjYfM5OMNyK2akGoF-11OJucWSMzyTSZQfbn-E_NdtULoVc5hIfnfLyEfOLsmDMmvo3uuKmq-g1ZcNaqJRNCviULxkstdc33yfuc7wqoRav2yL5USjVMVwvycJMwwAQ9dR7WIWafaezo7WaMYw958pb22E30FiFNNG-CS3FAGgOd-ylBjnNwdM4-rGkBfOetLy4fJux7v8Zg8YRexXvqImY6YupiGqD8UhuHERLSKVKgdk5ljOK3CTFsZWOK6wTDgKcfyLsO-owfn94D8ufix-_zq-X1r8uf52fXSyuVrpdS1ZYLhZo7Bw3UHaxUC0KuWsWcdauGYcWtBuc4q9BiVetG1G2nW65BcC4PyNedt7T-O2OezOCzLWtAwDhnI7Ribau5FAX98gq9i3MKZTojWaNrJculC3W4o2yKOSfszJj8AGljODPb1MzozDa1Qn5-8s2rAd0L9xxTAY52wL3vcfM_j7n5_k_3COONowE</recordid><startdate>202406</startdate><enddate>202406</enddate><creator>Day, Thomas G.</creator><creator>Budd, Samuel</creator><creator>Tan, Jeremy</creator><creator>Matthew, Jacqueline</creator><creator>Skelton, Emily</creator><creator>Jowett, Victoria</creator><creator>Lloyd, David</creator><creator>Gomez, Alberto</creator><creator>Hajnal, Jo V.</creator><creator>Razavi, Reza</creator><creator>Kainz, Bernhard</creator><creator>Simpson, John M.</creator><general>Wiley Subscription Services, Inc</general><scope>24P</scope><scope>CGR</scope><scope>CUY</scope><scope>CVF</scope><scope>ECM</scope><scope>EIF</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7QP</scope><scope>7T5</scope><scope>7T7</scope><scope>7TK</scope><scope>7TM</scope><scope>8FD</scope><scope>C1K</scope><scope>FR3</scope><scope>H94</scope><scope>K9.</scope><scope>P64</scope><scope>RC3</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0001-8391-7903</orcidid></search><sort><creationdate>202406</creationdate><title>Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?</title><author>Day, Thomas G. ; Budd, Samuel ; Tan, Jeremy ; Matthew, Jacqueline ; Skelton, Emily ; Jowett, Victoria ; Lloyd, David ; Gomez, Alberto ; Hajnal, Jo V. ; Razavi, Reza ; Kainz, Bernhard ; Simpson, John M.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c3785-375c127e81dda6a5fab79a23b970dcdb60e41c8add104ece4586259f8918a2113</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Artificial Intelligence</topic><topic>Cardiovascular diseases</topic><topic>Female</topic><topic>Fetuses</topic><topic>Heart</topic><topic>Heart diseases</topic><topic>Humans</topic><topic>Hypoplastic Left Heart Syndrome - diagnostic imaging</topic><topic>Medical diagnosis</topic><topic>Pregnancy</topic><topic>Prenatal diagnosis</topic><topic>Sensitivity and Specificity</topic><topic>Ultrasonic imaging</topic><topic>Ultrasonography, Prenatal - methods</topic><topic>Ultrasonography, Prenatal - standards</topic><topic>Ultrasound</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Day, Thomas G.</creatorcontrib><creatorcontrib>Budd, Samuel</creatorcontrib><creatorcontrib>Tan, Jeremy</creatorcontrib><creatorcontrib>Matthew, Jacqueline</creatorcontrib><creatorcontrib>Skelton, Emily</creatorcontrib><creatorcontrib>Jowett, Victoria</creatorcontrib><creatorcontrib>Lloyd, David</creatorcontrib><creatorcontrib>Gomez, Alberto</creatorcontrib><creatorcontrib>Hajnal, Jo V.</creatorcontrib><creatorcontrib>Razavi, Reza</creatorcontrib><creatorcontrib>Kainz, Bernhard</creatorcontrib><creatorcontrib>Simpson, John M.</creatorcontrib><collection>Wiley Online Library Open Access</collection><collection>Medline</collection><collection>MEDLINE</collection><collection>MEDLINE (Ovid)</collection><collection>MEDLINE</collection><collection>MEDLINE</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>Calcium & Calcified Tissue Abstracts</collection><collection>Immunology Abstracts</collection><collection>Industrial and Applied Microbiology Abstracts (Microbiology A)</collection><collection>Neurosciences Abstracts</collection><collection>Nucleic Acids Abstracts</collection><collection>Technology Research Database</collection><collection>Environmental Sciences and Pollution Management</collection><collection>Engineering Research Database</collection><collection>AIDS and Cancer Research Abstracts</collection><collection>ProQuest Health & Medical Complete (Alumni)</collection><collection>Biotechnology and BioEngineering Abstracts</collection><collection>Genetics Abstracts</collection><collection>MEDLINE - Academic</collection><jtitle>Prenatal diagnosis</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Day, Thomas G.</au><au>Budd, Samuel</au><au>Tan, Jeremy</au><au>Matthew, Jacqueline</au><au>Skelton, Emily</au><au>Jowett, Victoria</au><au>Lloyd, David</au><au>Gomez, Alberto</au><au>Hajnal, Jo V.</au><au>Razavi, Reza</au><au>Kainz, Bernhard</au><au>Simpson, John M.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme?</atitle><jtitle>Prenatal diagnosis</jtitle><addtitle>Prenat Diagn</addtitle><date>2024-06</date><risdate>2024</risdate><volume>44</volume><issue>6-7</issue><spage>717</spage><epage>724</epage><pages>717-724</pages><issn>0197-3851</issn><issn>1097-0223</issn><eissn>1097-0223</eissn><abstract>Background
Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed the performance of the current national screening programme in detecting hypoplastic left heart syndrome (HLHS) to compare with our own AI model.
Methods
Current screening programme performance was calculated from local and national sources. AI models were trained using four‐chamber ultrasound views of the fetal heart, using a ResNet classifier.
Results
Estimated current fetal screening programme sensitivity and specificity for HLHS were 94.3% and 99.985%, respectively. Depending on calibration, AI models to detect HLHS were either highly sensitive (sensitivity 100%, specificity 94.0%) or highly specific (sensitivity 93.3%, specificity 100%). Our analysis suggests that our highly sensitive model would generate 45,134 screen positive results for a gain of 14 additional HLHS cases. Our highly specific model would be associated with two fewer detected HLHS cases, and 118 fewer false positives.
Conclusion
If used independently, our AI model performance is slightly worse than the performance level of the current screening programme in detecting HLHS, and this performance is likely to deteriorate further when used prospectively. This demonstrates that collaboration between humans and AI will be key for effective future clinical use.
Key points
What is already known on this topic?
Artificial intelligence (AI) can be used to interpret medical images and make diagnoses, including detecting fetal congenital heart disease (CHD) by ultrasound.
The sensitivity of the current English screening programme for fetal cardiac malformations is publicly available, but specificity is not reported.
What this study adds?
The current screening programme in our region is operating at a very high specificity for fetal hypoplastic left heart syndrome (HLHS).
Using a curated retrospective dataset, it is possible to train AI models to detect HLHS with a performance approaching that of the current screening programme.
Current AI models do not have high enough specificity to be used independently for screening for fetal CHD, meaning that human‐AI interaction when performing or interpreting ultrasound will be important to select cases for specialist referral.</abstract><cop>England</cop><pub>Wiley Subscription Services, Inc</pub><pmid>37776084</pmid><doi>10.1002/pd.6445</doi><tpages>8</tpages><orcidid>https://orcid.org/0000-0001-8391-7903</orcidid><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0197-3851 |
| ispartof | Prenatal diagnosis, 2024-06, Vol.44 (6-7), p.717-724 |
| issn | 0197-3851 1097-0223 1097-0223 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_2870998132 |
| source | MEDLINE; Wiley Online Library Journals Frontfile Complete |
| subjects | Artificial Intelligence Cardiovascular diseases Female Fetuses Heart Heart diseases Humans Hypoplastic Left Heart Syndrome - diagnostic imaging Medical diagnosis Pregnancy Prenatal diagnosis Sensitivity and Specificity Ultrasonic imaging Ultrasonography, Prenatal - methods Ultrasonography, Prenatal - standards Ultrasound |
| title | Prenatal diagnosis of hypoplastic left heart syndrome on ultrasound using artificial intelligence: How does performance compare to a current screening programme? |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-04T03%3A24%3A08IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Prenatal%20diagnosis%20of%20hypoplastic%20left%20heart%20syndrome%20on%20ultrasound%20using%20artificial%20intelligence:%20How%20does%20performance%20compare%20to%20a%20current%20screening%20programme?&rft.jtitle=Prenatal%20diagnosis&rft.au=Day,%20Thomas%20G.&rft.date=2024-06&rft.volume=44&rft.issue=6-7&rft.spage=717&rft.epage=724&rft.pages=717-724&rft.issn=0197-3851&rft.eissn=1097-0223&rft_id=info:doi/10.1002/pd.6445&rft_dat=%3Cproquest_cross%3E2870998132%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=3068573028&rft_id=info:pmid/37776084&rfr_iscdi=true |