For mask

[AlexNmSED]

I noticed a difference in the tumor masks between the Duke dataset and the I-SPY2 and I-SPY1 datasets. I am having trouble understanding the masking mechanism used in I-SPY2 and I-SPY1. Can the masks they provide be considered as tumor masks? However, I observed that the masks provided by I-SPY2 and I-SPY1 are quite scattered. Could someone clarify the nature of these masks and how they should be interpreted?

Thank you for your help.

[naomifridman]

Hi, The masks in spy1 and spy2 are masks of the primary tumor. Sometimes they are scattered. It's not a totally manual segmentation, and I attach the explanation of the segmentation pipeline. The Duke dataset contains only the bounding box of the primary tumor. Best, Naomi

…

On Sat, Aug 30, 2025 at 11:27 AM AlexNmSED ***@***.***> wrote: *AlexNmSED* created an issue (naomifridman/BreastDCEDL#3) <#3> I noticed a difference in the tumor masks between the Duke dataset and the I-SPY2 and I-SPY1 datasets. I am having trouble understanding the masking mechanism used in I-SPY2 and I-SPY1. Can the masks they provide be considered as tumor masks? However, I observed that the masks provided by I-SPY2 and I-SPY1 are quite scattered. Could someone clarify the nature of these masks and how they should be interpreted? Thank you for your help. — Reply to this email directly, view it on GitHub <#3>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7M4XJLCM4C7FUJILWFHZL3QFOA7AVCNFSM6AAAAACFGMOOO2VHI2DSMVQWIX3LMV43ASLTON2WKOZTGM3DQOJQG42TGNI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[AlexNmSED]

Hi Naomi,

Thank you for your quick response and the helpful explanation!
I have a couple of follow-up questions.

Does this mean that I can generate the bounding box of the primary tumor for the I-SPY1 and I-SPY2 datasets based on the provided masks? Also, could you let me know when the I-SPY2 dataset on TCIA is expected to be updated? It would be great if you could notify me once the update is available.

Best regards,
Xinyu
PhD candidate, University of Jyväskylä, Finland

[naomifridman]

Hi Xinyu, "Does this mean that I can generate the bounding box of the primary tumor for the I-SPY1 and I-SPY2 datasets based on the provided masks?" Yes you can. I believe the spy2 will be uploaded to TCIA, in 2-3 weeks. I will be happy to notify you. Best, N

…

On Sat, Aug 30, 2025 at 3:32 PM AlexNmSED ***@***.***> wrote: *AlexNmSED* left a comment (naomifridman/BreastDCEDL#3) <#3 (comment)> Hi Naomi, Thank you for your quick response and the helpful explanation! I have a couple of follow-up questions. Does this mean that I can generate the bounding box of the primary tumor for the I-SPY1 and I-SPY2 datasets based on the provided masks? Also, could you let me know when the I-SPY2 dataset on TCIA is expected to be updated? It would be great if you could notify me once the update is available. Best regards, Xinyu PhD candidate, University of Jyväskylä, Finland — Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7M4XKYZCW3V7EGLMFIOCD3QGKWRAVCNFSM6AAAAACFGMOOO2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTEMZZGIZTKNJZGI> . You are receiving this because you commented.Message ID: ***@***.***>

[AlexNmSED]

Hi Naomi,

Thank you for your help.

Best regards,
Xinyu
PhD candidate, University of Jyväskylä, Finland

[AlexNmSED]

Hi Naomi,

Whether the spy2 has been uploaded to TCIA? Thank you for your help.

Best regards,
Xinyu
PhD candidate, University of Jyväskylä, Finland

[naomifridman]

Hi Xinyu,
The TCIA upload is nearly ready and should be available within the next few days. In the meantime, I’ve uploaded a minimal version to Zenodo. It includes three time points (pre-injection early post, and late post), cropped to 256×256 around the primary tumor, with two slices padding around the tumor region.

[naomifridman]

https://zenodo.org/records/17274053

[AlexNmSED]

Hi Naomi,

Thank you for your help. This is a nice work.

[AlexNmSED]

Hello Naomi,

Sorry to bother you again. Would it be possible for you to share the entire preprocessing code for spy2? Our specific research focus is on temporal issues, and I need data from different time points of spy2. I would like to follow your processing steps to obtain the remaining data from spy2.

Thank you again.

Best regards,
Xinyu
PhD candidate, University of Jyväskylä, Finland

[naomifridman]

What code ? To create nitro from Dicom ?

…

Sent from my iPhone

On Thu, 23 Oct 2025 at 5:27 AlexNmSED ***@***.***> wrote: *AlexNmSED* left a comment (naomifridman/BreastDCEDL#3) <#3 (comment)> Hello Naomi, Sorry to bother you again. Would it be possible for you to share the entire preprocessing code for spy2? Our specific research focus is on temporal issues, and I need data from different time points of spy2. I would like to follow your processing steps to obtain the remaining data from spy2. Thank you again. Best regards, Xinyu PhD candidate, University of Jyväskylä, Finland — Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7M4XIS4DE63MEZXQEVWVL3ZA4KXAVCNFSM6AAAAACFGMOOO2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIMZUHA2DGNZZGA> . You are receiving this because you commented.Message ID: ***@***.***>

[naomifridman]

The full preprocessing code for SPY2 is a tangled mess of if-else blocks, case-specific functions, and manual lookups. It's deeply fragmented and organizing it properly would be an enormous effort.

[AlexNmSED]

Yes, processing SPY2 is indeed quite a challenge. You’ve made a great contribution to the community with your efforts!

[naomifridman]

I-SPY2 (Full): [Download from breastdcedl-ispy2 TCIA][(https://www.cancerimagingarchive.net/analysis-result/breastdcedl-ispy2/)]