Attachment 6 NIMH20003949RFQA01 Questions and Answers.pdf

NIMH20003949RFQA01 Questions and Answers:

Question: If this is a follow-on requirement, could you please provide the previous contract or task order number?

Answer: NA. This is a new requirement.

Question: Is there an incumbent contractor currently providing these services? If yes, please provide us with the contract details.

Answer: NA. This is a new requirement.

Question: What will be the set-aside and Contract value for this opportunity?

Answer: This is a total small business set-aside. The value will be determined at award.

Question: May the Offeror use Key Personnel’s personal experience to satisfy the two Past

Performance references?

Answer: No. Performance references should be professional.

Question: Will the government please clarify the number of Key Personnel who must meet the qualification requirements as mentioned in the Level of Effort section of Attachment?

Answer: The Government is not prescribing a specific number of Key Personnel, or a Skill Mix;

however, Key Personnel contained in the successful offeror’s technical approach will be added to the TO via KEY PERSONNEL, HHSAR 352.237-75 (December 2015)

Question: What is the timelines expectation of the project?

Answer: The timelines, to the best of NIH’s understanding, are specified in the statement of work.

Question: What are the performance requirements?

Answer: The performance requirements, to the best of NIH’s understanding, are specified in the statement of work. Further requirements are addressed via the answers contained in this document.

Question: Will there be a virtual site tour to cover what the day in the life of their users go through?

Answer: No.

Question: What vendor systems are currently being used? PACS/RIS etc.

Answer: For MR scanners, Siemens, General Electric, Philips, and Bruker. Communications with a

CareStream PACS system should also be expected. Most data are aggregated and made available on a pair of custom, open-source stacks, utilizing in-house Python-based tools and DCMTK for organizing data, and MySQL and PHP to make these data available via a secure web site.

Question: What tech stack will the government expect?

Answer: NIH is not mandating any particular tool or combination of tools. Preference will be given to leveraging existing, proven, open source tools and open cross-platform protocols.

Question: Can the vendor choose the languages and operating systems used?

Answer: There will be no language restrictions, but preference will be given to widely used and openly available cross-platform languages (including, but not limited to C/C++/Python/Javascript), protocols, technologies, and APIs. Offers utilizing modern versions of RedHat or Ubuntu Linux as the base operating system are preferred.

Question: Will the government allow on-site AWS technology like outpost?

Answer: NIH will consider these options, if all requirements can be met.

Question: Will there be any HL7 (Health Level Seven), HIS (Health Information System) or RIS

(Radiology Information System) integrations involved?

Answer: Yes.

Question: What is the value of the contract (NIMH20003949RFQ)?

Answer: NA. This is a new requirement. NIMH20003949RFQ is the Request for Quote Notice ID of the new requirement.

Question: Is this a new requirement? If not, who is existing incumbent?

Answer: Yes. NA.

Question: What is the estimated date of official RFQ/RFP solicitation announcement? How long it will remain open?

Answer: The RFQ will be open from 7/13/20 until 8/4/20 as noted in the solicitation.

Question: The SOW identified the tools AFNI, FreeSurfer, and FSL; are these tools required or are they

FMRIF recommended tools?

Answer: The local user base utilizes and are already familiar with these tools. Offers that do not or cannot utilize or incorporate these tools will not be scored highly.

Question: What is the key functionality that the existing system does not provide today

Answer: The primary new functionality would be to provide interfaces for managing users’ access to data (i.e. assigning role-based privileges set at the project level, e.g. admin, read/write, read-only, and use groups to organize users into labs, teams, for example). Secondary new functionality would be to provide tools for version tracking and maintaining provenance of data and analysis pipelines. (Both of these are listed in the Statement of Work).

Question: This solution is for research purposes only and HIPPA regulations are not needed?

Answer: The systems will house and manage PII, so HIPAA compliance will be needed.

Question: How many scans a year are anticipated

Answer: The number of scans is extremely difficult to estimate. The system should be able to process on the order of 10,000 sessions per year, with each session having anywhere from 5 to 20 scans.

Question: Is there a need for any analytics within the scan or purely document storage, workflow and

Enablement

Answer: Yes – at a minimum, users may want to obtain QA metrics on image quality. For example, the open source MRIQC (https://github.com/poldracklab/mriqc/releases) is a likely candidate to perform within-scan analytics. Offerors providing a wider range of capabilities beyond data storage and management will receive higher scores.

Question: How many users of the systems

Answer: There may be 250 – 300 users with access to the system. We expect typical utilization to be in the range of 30 – 50 users at any given time.

Question: How many different security groups

Answer: There will be approximately 5 – 6 tiers of security, permissions, and privileges needed for the platform to be adequately managed. There will be approximately 50 labs, with sub-projects, into each of which data can be siloed.

Question: Is the data and workflow management platform meant to be installed on-premises (using local computational resources) or on a virtual system (using computational resources on a cloud vendor), as long as it is offered remote management capabilities?

Answer: Current requirements specify an installation to on-premise physical or virtual resources to manage and warehouse data. The platform can be managed remotely, pending setup of MOUs and ISAs between NIH and the offeror.

Question: As part of the data capturing process into the platform there are "plugged" data preparation steps for the anonymization of the files (configurable). When the requirement mentions that users can download the "original" data, does this mean that users can download data as uploaded, so even before anonymization? Or it relates to download original data without any modifications to the "pixel data"?

Answer: Original data (before anonymization) should be accessible. Configurable anonymization must be an option, to be able to generate data for sharing.

Question: If the platform is already using commercial cloud compute resources, should it be able to interface with local on-premise HPC systems?

Answer: Yes – this is one of the Specific Requirements in the Statement of Work.

Question: Would it be possible to receive a list of the imaging scanners that the group would like to bring in data from? It would be helpful to know the number of scanners as well as the company/model of each.

Answer: Siemens (Magnetom 7T, Terra 7T, Skyra, Prisma), General Electric (MR-750, Premier), Philips, and Bruker scanners may send data to this system. Communications with a CareStream PACS system should also be possible.

Question: Could you please provide an estimate of how many new, prospective sessions that the group is planning to bring into this platform on an annual basis over the next 5 years. Additionally, if known, could you please indicate the estimated number of prospective scans that would be research sessions versus phantom or test scans.

Answer: The number of scans is extremely difficult to estimate. The system should be able to process on the order of 10,000 sessions per year, with each session having anywhere from 5 to 20 scans.

Approximately one third of these can be phantom / test scans.