Correction: On the construction of a large-scale database of AI-assisted annotating lung ventilation-perfusion scintigraphy for pulmonary embolism (VQ4PEDB)

Amir Jabbarpour¹Eric Moulton²Sanaz Kaviani³Siraj Ghassel²Wanzhen Zeng⁴Ramin Akbarian⁴Anne Couture⁵Aubert Roy⁶Richard Liu⁷Yousif Lucinian⁵Nuha Hejji⁴Sukainah Alsulaiman⁴Farnaz Shirazi⁴Eugene Leung⁴Sierra Bonsall⁸Samir Arfin⁸Bruce Gray⁸Ran Klein^1*

• ¹The Ottawa Hospital, Ottawa, Canada
• ²University of Ottawa School of Electrical Engineering and Computer Science, Ottawa, Canada
• ³Jubilant DraxImage Inc, Kirkland, Canada
• ⁴Nuclear Medicine and Molecular Imaging, Faculty of Medicine, University of Ottawa, Ottawa, Canada
• ⁵Hopital Maisonneuve-Rosemont, Montreal, Canada
• ⁶McGill University Faculty of Medicine and Health Sciences, Montreal, Canada
• ⁷Department of Nuclear Medicine, Jewish General Hospital, Montreal, Canada
• ⁸Department of Medical Imaging, St. Michael’s Hospital, Unity Health Toronto, Toronto, Canada

A correction refers to a change to their article that the author wishes to publish after publication. The publication of this article is subject to Frontiers' editorial approval.Instructions:• Please read through all the templates before choosing • Pick the most relevant text template(s) from the following page and delete all others.• Edit the text as necessary, ensuring that the original incorrect text is included for the record, please see the below. • Please do not use any extra formatting when editing the templates, and only modify the red text unless absolutely necessary • Submit to Frontiers following the instructions on this page.When the original text contained incorrect information, to preserve the scientific record, please include that text when editing the below templates. For example:There was a mistake in the Funding statement, an incorrect number was used.The correct number is "2015C03Bd051.". The publisher apologizes for this mistake.The original version of this article has been updated. We have incorrectly opined about the performance of a private company's third party software (Segmed Inc). The performance related to Canadian zip codes as well as other types of data was due to a user error and improper configuration of the tool by the authors. We have made the following edits:[Methods, Anonymization of DICOM files and de-identification of clinical reports, second paragraph]. This paragraph should be replaced with:"For clinical report texts, we adopted and compounded the effect of the following three independent approaches as a conservative de-identification strategy: (1) Segmed Inc.'s Python-based web server was used to remove PII/PHI from clinical reports, (2) RegEx rules were used to remove Canadian formatted addresses and postal codes in Python, and (3) resulting texts were fed to a Microsoft Copilot agent that was instructed to list suspected people names, addresses, street names, 5-8 digit numbers, business names, clinic names and occupations. The agent was further prompted to ignore medical terms. The resulting terms were then manually screened for relevance, the terms were searched for in the text and then replaced with "[Anon]"."[Discussion, Data Ingestion, first paragraph]: This paragraph should be replaced with:"QA revealed our unstructured text data to be properly de-identified, highlighting the effectiveness of our multiple layers of de-identification approaches. Structured DICOM data from hospital sources proved straightforward to robustly de-identify using our strategy. Also, during our QA process, various non-structured DICOM tags, such as series description, used during splitting process were identified and addressed accordingly to preserve integrity of workflow." Accordingly, please remove reference 27. for a reason not seen here, please contact the journal's editorial office.