Create matrix of project deliverables and the software packages/repos that support those deliverables

[jshoughtaling]

In an effort to eliminate ambiguity regarding development efforts and the deliverables they support, we will catalog and align deliverables with the packages and display them on this site.

[jshoughtaling]

@del42 - I created a collapsible table structure below to capture items per team and per major deliverable.

Let me know if you think this is a reasonable way to represent the data and I'll do it for the other teams and then publish it on the developer site.

Data Acquisition

Expand Aims

1. Site startup

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
1.1) Project Readiness
1.2) Ensure IRB readiness

2. Cohort Sampling and Size Justification

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
2.1) Identify intellectual property concerns at each site
2.2) Ensure diversity in site data
2.3) Obtain population- representative inferences	JARED	container-apps [atlas]

3. Federation process to access data from all patients

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
3.1) Ensure federated process to identify cohorts and request data	JARED	container-apps [atlas]
3.2) Document federated processes, with tooling module
3.3) Define and prepare meta dataset	JARED	container-apps [atlas]

4. Structured Data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
4.1) Prepare OMOP Data	JARED	container-apps [etl]	Central Processing SOP
4.1) Prepare to query OMOP data	JARED	container-apps [etl]

5. Obtaining High-resolution physiological data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
5.1) Integrate OMOP vocabulary for physiological data	JARED	chorus-mapping [vocab]	SOP
5.2) Standardize signal processing in waveforms	BRIAN	chorus_waveform
5.3) Quality control

6. Collection and processing of image data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
6.1) Ensure collection of image data at all sites
6.2) Ensure generation of metadata from images
6.3) Ensure collection of image data at all sites
6.4) Ensure linkage to other data domain	JARED	container-apps [registry]	MM SOP

7. Collection and Processing of Clinical notes

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
7.1) Customize NLP algorithm
7.2) Validate the NLP tool using honest brokers
7.3) Unstructured EHR extraction
7.4) Implement NLP tool

8. Mining for socioeconomic status data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
8.1) Identify common data SDoH elements from collective experience of sites
8.2) Identify and resolve discrepancies in performance of SDoH elements

9. Collection of contextual SDoH data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
9.1) Review SDoH variables and data sources
9.2) Prepare geospatial crosswalk datasets

10. Linkage among EHR

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
10.0)Design and communicate linkage SOP	JARED		MM SOP
10.1) Retrieve EHR data through MRNs at each site	JARED	container-apps [registry]	MM SOP
10.2) Minimal EHR dataset elements created and communicated to all sites
10.3) EHR data extracted	JARED
10.4) Site-specific EHR extracts validated
10.5) Implement tools to accurately link EHR and physiologic data	JARED	container-apps [registry]
10.6) Gap analysis of linkage performed	JARED

11. Deidentification of data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
11.1) Prepare safe harboring approach
11.2) Apply SOP on safe harboring approach
11.3) Quality control of linked safe-harbor datasets

12. Event Annotation

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
12.1) Develop semi-automatic predictive monitoring model to identify events
12.2) ndividual chart review of potential events by clinicians
12.3) Develop phenotype algorithm in OHDSI Phenotype repository	JARED	container-apps [atlas]
12.4) Implement phenotype algorithm in OHDSI Phenotype repository	JARED	container-apps [atlas]
12.5) Generate silver-standard labels at scale automatically	JARED
12.6) Implement annotation pipeline to combine physiological data for resolved timeline
12.7) Store results both locally and centrally
12.8) Create datasheet describing each cohort	JARED	container-apps [atlas], CHoRUSReports

13. Site-specific metadata

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
13.1) Encode de-identified hospital number on each data set
13.2) Report site-specific metadata from full patient dataset

14. Quality control

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
14.1) Write protocols for quality controls checks at all sites	JARED	CHoRUSReports	Quality Central
14.2) Quantify variability when it exists	JARED	CHoRUSReports	Quality Central
14.3) Use the Data Quality Dashboard from OHSDI to evaluate datasets	JARED	container-apps [ares, www-dgs]	Quality Central

15. Privacy Check

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
15.1) Examine unique identifiability with combinations of attributes to check potential risk of linkability
15.2) Evaluate I-diversity and t-closeness
15.3) For unstructured data, extract concepts through whitelist mechanism and test for linkability

16. Generate synthetic data

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
16.1) Use generative adversarial network (GAN) approaches to produce site-specific synthetic data sets
16.2) Post the data and the source code

17. Hold-out validation dataset

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
17.1) Set aside 20% of patients for algorithm testing
17.2) Pursue qualification of reserved dataset for FDA's Medical Device Development Tool program

18. Data storage platform - CHoRUS

SUBAIM	INVOLVED	SOFTWARE	DOCUMENTATION
18.1) Organize files by year/month, then Subject ID, data modalities	JARED	container-apps [etl]	Central Processing SOP
18.2) Store dataset in in site-specific staging area	JARED	container-apps [etl]	Central Processing SOP
18.3) Implement automated data integrity check; review error logs	JARED	CHoRUSReports	Quality Central
18.4) Copy files to data lake using secure FTP	JARED	container-apps [etl]	Central Processing SOP