I am trying to match the bench mark values

[Kundan Singh Chufal]

I am trying to match the bench mark values for IBSI CT phantom.

Many values are not matching and while some are beyond the tolerance. i have attached my parameter json file. Please advise if i need any modification.

[ap...@mskcc.org]

Here is sample settings file for IBSI phase 2 phantom.

--
You received this message because you are subscribed to the Google Groups "CERR FORUM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cerr-forum+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cerr-forum/391a31de-80b7-4c2f-ac9f-7b785d1b495eo%40googlegroups.com.

*** Only open attachments or links from trusted senders. Report phishing to inf...@mskcc.org ***

 

=====================================================================

Please note that this e-mail and any files transmitted from
Memorial Sloan Kettering Cancer Center may be privileged, confidential,
and protected from disclosure under applicable law. If the reader of
this message is not the intended recipient, or an employee or agent
responsible for delivering this message to the intended recipient,
you are hereby notified that any reading, dissemination, distribution,
copying, or other use of this communication or any of its attachments
is strictly prohibited. If you have received this communication in
error, please notify the sender immediately by replying to this message
and deleting this message, any attachments, and all copies and backups
from your computer.

[kundan]

Hi Aditya

I have done benchmarking with IBSI git repo for digital CT phantom as per your settings. I am attaching settings file as well as benchmarked excel sheet. If it seems ok then i will proceed to extract features from my CT dataset.

Thanks and regards 

To view this discussion on the web visit https://groups.google.com/d/msgid/cerr-forum/4c1d64b4d4884e2088dac98bf60a1484%40mskcc.org.

--

Dr Kundan Singh Chufal

Sr Consultant Radiation Oncology

Chief Breast and Thorax Radiation Services

Rajiv Gandhi Cancer Institute and Research Centre

Sector-5, Rohini 110085, Delhi, India 

[Aditya Apte]

Thanks Kundan for the comparison file.

In general, some features are quite sensitive to the selection of resampling grid. For example, selection of slice z-values based on the original scan grid vs the cropped scan might result in different slice positions. The IBSI benchmark resampled from the cropped ROI, whereas CERR resampled from the original scan. Hence, some metrics (min/max or those that are influenced by outliers) are beyond or borderline tolerance. 

We are in the process of pushing an update to allow padding scan cropped around the structure mask. This will help with (1) application of filters such as LoG and (2) specification of zero padding for resampling so that the features are closer to IBSI benchmark values (although such sensitive features might not be useful in modeling).

To view this discussion on the web visit https://groups.google.com/d/msgid/cerr-forum/CAD5R6Dp5pQRhLfrF89ShN1xHfKnopTfyK1pWm0amyWBPWCHdGg%40mail.gmail.com.

<final_cerr_benchmark.xlsx>

<paramsForCTfeatures1_configC.json>

<paramsForCTfeatures2_config_C.json>

[Kundan Singh Chufal]

Thanks Aditya 

Just need one clarification.

For wavelet filter what is the difference between different types. 

For example coif1 coif2  and so on

Should I extract for all subtypes 

Also for direction like HHH, HLH and so on there are 9 combination. And there is one option of ‘all’. 

So should I extract using all option and it will incorporate all direction or I have to extract for each direction 

Currently I am extracting for all filters with all possible Filter combination which is giving an output with more than 10000 features.

Just need your input and comment for the same.

Sent from my iPhone

On 17-Jun-2020, at 7:24 AM, Aditya Apte <adity...@gmail.com> wrote:

Thanks Kundan for the comparison file.

To view this discussion on the web visit https://groups.google.com/d/msgid/cerr-forum/6EF90037-12AD-4972-9FC2-BF5941FEB56A%40gmail.com.