Clinical/Longitudinal Layers of Data + Labels for Healthy Cohort?
Hi! First off wanted to say congrats on the release!
I was wondering if clinical/longitudinal data will be made available and/or labels defining which patients are verified healthy?
Also if any other original data posted on the timeline will be released. Thank you!
Hi @k3alvarez ,
Thanks for the kind words and for the question!
Longitudinal data: yes, it is in the release. Every scan in the metadata CSV carries a de-identified patient ID, accession number, and exam date, so you can reconstruct each patient's timeline across scans. Radiology reports, pathology results (ICD-10 codes), clinical variables, and the tumor annotations are all released as well.
Verified healthy: yes, it is in the release as well. Since the tumor annotations are released, the verified healthy patients are the ones with no tumor annotation.
Could you let me know what you mean by "other original data"? To my knowledge, we have released everything promised in the paper, plus extra CT scans and patients as a bonus. But I would love to hear the wishlist from researchers like you. If there is something that would make CancerVerse more useful for your work, tell me and I will see what I can do!
Thanks,
Wenxuan
I guess maybe just a clarification if this information will be made available in a more structured way, had the older version pulled up, but then refreshed midway, but I guess this was a snippet "NM stage, treatments received, surgical resections, RECIST treatment response, metastatic sites, disease-status narratives, follow-up recommendations, comorbidities, and devices/implants."
Hi @k3alvarez ,
That information is already available inside the free-text radiology reports in the newest version. If you would like a cleaned-up structured version, please contact zzhou82@jh.edu for more details.
Thanks,
Wenxuan
Clarification on Date De-identification in CancerVerse Reports
Hi @BodyMaps team,
As mentioned, the data has been de-identified, including the Patient IDs, accession numbers, and exam dates in the metadata CSV file.
I wanted to clarify how dates were handled, particularly the dates appearing within the reports text column of the metadata CSV, such as dates of previous scans or procedures.
- Are the dates mentioned in the report text the original dates, or were they shifted as part of the de-identification process?
Any details or documentation on the de-identification process would be really helpful. Thanks!
Hi @shravaniiiiii ,
It looks like you also posted this as a separate discussion. I will reply there, so please follow that thread for my answer.
Thanks,
Wenxuan