Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage
•This study assessed if deep learning models (DLMs) improve neurosurgery residents’ ability to detect aneurysms on CTA.•Residents’ detection sensitivities without DLM: 77.8% (1st year), 86.5% (3rd year), 87.3% (5th year).•With DLM: sensitivities rose to 97.6%, 95.2%, and 98.4%, averaging a 13.2% inc...
Gespeichert in:
| Veröffentlicht in: | Journal of clinical neuroscience 2025-02, Vol.132, p.110971, Article 110971 |
|---|---|
| Hauptverfasser: | , , , , , , , , , , , , , , , |
| 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 | 110971 |
| container_title | Journal of clinical neuroscience |
| container_volume | 132 |
| creator | Goertz, Lukas Jünger, Stephanie T. Reinecke, David von Spreckelsen, Niklas Shahzad, Rahil Thiele, Frank Laukamp, Kai Roman Timmer, Marco Gertz, Roman Johannes Gietzen, Carsten Kaya, Kenan Grunz, Jan-Peter Schlamann, Marc Kabbasch, Christoph Borggrefe, Jan Pennig, Lenhard |
| description | •This study assessed if deep learning models (DLMs) improve neurosurgery residents’ ability to detect aneurysms on CTA.•Residents’ detection sensitivities without DLM: 77.8% (1st year), 86.5% (3rd year), 87.3% (5th year).•With DLM: sensitivities rose to 97.6%, 95.2%, and 98.4%, averaging a 13.2% increase.•DLMs assist neurosurgeons in making treatment decisions when immediate radiology consults are not available.
The purpose of this study was to evaluate the effectiveness of a deep learning model (DLM) in improving the sensitivity of neurosurgery residents to detect intracranial aneurysms on CT angiography (CTA) in patients with aneurysmal subarachnoid hemorrhage (aSAH).
In this diagnostic accuracy study, a set of 104 CTA scans of aSAH patients containing a total of 126 aneurysms were presented to three blinded neurosurgery residents (a first-year, third-year, and fifth-year resident), who individually assessed them for aneurysms. After the initial reading, the residents were given the predictions of a dedicated DLM previously established for automated detection and segmentation of intracranial aneurysms. The detection sensitivities for aneurysms of the DLM and the residents with and without the assistance of the DLM were compared.
The DLM had a detection sensitivity of 85.7%, while the residents showed detection sensitivities of 77.8%, 86.5%, and 87.3% without DLM assistance. After being provided with the DLM’s results, the residents’ individual detection sensitivities increased to 97.6%, 95.2%, and 98.4%, respectively, yielding an average increase of 13.2%. The DLM was particularly useful in detecting small aneurysms. In addition, interrater agreement among residents increased from a Fleiss κ of 0.394 without DLM assistance to 0.703 with DLM assistance.
The results of this pilot study suggest that deep learning models can help neurosurgeons detect aneurysms on CTA and make appropriate treatment decisions when immediate radiological consultation is not possible. |
| doi_str_mv | 10.1016/j.jocn.2024.110971 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_3146776570</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S0967586824005101</els_id><sourcerecordid>3146776570</sourcerecordid><originalsourceid>FETCH-LOGICAL-c237t-aba53bb95e5cc59415eeb0ee8fe5c36ae3128b85048f74338451f2a8b653b0fa3</originalsourceid><addsrcrecordid>eNp9kc-O1DAMhyMEYoeFF-CAcuTSIWmaNpW4oOWvtBIXOEdp6s541CZDnK7UV-FpyWgWjpwi2Z9_iv0x9lqKvRSyfXfan6IP-1rUzV5K0XfyCdtJreqqbrV6ynaib7tKm9bcsBdEJyFE3yjxnN2o0lBGmR37_RHgzGdwKWA4VI4IKbvggRMeAk7oXcjzxjH4BI6AeD4CHyGDzxgDJwiEGR8wbzxOPMCaIq3pAGnjCQhHCJn4FFNJyMn55AK6mbsLuNFCpcxpHVxpHUPEkR9hiSkd3QFesmeTmwlePb637OfnTz_uvlb33798u_twX_ladblyg9NqGHoN2nvdN1IDDALATKWgWgdK1mYwWjRm6hqlTKPlVDszlBsNYnLqlr295p5T_LUCZbsgeZjn8sm4klWyabuu1Z0oaH1FfdmSEkz2nHBxabNS2IsTe7IXJ_bixF6dlKE3j_nrsMD4b-SvhAK8vwJQtnxASJY8QnEwYipntmPE_-X_AecFozQ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3146776570</pqid></control><display><type>article</type><title>Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage</title><source>Elsevier ScienceDirect Journals</source><creator>Goertz, Lukas ; Jünger, Stephanie T. ; Reinecke, David ; von Spreckelsen, Niklas ; Shahzad, Rahil ; Thiele, Frank ; Laukamp, Kai Roman ; Timmer, Marco ; Gertz, Roman Johannes ; Gietzen, Carsten ; Kaya, Kenan ; Grunz, Jan-Peter ; Schlamann, Marc ; Kabbasch, Christoph ; Borggrefe, Jan ; Pennig, Lenhard</creator><creatorcontrib>Goertz, Lukas ; Jünger, Stephanie T. ; Reinecke, David ; von Spreckelsen, Niklas ; Shahzad, Rahil ; Thiele, Frank ; Laukamp, Kai Roman ; Timmer, Marco ; Gertz, Roman Johannes ; Gietzen, Carsten ; Kaya, Kenan ; Grunz, Jan-Peter ; Schlamann, Marc ; Kabbasch, Christoph ; Borggrefe, Jan ; Pennig, Lenhard</creatorcontrib><description>•This study assessed if deep learning models (DLMs) improve neurosurgery residents’ ability to detect aneurysms on CTA.•Residents’ detection sensitivities without DLM: 77.8% (1st year), 86.5% (3rd year), 87.3% (5th year).•With DLM: sensitivities rose to 97.6%, 95.2%, and 98.4%, averaging a 13.2% increase.•DLMs assist neurosurgeons in making treatment decisions when immediate radiology consults are not available.
The purpose of this study was to evaluate the effectiveness of a deep learning model (DLM) in improving the sensitivity of neurosurgery residents to detect intracranial aneurysms on CT angiography (CTA) in patients with aneurysmal subarachnoid hemorrhage (aSAH).
In this diagnostic accuracy study, a set of 104 CTA scans of aSAH patients containing a total of 126 aneurysms were presented to three blinded neurosurgery residents (a first-year, third-year, and fifth-year resident), who individually assessed them for aneurysms. After the initial reading, the residents were given the predictions of a dedicated DLM previously established for automated detection and segmentation of intracranial aneurysms. The detection sensitivities for aneurysms of the DLM and the residents with and without the assistance of the DLM were compared.
The DLM had a detection sensitivity of 85.7%, while the residents showed detection sensitivities of 77.8%, 86.5%, and 87.3% without DLM assistance. After being provided with the DLM’s results, the residents’ individual detection sensitivities increased to 97.6%, 95.2%, and 98.4%, respectively, yielding an average increase of 13.2%. The DLM was particularly useful in detecting small aneurysms. In addition, interrater agreement among residents increased from a Fleiss κ of 0.394 without DLM assistance to 0.703 with DLM assistance.
The results of this pilot study suggest that deep learning models can help neurosurgeons detect aneurysms on CTA and make appropriate treatment decisions when immediate radiological consultation is not possible.</description><identifier>ISSN: 0967-5868</identifier><identifier>ISSN: 1532-2653</identifier><identifier>EISSN: 1532-2653</identifier><identifier>DOI: 10.1016/j.jocn.2024.110971</identifier><identifier>PMID: 39673838</identifier><language>eng</language><publisher>Scotland: Elsevier Ltd</publisher><subject>Artificial intelligence ; Convolutional neural networks ; CT-angiography ; Neurosurgical training</subject><ispartof>Journal of clinical neuroscience, 2025-02, Vol.132, p.110971, Article 110971</ispartof><rights>2024</rights><rights>Copyright © 2024. Published by Elsevier Ltd.</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><cites>FETCH-LOGICAL-c237t-aba53bb95e5cc59415eeb0ee8fe5c36ae3128b85048f74338451f2a8b653b0fa3</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.sciencedirect.com/science/article/pii/S0967586824005101$$EHTML$$P50$$Gelsevier$$H</linktohtml><link.rule.ids>314,776,780,3537,27901,27902,65306</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/39673838$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Goertz, Lukas</creatorcontrib><creatorcontrib>Jünger, Stephanie T.</creatorcontrib><creatorcontrib>Reinecke, David</creatorcontrib><creatorcontrib>von Spreckelsen, Niklas</creatorcontrib><creatorcontrib>Shahzad, Rahil</creatorcontrib><creatorcontrib>Thiele, Frank</creatorcontrib><creatorcontrib>Laukamp, Kai Roman</creatorcontrib><creatorcontrib>Timmer, Marco</creatorcontrib><creatorcontrib>Gertz, Roman Johannes</creatorcontrib><creatorcontrib>Gietzen, Carsten</creatorcontrib><creatorcontrib>Kaya, Kenan</creatorcontrib><creatorcontrib>Grunz, Jan-Peter</creatorcontrib><creatorcontrib>Schlamann, Marc</creatorcontrib><creatorcontrib>Kabbasch, Christoph</creatorcontrib><creatorcontrib>Borggrefe, Jan</creatorcontrib><creatorcontrib>Pennig, Lenhard</creatorcontrib><title>Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage</title><title>Journal of clinical neuroscience</title><addtitle>J Clin Neurosci</addtitle><description>•This study assessed if deep learning models (DLMs) improve neurosurgery residents’ ability to detect aneurysms on CTA.•Residents’ detection sensitivities without DLM: 77.8% (1st year), 86.5% (3rd year), 87.3% (5th year).•With DLM: sensitivities rose to 97.6%, 95.2%, and 98.4%, averaging a 13.2% increase.•DLMs assist neurosurgeons in making treatment decisions when immediate radiology consults are not available.
The purpose of this study was to evaluate the effectiveness of a deep learning model (DLM) in improving the sensitivity of neurosurgery residents to detect intracranial aneurysms on CT angiography (CTA) in patients with aneurysmal subarachnoid hemorrhage (aSAH).
In this diagnostic accuracy study, a set of 104 CTA scans of aSAH patients containing a total of 126 aneurysms were presented to three blinded neurosurgery residents (a first-year, third-year, and fifth-year resident), who individually assessed them for aneurysms. After the initial reading, the residents were given the predictions of a dedicated DLM previously established for automated detection and segmentation of intracranial aneurysms. The detection sensitivities for aneurysms of the DLM and the residents with and without the assistance of the DLM were compared.
The DLM had a detection sensitivity of 85.7%, while the residents showed detection sensitivities of 77.8%, 86.5%, and 87.3% without DLM assistance. After being provided with the DLM’s results, the residents’ individual detection sensitivities increased to 97.6%, 95.2%, and 98.4%, respectively, yielding an average increase of 13.2%. The DLM was particularly useful in detecting small aneurysms. In addition, interrater agreement among residents increased from a Fleiss κ of 0.394 without DLM assistance to 0.703 with DLM assistance.
The results of this pilot study suggest that deep learning models can help neurosurgeons detect aneurysms on CTA and make appropriate treatment decisions when immediate radiological consultation is not possible.</description><subject>Artificial intelligence</subject><subject>Convolutional neural networks</subject><subject>CT-angiography</subject><subject>Neurosurgical training</subject><issn>0967-5868</issn><issn>1532-2653</issn><issn>1532-2653</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2025</creationdate><recordtype>article</recordtype><recordid>eNp9kc-O1DAMhyMEYoeFF-CAcuTSIWmaNpW4oOWvtBIXOEdp6s541CZDnK7UV-FpyWgWjpwi2Z9_iv0x9lqKvRSyfXfan6IP-1rUzV5K0XfyCdtJreqqbrV6ynaib7tKm9bcsBdEJyFE3yjxnN2o0lBGmR37_RHgzGdwKWA4VI4IKbvggRMeAk7oXcjzxjH4BI6AeD4CHyGDzxgDJwiEGR8wbzxOPMCaIq3pAGnjCQhHCJn4FFNJyMn55AK6mbsLuNFCpcxpHVxpHUPEkR9hiSkd3QFesmeTmwlePb637OfnTz_uvlb33798u_twX_ladblyg9NqGHoN2nvdN1IDDALATKWgWgdK1mYwWjRm6hqlTKPlVDszlBsNYnLqlr295p5T_LUCZbsgeZjn8sm4klWyabuu1Z0oaH1FfdmSEkz2nHBxabNS2IsTe7IXJ_bixF6dlKE3j_nrsMD4b-SvhAK8vwJQtnxASJY8QnEwYipntmPE_-X_AecFozQ</recordid><startdate>20250201</startdate><enddate>20250201</enddate><creator>Goertz, Lukas</creator><creator>Jünger, Stephanie T.</creator><creator>Reinecke, David</creator><creator>von Spreckelsen, Niklas</creator><creator>Shahzad, Rahil</creator><creator>Thiele, Frank</creator><creator>Laukamp, Kai Roman</creator><creator>Timmer, Marco</creator><creator>Gertz, Roman Johannes</creator><creator>Gietzen, Carsten</creator><creator>Kaya, Kenan</creator><creator>Grunz, Jan-Peter</creator><creator>Schlamann, Marc</creator><creator>Kabbasch, Christoph</creator><creator>Borggrefe, Jan</creator><creator>Pennig, Lenhard</creator><general>Elsevier Ltd</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope></search><sort><creationdate>20250201</creationdate><title>Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage</title><author>Goertz, Lukas ; Jünger, Stephanie T. ; Reinecke, David ; von Spreckelsen, Niklas ; Shahzad, Rahil ; Thiele, Frank ; Laukamp, Kai Roman ; Timmer, Marco ; Gertz, Roman Johannes ; Gietzen, Carsten ; Kaya, Kenan ; Grunz, Jan-Peter ; Schlamann, Marc ; Kabbasch, Christoph ; Borggrefe, Jan ; Pennig, Lenhard</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c237t-aba53bb95e5cc59415eeb0ee8fe5c36ae3128b85048f74338451f2a8b653b0fa3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2025</creationdate><topic>Artificial intelligence</topic><topic>Convolutional neural networks</topic><topic>CT-angiography</topic><topic>Neurosurgical training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Goertz, Lukas</creatorcontrib><creatorcontrib>Jünger, Stephanie T.</creatorcontrib><creatorcontrib>Reinecke, David</creatorcontrib><creatorcontrib>von Spreckelsen, Niklas</creatorcontrib><creatorcontrib>Shahzad, Rahil</creatorcontrib><creatorcontrib>Thiele, Frank</creatorcontrib><creatorcontrib>Laukamp, Kai Roman</creatorcontrib><creatorcontrib>Timmer, Marco</creatorcontrib><creatorcontrib>Gertz, Roman Johannes</creatorcontrib><creatorcontrib>Gietzen, Carsten</creatorcontrib><creatorcontrib>Kaya, Kenan</creatorcontrib><creatorcontrib>Grunz, Jan-Peter</creatorcontrib><creatorcontrib>Schlamann, Marc</creatorcontrib><creatorcontrib>Kabbasch, Christoph</creatorcontrib><creatorcontrib>Borggrefe, Jan</creatorcontrib><creatorcontrib>Pennig, Lenhard</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>Journal of clinical neuroscience</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Goertz, Lukas</au><au>Jünger, Stephanie T.</au><au>Reinecke, David</au><au>von Spreckelsen, Niklas</au><au>Shahzad, Rahil</au><au>Thiele, Frank</au><au>Laukamp, Kai Roman</au><au>Timmer, Marco</au><au>Gertz, Roman Johannes</au><au>Gietzen, Carsten</au><au>Kaya, Kenan</au><au>Grunz, Jan-Peter</au><au>Schlamann, Marc</au><au>Kabbasch, Christoph</au><au>Borggrefe, Jan</au><au>Pennig, Lenhard</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage</atitle><jtitle>Journal of clinical neuroscience</jtitle><addtitle>J Clin Neurosci</addtitle><date>2025-02-01</date><risdate>2025</risdate><volume>132</volume><spage>110971</spage><pages>110971-</pages><artnum>110971</artnum><issn>0967-5868</issn><issn>1532-2653</issn><eissn>1532-2653</eissn><abstract>•This study assessed if deep learning models (DLMs) improve neurosurgery residents’ ability to detect aneurysms on CTA.•Residents’ detection sensitivities without DLM: 77.8% (1st year), 86.5% (3rd year), 87.3% (5th year).•With DLM: sensitivities rose to 97.6%, 95.2%, and 98.4%, averaging a 13.2% increase.•DLMs assist neurosurgeons in making treatment decisions when immediate radiology consults are not available.
The purpose of this study was to evaluate the effectiveness of a deep learning model (DLM) in improving the sensitivity of neurosurgery residents to detect intracranial aneurysms on CT angiography (CTA) in patients with aneurysmal subarachnoid hemorrhage (aSAH).
In this diagnostic accuracy study, a set of 104 CTA scans of aSAH patients containing a total of 126 aneurysms were presented to three blinded neurosurgery residents (a first-year, third-year, and fifth-year resident), who individually assessed them for aneurysms. After the initial reading, the residents were given the predictions of a dedicated DLM previously established for automated detection and segmentation of intracranial aneurysms. The detection sensitivities for aneurysms of the DLM and the residents with and without the assistance of the DLM were compared.
The DLM had a detection sensitivity of 85.7%, while the residents showed detection sensitivities of 77.8%, 86.5%, and 87.3% without DLM assistance. After being provided with the DLM’s results, the residents’ individual detection sensitivities increased to 97.6%, 95.2%, and 98.4%, respectively, yielding an average increase of 13.2%. The DLM was particularly useful in detecting small aneurysms. In addition, interrater agreement among residents increased from a Fleiss κ of 0.394 without DLM assistance to 0.703 with DLM assistance.
The results of this pilot study suggest that deep learning models can help neurosurgeons detect aneurysms on CTA and make appropriate treatment decisions when immediate radiological consultation is not possible.</abstract><cop>Scotland</cop><pub>Elsevier Ltd</pub><pmid>39673838</pmid><doi>10.1016/j.jocn.2024.110971</doi></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0967-5868 |
| ispartof | Journal of clinical neuroscience, 2025-02, Vol.132, p.110971, Article 110971 |
| issn | 0967-5868 1532-2653 1532-2653 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_3146776570 |
| source | Elsevier ScienceDirect Journals |
| subjects | Artificial intelligence Convolutional neural networks CT-angiography Neurosurgical training |
| title | Deep learning-assistance significantly increases the detection sensitivity of neurosurgery residents for intracranial aneurysms in subarachnoid hemorrhage |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T18%3A08%3A09IST&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=Deep%20learning-assistance%20significantly%20increases%20the%20detection%20sensitivity%20of%20neurosurgery%20residents%20for%20intracranial%20aneurysms%20in%20subarachnoid%20hemorrhage&rft.jtitle=Journal%20of%20clinical%20neuroscience&rft.au=Goertz,%20Lukas&rft.date=2025-02-01&rft.volume=132&rft.spage=110971&rft.pages=110971-&rft.artnum=110971&rft.issn=0967-5868&rft.eissn=1532-2653&rft_id=info:doi/10.1016/j.jocn.2024.110971&rft_dat=%3Cproquest_cross%3E3146776570%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=3146776570&rft_id=info:pmid/39673838&rft_els_id=S0967586824005101&rfr_iscdi=true |