SOLICITATION NOTICE
T -- Modification - : Computational Photography Project for Pill Identification: Content-Based Information Retrieval of a Digital Image Inventory of the Nation’s Oral Solid-Dose Prescription Medications
- Notice Date
- 2/22/2010
- Notice Type
- Presolicitation
- NAICS
- 541990
— All Other Professional, Scientific, and Technical Services
- Contracting Office
- Department of Health and Human Services, National Institutes of Health, National Library of Medicine, 6707 Democracy Blvd., Suite 105, Bethesda, Maryland, 20894, United States
- ZIP Code
- 20894
- Solicitation Number
- NLM-10-057-CYC
- Archive Date
- 3/17/2010
- Point of Contact
- Cara Y Calimano, Phone: 301 496-6127, Karen D Riggs, Phone: 301 496-6546
- E-Mail Address
-
CalimaC@mail.nlm.nih.gov, kr33v@nih.gov
(CalimaC@mail.nlm.nih.gov, kr33v@nih.gov)
- Small Business Set-Aside
- N/A
- Description
- Original Title : NLM Project Proposal: Computational Photography Project for Pill Identification: an NLM Public Safety Initiative - a Digital Image Inventory of the Nation's Oral Solid-Dose Prescription Medications Revised Title - NLM Project Synopsis - MODIFICATION: Computational Photography Project for Pill Identification: Content-Based Information Retrieval of a Digital Image Inventory of the Nation's Oral Solid-Dose Prescription Medications This project is funded through the American Recovery and Reinvestment Act of 2009 (ARRA) and is subject to the reporting requirements of the act. This is a combined synopsis/solicitation. This acquisition will be conducted under the procedures for acquiring commercial items authorized in FAR Part 12. This announcement constitutes the formal Request for Information (RFI) and a separate written solicitation WILL be issued. The solicitation number is RFI No. NLM-10-057-CYC. This acquisition is not a small business set-aside and the NAICS Code is 541990. In a national effort to promote patient safety, the National Library of Medicine (NLM) proposes to create a comprehensive, public digital image inventory of the nation's commercial prescription solid dose medications. The primary intention of this effort is create a test data collection for the advancement of automatic pharmaceutical identification through computer vision from photographic data. NLM expects to promote computer vision research applied to the domain of content-based information retrieval (CBIR) of solid-dose pharmaceuticals, and anticipates the need for generating a test environment, including variations of photographs of the same drug or sample under different environments. We invite ideas and plans for the structure for such a test collection subset of the overall imaging effort including a discussion of issues such as: (1) the number of validated sample images needed to evaluate automated drug identification algorithms, (2) the different photographic conditions and equipment necessary to effectively support a test image collection, and (3) a plan for deploying and maintaining such a collection in support of a grand challenge or other mechanism for promoting computer vision for solid-dose pharmaceutical identification. These data will be cross-correlated with high-resolution imagery of the same products, providing a baseline and authoritative measurement of attributes of the samples including color, shape, and surface properties such as scoring, coating, or debossed text. Our current effort is a continuation and extension of our previous project providing the Veterans Administration with a complete digital photographic catalog of the VA Centralized Mail Order Pharmacy formulary for prescription tablets and capsules. We intend to create a collection of digital photographs of prescription tablets and capsules, acquiring the images through a Simplified Acquisition Procedure. This will be an award to a contractor or vendor who will be responsible for formulating a protocol with the participation of NLM staff for acquiring the sample medications and creating images of medications to promote the development of automatic computer-vision-based identification of solid-dose pharmaceutical samples including variations in lighting, camera angle, background, photographic equipment, image resolution, and other attributes. We believe that there are between 5,000 and 32,000 prescription solid-dose oral medications to be cataloged through this process. In addition to these images for research support, the contractor will be responsible for creating high resolution digital photographs of the front (obverse) and back (reverse) surfaces of the samples using conventional techniques for color management, confirming that the images match the description of the medication, developing and matching the images of the samples to relevant metadata (including size descriptions, dimensions, color, and the provenance of the sample), and transmitting those data (images and metadata) to the NLM. The contractor will deliver matching test images accompanying the reference images to develop test collections for CBIR research. The CBIR test data will be crafted into training and validation collections to support research in computer vision. The accompanying high-resolution digital image collection from this effort will be placed by NLM in the public domain. The impact of these collections will target patient safety including assisting in drug identification for poison control hotlines as well as online services for the public and nursing care professionals to confirm the identification of a drug to help prevent mistakes in medication. Other uses of the image collections include education and applications in pharmacy pedagogy such as its inclusion in textbooks and digital teaching materials. These applications demand high resolution, high quality images with good color reproduction to faithfully represent the medications through imagery. Smaller images and subsets of the collection can be generated from the original high-resolution data for other applications that do not require the same level of fidelity; however, we are committed to acquiring the best data possible through this mechanism. NLM intends to identify a contractor or assemble a contractor-and-subcontractor team to address these issues, with the emphasis on experienced digital macro-photography talent as well as credentials and experience necessary to safely handle oral solid-dose prescription medications. The matching of the images to the correct identification of the medications is at least as critical as the dispensing of the medications to patients for treatment in that the impact of a mis-identification could have grave consequences. The need for quality control and verification in this project is likely to be significantly higher since the distribution of the resulting data could have widespread public impact, both a positive one if the collection is comprehensive and detailed, and significant negative impacts if there are errors in the database. The contractor should have clear explicit procedures for verifying the data and its correct correspondence to the medications from which it was derived. NLM expects the final team to incorporate photographers, assistants familiar with the handling of pharmaceuticals, a pharmacist for the appropriate qualifications in receiving, handling, and disposing of prescription medications, administrative technicians for the verification and electronic delivery of the digital images and their associated data, and information technology support for image processing, the extraction metadata, and technical support of the digital camera, its software, data storage and backup, as well as internet support to enable the electronic delivery of the digital data. Finally, and perhaps most importantly, the final team should include computer science personnel to help craft the CBIR component of the project. We anticipate that this first award will encompass the acquisition, photography, and the electronic delivery of the resulting digital data of between 4,500 and 6,000 prescription medications. Proposals addressing this requirement should not surpass 1 year in duration. Up to four (4) options for additional services (the additional acquisition of images of pharmaceutical samples in lots of 6,000 samples) should be provided by teams responding to this requirement. The review and award process will follow a two-step process: 1. A Request for Information (RFI) - Search for Sources will be conducted: Any offeror or group wishing to participate in the competition process MUST reply to the search for sources. Potential sources for this work will be asked for examples of their previous work and statements of their capabilities to meet the challenges posed by this project. If a substantial number of contractors respond to this RFI, NLM may use a Technical Evaluation Group (TEG) to evaluate the potential sources on their existing facilities, technical approach, capabilities for the acquisition and disposal of samples, expertise of available personnel, consultants, and subcontractors (including photographers, computer scientists, image processors, pharmacists, and technicians), and past performance. NIH personnel reserve the right to visit potential sites as part of the search for sources; however, a site visit is not required for a future award. 2. A Request for Quotes (RFQ) will be issued for a photographic collection. Only sources identified as competitive through the search for sources will be solicited for quotes for the competition. A complete Special Emphasis Panel (SEP) will be convened to make the award including any options to the contract or purchase order. These/this award(s) will be executed quickly through a Simplified Acquisitions Mechanism under FAR Part 13. Response to the Search for Sources: NLM will review the responses to the search for sources based on an evaluation of the responder's capabilities to execute both the expected photographic collection as well as the optional creation of the computer vision test data set. Responses to the Request for Information (RFI) that constitutes the search for sources will evaluate: 1. Proposed methods for acquiring the data: The response will be evaluated based on the overall technique. A complete proposal is not required, but rather an overview of the proposed method. How does the potential contractor plan to acquire sample medications? How will the contractor ensure that all images are properly identified with the correct medication? How will they process the image data for color management, size accuracy, and background removal? 2. Qualifications of existing personnel: The response will be evaluated based on the technical merits of the named existing personnel. Do they have experience in digital photography? Do they have expertise in managing digital images for record photography, curatorship, or some other related enterprise? 3. Existing facilities: Since this award process is not intended to provide funds for new equipment acquisition, the responders should describe their current capabilities for acquiring the data for the proposed photographic collection. If modifications of any equipment is required, the responses should describe the extent of the modifications, what capabilities are possible using the current configurations, and what capabilities will be enabled by proposed changes to the equipment. 4. Past performance: The responses should provide sources for relevant references of past performance in data acquisition. 5. Samples of previous work: The responses should include photographic or digital illustrations and evidence of previous results. These results should demonstrate the capability of the responder to provide the proposed data with accurate identification of the contents of individual images. Page Limit to responses to the RFI: Responses to the RFI search for sources should be limited to 10 pages of textual information (with figures and tables). Additional illustrations, photographs and evidence of results (presented in photographs or digital information) must accompany the response. The total response should not exceed 15 pages (text plus pictures). Due Date: Responses to the search for sources RFI are due on March 2, 2010. Electronic copies of the RFI response are due by the application deadline. In addition to the official electronic submission, two printed copies of the Sources Sought response are required. These printed hard copies should be printed, 3-hole punched, and submitted in binders. NLM anticipates an official deadline for electronic submission by noon (12:00 PM), Tuesday March 2, 2010, with an official deadline for printed hard copy submission by close of business, Friday March 5, 2010. Electronic copies of submissions should be sent via e-mail to the Contracting Specialist, Cara Calimano, at calimac@mail.nlm.nih.gov. The submission may be split into multiple e-mails to accommodate e-mail servers. At the completion of the submission of responses to the Sources Sought RFI, even if only a single e-mail is transmitted, an additional confirmation text-only e-mail should be transmitted, communicating that a complete submission has been sent and detailing the contents of each file and how many separate e-mail files should be expected. Paper copies may be submitted in writing to the National Library of Medicine, Office of Acquisitions Management, 6707 Democracy Blvd. Suite 105, Bethesda, Maryland 20894, Attention: Cara Calimano, Contract Specialist. Response to the RFQ: NLM will solicit quotes from sources identified through the Sources Sought RFI. You must respond to the RFI to be considered for the award. If the Sources Sought RFI generates a large number of responses, NLM will use a Technical Evaluation Group to determine which offerors will be invited to respond to the RFQ. Technical proposals should cover: (i) Technical details for: 1. Sample Acquisition • NLM will prioritize the list of images of samples to be acquired, beginning with the most widely prescribed drugs as well as the those medications that pose a public threat either through abuse or as a poisoning hazard. NLM proposes the use of DailyMed as the source of the initial list of prescription medications (4,676 medications listed as of 10-June-2009) • If samples are offered by participating manufacturers or vendors, the contractor will accept these samples as available. • The contractor will be responsible for establishing a source of medications and ensuring that their supply chain remains viable through the life of this project. • The contractor will establish and maintain procedures that comply with all state and federal regulations for the management, handling, and disposal of prescription medications. 2. Photographic quality • The contractor will adhere to image quality guidelines established by the NLM pertaining to issues of: color management (ensuring that the images provided reflect the actual color of the medication), image resolution (Minimum of 6 MPixels for the field including the pharmaceutical sample), file formats, color background (18% grey color background), and fiducial marks for size, color, and white balance. NLM will provide a standard color reference for matching colors in the raw digital images. 3. Quality control and Data transfer • The contractor will develop quality control procedures with the NLM for the verification of acceptable images, matching the images with the correct metadata, and data transmission. • Images and metadata will be transferred to NLM, subject to verification. 4. Transition plan - project completion • After the data collection is firmly established and the period of performance nears its end, the contractor may be required to transfer the program another contractor. The plan should address a process for disclosing all documentation, file formats, database organization, and open-source software. This transition is essential for the long-term sustainability of the project. (ii) Qualifications of key personnel to execute this project. 1. Background expertise - the extent to which the key personnel and the project team as a whole have access to the range of educational backgrounds in macrophotography, in particular in photography for documenting scientific or records of physical samples. 2. Experience - the extent to which the key personnel and the project team include significant experience and expertise in the production workflows for documentation photography. (iii) The commitment and experience of the institution or offeror in building and releasing public data resources for patient safety in pharmaceutical information. (iv) Proposed relevant facilities Page Limit to Technical Proposals in response to the RFQ: The Technical Proposal should include examples digital images of pharmaceutical samples, demonstrating the capabilities of the respondent. These images are not included in the 15-page limit. Beyond these digital images 15 pages of technical proposal, resumes of key investigators, reprints of relevant technical papers, and other supplemental materials can be provided outside of the page limit. Reviewers are instructed that the 15-page technical proposal and the representative digital images are required reading, and supplemental materials are available as part of the support documentation but not required. NLM is specifically seeking contractors with the discipline necessary to conduct this project, adhering to high standards of quality and consistency necessary to acquire high quality reference images through digital macrophotography and to verify and validate that images of these solid-dose pharmaceuticals are accurately tied to their respective medications. Regular progress evaluations may be initiated, and frequent interaction with the NIH staff responsible for the imaging informatics and release of the data to the public domain is expected. The deliverables for this proposed project are the photographs of the solid-dose pharmaceutical samples, the metadata and information for correctly identifying and correlating the images with their corresponding medications. The resulting images and the final data will be owned by the National Institutes of Health and distributed to the public. The NIH reserves the right to include a site visit as part of the proposal evaluation. The requirement for the photographic collection will be procured in accordance with FAR Part 13 - Simplified Acquisitions Procedures. NLM anticipates that the first award of the photographic collection (up to 6,000 samples) will be completed within one year and costs will not exceed $500,000 including the cost of acquiring and disposing of the sample medications. All quotes will be evaluated on technical merit, qualifications of the personnel (including photographers, computer scientists, image processors, pharmacists, and technicians), cost, and past performance on research and development projects. Request for Quotations (RFQ) Number NLM 10-058/CYC covering the photographic collection will be available on or about March 15, 2010. RFQ NLM 10-058/CYC may be obtained by writing to the National Library of Medicine, Office of Acquisitions Management, 6707 Democracy Blvd. Suite 105, Bethesda, Maryland 20894; Attention: Cara Calimano, Contract Specialist or by email request to calimac@mail.nlm.nih.gov. Facsimiles WILL NOT BE ACCEPTABLE. Quotes sent in response to the RFQ must be valid for up to 120 days. Awards for the photographic collection are anticipated on or about May 28, 2010. Only sources that are invited by NLM from among the respondents to the Search for Sources/Request for Information that were found competitive by the sources review may submit a proposal for consideration by the NLM. Responses to the RFQ will be due on April 29, 2010. Electronic copies of the technical proposal are due by the application deadline. In addition to the official electronic submission, two printed copies of the technical proposal are required. These printed hard copies should be printed, 3-hole punched, and submitted in binders. NLM anticipates an official deadline for electronic submission by noon (12:00 PM), Thursday April 29, 2010, with an official deadline for printed hard copy submission by close of business, Tuesday May 4, 2010. Electronic copies of submissions should be sent via e-mail to the Contracting Specialist, Cara Calimano, at calimac@mail.nlm.nih.gov. The submission may be split into multiple e-mails to accommodate e-mail servers. At the completion of the submission of responses to the RFQ, even if only a single e-mail is transmitted, an additional confirmation text-only e-mail should be transmitted, communicating that a complete submission has been sent and detailing the contents of each file and how many separate e-mail files should be expected. Paper copies may be submitted in writing to the National Library of Medicine, Office of Acquisitions Management, 6707 Democracy Blvd. Suite 105, Bethesda, Maryland 20894; Attention: Cara Calimano, Contract Specialist. All responses must be in writing and must include pricing information and meet the criteria requirement. All responses from responsible sources will be considered. Please reference the solicitation number "NLM-10-057-CYC" on your proposal and/or inquiries. Facsimile submissions are not authorized and collect calls with NOT be accepted. Contracting Office Address: 6707 Democracy Blvd., Suite 105 Bethesda, Maryland 20894 United States Primary Point of Contact.: Cara Calimano cc436e@nih.gov
- Web Link
-
FBO.gov Permalink
(https://www.fbo.gov/spg/HHS/NIH/OAM/NLM-10-057-CYC/listing.html)
- Place of Performance
- Address: 6707 Democracy Blvd. Ste. 105, Bethesda, Maryland, 20892, United States
- Zip Code: 20892
- Zip Code: 20892
- Record
- SN02071309-W 20100224/100222235953-cf6132c2286d98d8d8d6d265da8c06f5 (fbodaily.com)
- Source
-
FedBizOpps Link to This Notice
(may not be valid after Archive Date)
| FSG Index | This Issue's Index | Today's FBO Daily Index Page |