OHDSI Home | Forums | Wiki | Github

Oncology radiology imaging integration into CDM

We are Medical Convergence Research Center of Wonkwang University Hospital in Korea(prof. Kwon-Ha Yoon).
We has been working on building medical image big data, and proposing a new table for radiology based on OMOP CDM.
It is a table related to medical imaging. We named this as Radiology_CDM and documented the definition of the attributes of each table.
The attributes of Radiology_CDM are defined to be obtained from Tag information of DICOM which is medical image standard.
We prepared for Radiology _CDM ver.1.0 supported by Ajou university hospital research team(prof. Rae Woong Park)

We would like to make a working group and discuss it together.

1 Like

@mediblue

Great work!
I think integration of radiology data into CDM would be one of the the most important works in CDM. We’ll join this working group, definitely!

@mkwong, i remember you were interested and already working on mapping ECG data?

@mediblue
We are Medical Convergence Research Center of Wonkwang University Hospital in Korea(prof. Kwon-Ha Yoon).
This file is our proposed Radiology_CDM (v1.0).
Radiology_CDM.xlsx (18.1 KB)

We would like to make a working group and discuss it together.

Suggested tables looks great. An optional reference to a radiology report in the NOTE table would be nice. Maybe add RADIOLOGY_NOTE_ID to RADIOLOGY_OCCURRENCE.

Also, yes, I know about FACT_RELATIONSHIP.

It’s very good point, @mgurley!
I think the note table should be revised to capture the information where it comes from
(current note table doesn’t store the pathology or radiology reports), as discussed here.
I made a suggestion and posted it to the github.

Please continue the EKG conversation in this thread.

Two things, @SiHyung_No: Could you be so kind and put this proposal into the OMOP CDM Github? We also need use cases. What kind of questions are we trying to answer, and how do all these fields support them.

@Christian_Reich

One of use cases would be developing machine learning algorithm analyzing X-ray, CT or MRI using radiology extension model.
I started to take the Neurohacking course in Coursera on every Saturday.
I plan to convert the sample data of this course into radiology CDM and apply the analysis on CDM, which I learn in the course.

Hi,

can you please clarify:
are we talking about CDM modification allowing us to put there images and reports
or
are we talking about the

@Dymshyts
I’m not sure that I understand correctly,
But what @SiHyung_No suggested is that storing path for the images and its informations, not image itself.
Image reports can be stored in the Note table after slight modification.

I’m sorry for the incomplete message posting, but you answered exactly what I wanted to know:)

Thanks

@yoon8302 had finished extracting waveform signals from ECG reports in PDF form . The new DB named ECG-ViEW III consists of 1 M of ECG reports and all the clinical information (demographic, condition, drug, selected measurement) in OMOP-CDM format. He will submit a manuscript on it soon. The DB will be opened to any researcher after acceptance of the manuscript. He also made a prototype of CDM extension model to contain the bio-signal waveform data.

Sorry for this TOO late reply. Extension models for radiology and bio-signals will be different each other. @yoon8302 will try to make a prototype for bio-signal CDM extension model, and he may want to submit it to our community.

Owing to the effort of @SiHyung_No , @mediblue , @NEONKID, the first sample data for radiology extension model is released.
This sample is consists of two tables

  1. Radiology_Occurrence: each row represents one radiologic procedure
  2. Radiology_Image: each row represents one image among the images from single radiologic procedure.

Samples:
Radiology_Ocurrence
Radiology_Image

2 Likes

Friends:

As you model this out don’t forget we need to have normalized information as concepts in the concept table. If Dicom has it all, bring it on. If not, pick from SNOMED or so. If nothing helps propose a new controlled vocabulary.

@Christian_Reich

We’re now figuring out which concepts we need to standardize. I’ve already picked RadLex ontology system from LOINC within OMOP concepts for standardized radiology procedure.
Definitely, we need to normalize all this information after completing the pilot model. Thank you for reminding again! :smile:

@SCYou Thanks for leading this work. How do you do so much?

It would provide the basis of a lot of important research if the concepts required for DICOM CT Radiation Dose Structured Reports (RDSRs) were among the concepts you standardized. Here in the US, an estimated 3-5% of incident cancers are attributable to medical radiation and there is ample evidence that CT dosing is frequently too high relative to benchmarks and that there is very high unwarranted variability in CT dosing. Capturing the required data elements for RDSRs would allow the OHDSI community to document and address these important aspects of CT use. Having contributed to some of this research I am interested in being able to continue it with OHDSI partners…

These mature open source toolkits for getting data from DICOM (which uses RadLex) are designed to pull the elements for RDSRs among other useful tasks. PixelMed toolkit: http://www.pixelmed.com/dicomtoolkit.html and GROK: http://dose-grok.sourceforge.net/

1 Like

Thank you for providing such a precious information, @Andrew

Columns for radiation dosage and the unit exist without details in the pilot radiology extension model above. In order to standardize dosage and the units according to the various modality further, RDSR will be very helpful. Thank you again.

1 Like

@SCYou

I am curious to find out the status of this, is there any more recent work on Radiology_CDM? I am interested trying to deploy your current model and code on some test image data at my site, and possibly make extensions to it to fill gaps for our use cases for enabling image dataset discovery for multi-site collaboration in a federated learning setting without sharing data.

1 Like
t