BMMR2: Breast Multiparametric MRI for prediction of NAC Response Forum

Go back to competition Back to thread list Post in this thread

> How to get timings for first and second post-contrast scans?

Hi,

I'm trying to extract the timing since injection for the first and second post-contrast scans for the DCE images. Based on Table 4 on page 9 of "ACRIN 6698 ISPY2 DWI and DCE MRI Data Descriptions_20210520.pdf", I thought this would be the second and third elements of the array in DICOM Tag (0117,1032). But when I looked at the those fields, I found that while they differ per subject, the T0/T1/T2 values are nearly/exactly the same for a single subject.

Here's an examples of a few fields:

subj,time_pt,timing1,timing2,study_date
ACRIN-6698-102212,T0,42.5295600891,129.78956604,20020404
ACRIN-6698-102212,T1,42.5295600891,129.6305541992,20020502
ACRIN-6698-102212,T2,42.5295600891,129.8195648193,20020706
ACRIN-6698-103939,T0,45.8485832214,139.580581665,20030319
ACRIN-6698-103939,T1,45.7175827026,139.1785888672,20030425
ACRIN-6698-103939,T2,48.2215042114,146.7155151367,20030717
ACRIN-6698-104268,T2,42.5,127.375,20020112
ACRIN-6698-104268,T0,42.4375,127.3125,20010915
ACRIN-6698-104268,T1,42.5,127.3775024414,20011111
ACRIN-6698-107700,T0,42.5295600891,129.692565918,20010315
ACRIN-6698-107700,T1,42.5295600891,129.510559082,20010419
ACRIN-6698-107700,T2,42.5295600891,129.4585571289,20010707
ACRIN-6698-116603,T0,42.5295600891,129.5365600586,20020328
ACRIN-6698-116603,T1,42.5295600891,129.5995635986,20020425
ACRIN-6698-116603,T2,42.5295600891,129.7035675049,20020627

It seems unlikely that the timings would be so close when the scans were taken weeks or months apart. Am I looking at the wrong fields? Or am I misinterpreting how to use this field?

Thanks.

Walt

Posted by: wmankowski @ Nov. 1, 2021, 6:53 p.m.

Hi Walt:
Sorry for the delay.
I think you meant fields (0117,1034) "EffectiveAcquisitionDelay"? (0117,1032) is defined in our private group, but I don't think it is included in these files.

Anyways, I believe the given timings are correct. Or at least as correct as we could manage given the highly variable ways the manufacturers store timing information.
I think the consistency is related to how the sites programmed in the protocols in the scanners. In some cases the timing was hard-coded perhaps. For example I checked the 1st patient you sent. It was done on a GE scanner and the scan duration (private attribute we use for timing calculations) is identical at 85.0591 seconds at all four study visits.

David

Posted by: Newitt @ Nov. 16, 2021, 2:16 a.m.

Thanks for the explanation. Yes, I did mean (0117,1034). That's what I used in my code, but I was looking at the documentation instead of my code when I made the post.

Walt

Posted by: wmankowski @ Nov. 16, 2021, 3:26 p.m.
Post in this thread