...

• Requirement statements: https://wiki.nci.nih.gov/x/KDxyAQ
• Use cases:https://wiki.nci.nih.gov/display/seminfra/Init1SD60-Metadata+for+statistical+practice

caGRID should support interoperability from non grid platforms.

Domain description:  A cancer researcher who is not familiar with the grid wants to collaborate with his peers from the cancer research community to identify tissue specimens, microarray, clinical trials and images of his interest. Being a total stranger to the grid he does not know the data standards and the models he needs to invoke to support his search.

• Patrick: This is a very interesting one that is different from the rest.  However, I am not sure the last sentence in the domain description adequately captures it.  To me, the issue is how a statistician finds the appropriate standards and then integrates them into their own statistical computing environment.  All of the other use cases we have focus on the metadata specialist or cancer researcher - this one focuses on the statistician who is performing the analysis and generating the data.  We need to drive home his issues and how they are to be resolved.  Also, I am not sure I understand the table of contents of objects.caGRID should support interoperability from non grid platforms.

Domain description:  A cancer researcher who is not familiar with the grid wants to collaborate with his peers from the cancer research community to identify tissue specimens, microarray, clinical trials and images of his interest. Being a total stranger to the grid he does not know the data standards and the models he needs to invoke to support his search.

Technical description: One persons object is Technical description: One persons object is another person's attribute". Depending on one's world view, a real life entity can be modeled in UML as Object, Attribute or value set. In caBIG models now, some model Race (and Ethnicity) as an Object while others model Race (and Ethnicity) as an attribute. This is problematic, because Race data that is modeled differently cannot be "seamlessly integrated" on caGrid (there needs to be a transform). One can start to use the SAIF language in terms of Conceptual, Platform Independent (logical) and Platform Specific (Implemented).  Given that the grid has CIMs, PIMs and PSMs for applications, and BAM and a DAM and other institutions, may have their own DAMs, BAMs, CIMs, PIMs and PSMs.  These elements need to be mapped to each other at whatever level needed, to get to some semantic interoperability.

...

• Requirement statement: https://wiki.nci.nih.gov/x/hAVyAQ
• Forum post: https://cabig-kc.nci.nih.gov/Vocab/forums/viewtopic.php?f=39&t=167
• Usecases: https://wiki.nci.nih.gov/display/seminfra/Init6SD12SD12UML+modelling+in+different+layer://wiki.nci.nih.gov/display/seminfra/Init6SD12SD12UML+modelling+in+different+layerPatrick: This one is very specific to a technology (caGrid).  It needs to be generalized.  I am not sure how the title matches the descriptions (since non-grid platforms are not mentioned).  Also, I am not sure how the domain description matches the technical description.  The technical description seems to focus on semantic relationships and transforms, whereas the domain description seems to focus on accessibility to non-technical users.Semantic search on the cancer grid.

Domain description:  A cancer researcher is looking for lung cancer specimens with a histologic picture of a 'oats cell carcinoma' in males aged between 45-55 years and who have a history of smoking for at least 10 years. He invokes a data service and queries caTissue instances in DFCI, TJU, LLU for specimens and rather than using a advance query, uses a semantic query like show me all specimens of lung cancer in  males aged 45-55 years who are smokers for at least 10 years and whose histologic  picture is that of a oat cell carcinoma.  The query runs on various instances of caTissue and comes back with the identified specimens that matched the criteria. Being able to facilitate a semantic search on the grid would facilitate greater cohesiveness of the research cancer research community.

...

• Use case https://wiki.nci.nih.gov/display/seminfra/Init4hm1.SD210-Triple+store+backend+for+LexEVS
• Requirement statement: https://wiki.nci.nih.gov/x/3AJyAQ
• Forum post: https://cabig-kc.nci.nih.gov/Vocab/forums/viewtopic.php?f=37&t=127Patrick: I think this one is very similar to some of the search use cases earlier.  Also, the technical description seems to focus on terminologies, whereas they are not really mentioned in the domain description.
  Integration of radiology, pathology, molecular and genomic data to better predict patient outcome and support clinical decision.

Domain Description: A patient reports to a hospital with a clinical condition of Glioblastoma multiforme. The treating oncologist wants to find out the likely outcome for this patient. He initiates a search based on patient presenting criteria in Imaging, histopathology and genomic data to look for cohort with matching criteria and survival rate to better predict outcome for his patients.

...

• Requirement statement: https://wiki.nci.nih.gov/x/HpN-AQ
• Forum post: https://wiki.nci.nih.gov/display/Imaging/TCGA+Enterprise+Use+CasePatrick: I am not sure which requirements in this are not captured in the other use cases.

  B. Forms Stories

Create and reuse forms

Domain Description: Forms provide a convenient paper-like electronic mechanism to capture data in a structured way.  For example, when a patient is placed on a clinical trial, data about the patient's demographics and eligibility for the trial need to be captured.  The trial investigator sits with the forms curator to generate this case report form.  The forms curator searches for existing demographics forms and form modules, and the investigator reviews them.  They identify an appropriate set of questions, and include them in the case report form.  They then move onto the eligibility checklist.  The investigator drafted the checklist, and it has been approved by the IRB.  The forms curator begins keying in the questions, some of which are identified as existing questions and reused, others of which are created completely new.  The form is marked complete and is available by the clinical research staff for gathering and enrolling new patients.

...

Given the numerous usability issues with the CDE browser the need is to come up with an alternative and a more efficient search workflow.

Cross reference

• New CDE browser workflow:

...

• https://wiki.nci.nih.gov/pages/viewpageattachments.action?pageId=24259415
• Requirement statement: https://wiki.nci.nih.gov/x/agRyAQ
• Forum post: https://cabig-kc.nci.nih.gov/Vocab/forums/viewtopic.php?f=43&t=109
  

D. Developer Stories

Iterative development and management of information models

...