SOURCES SOUGHT
R -- Development of a Task-related and Resting State Realistic fMRI Simulator for Benchmarking fMRI Data
- Notice Date
- 8/12/2015
- Notice Type
- Sources Sought
- NAICS
- 541712
— Research and Development in the Physical, Engineering, and Life Sciences (except Biotechnology)
- 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
- NIHLM2015604
- Archive Date
- 8/19/2015
- Point of Contact
- Suet Vu, Phone: 301-496-6546
- E-Mail Address
-
vus@mail.nih.gov
(vus@mail.nih.gov)
- Small Business Set-Aside
- Total Small Business
- Description
- GENERAL INFORMATION INTRODUCTION: This is a Small Business Sources Sought notice. This is NOT a solicitation for proposals, proposal abstracts, or quotations. The purpose of this notice is to obtain information regarding: (1) the availability and capability of qualified small business sources; (2) whether they are small businesses; HUBZone small businesses; service-disabled, veteran-owned small businesses; 8(a) small businesses; veteran-owned small businesses; woman-owned small businesses; or small disadvantaged businesses; and (3) their size classification relative to the North American Industry Classification System (NAICS) code for the proposed acquisition. Your responses to the information requested will assist the Government in determining the appropriate acquisition method, including whether a set-aside is possible. An organization that is not considered a small business under the applicable NAICS code should not submit a response to this notice. The National Institutes of Health (NIH), National Library of Medicine (NLM) is conducting a market survey to help determine the availability and technical capability of qualified small businesses, HUBZone small businesses; service- disabled, veteran-owned small businesses; 8(a) small businesses; veteran-owned small businesses; woman-owned small businesses; or small disadvantaged businesses capable of serving the needs identified below. Background: There is a particular need today to be able to generate a synthetic yet realistically modeled "ground truth" for various fMRI benchmarking. Due to the massive amount of noisy and complex data collected in a single fMRI study, the statistical analysis of fMRI data is challenging but essential for understanding and obtaining an accurate interpretation of the data (Lindquist, 2008). Many fMRI analysis techniques are ad hoc and without a solid basis or any ground truth; however, sufficiently modeled ground truth data is needed to validate the assumptions and the statistical methodology used to analyze the data, as well as to access statistical measures such as specificity and sensitivity. Such a ground truth could be estimated from intracranial EEG (iEEG) observations; however these are unobtainable for most human subjects. Recent research suggests that dynamic functional connectivity (FC) patterns may reflect intrinsic characteristics of neural brain functions; however, this research area is still exploratory and evolving (Huchison et al., 2013). Therefore, development of a truly realistic fMRI simulator requiring full knowledge of inherent signal as well as noise characteristics is not quite possible at this time. Development of any realistically modeled fMRI simulator must allow iterative processes for upgrading the model as more information becomes available. Currently there are four well-known fMRI simulators, two in-house and two cross-platform applications; these are POSSUM (Drobnjak et. Al., 2006), NeuRosim (Welvaert et al.,2011), DCM (Marreiros et al., 2008), and SimTB (Erhardt et al. 2011). However, each of the above four simulators has its own limitations. The in-house simulator for FSL (FMRIB's Software Library) is called POSSUM (Physics-Oriented Simulated Scanner for Understanding MRI), which is more of an fMRI sequence simulator that takes T2* 4D data, typically back-calculated from data, as input. POSSUM requires a gradient echo pulse sequence, a segmented object with known tissue parameters, and a motion sequence; making it possible to simulate spin history effects, motion during the readout periods and interactions that these have with B0 inhomogeneities (Drobnjak, 2006). However, the current limitations of POSSUM are not including physiological noise, inability to incorporate any general pulse sequences, and not being friendly to general users such as psychologists and clinicians. The in-house fMRI simulator in SPM is a DCM (Dynamical Causal Modelling) connectivity simulator used to infer directed connectivity among brain regions (Marreiros et al., 2008). The MATLAB toolbox, SimTB, is useful for simulating the ICA components for group functional connectivity studies, but only operates on a 2D "cartoon" brain and lacks much of MR physics (Erhardt et al. 2011). The R-based neuRosim fMRI data simulator is still a work in progress and was created to allow for the experimental fMRI design to be specified, activation regions defined, and various statistical properties and noise simulated (Welvaert et al., 2011). Despite the creation of these four fMRI simulators, a comprehensive fMRI data generation program that can be used to fully optimize experimental fMRI data collection and provide a realistically modeled "ground truth" is still lacking (Welvaert, 2014). There are no current fMRI simulators that simultaneously employ realistic characterizations of the underlying MR physics and assumptions about the spatial and temporal frequencies of neural activity, including resting state networks (RSNs). RSNs are observed in many task-related fMRI studies (Raichle et al. 2001; Frannson et al. 2005), and it is essential to benchmark resting state fMRI (rfMRI) before any task-related fMRI time series can be properly simulated; also, rfMRI can be a study by itself (Biswall et al. 2001; Deco et al. 2011; Smith et al. 2013). Because of this lack of an accessible simulator, neuroscientists often only optimize parts of their experimental protocol, such as stimulus presentation (Dale et al., 1999), without considering fine-tuning of the image acquisition parameters or subsequent analysis for specific research goals, or their power and false-positive rates (e.g., Mumford et al. 2014). In addition, all of the above four simulators are mostly used by researchers, for specific research applications, and not by more "general" users (such as psychologists) who collect fMRI data, but without any validation of the collected data or design protocol used. The fMRI simulator proposed here provides a user friendly environment (currently MATLAB based with a final GUI) with capability of generating synthetic fMRI activation and time-series through Bloch equations solver (Figure 1) based on the known signal and noise characteristics available at this time, but having capability of being iteratively upgraded as more realistic modeling of signal and noise in fMRI are available. This would allow not only the researchers in neuroscience to validate protocol selection for their experimental goals tailored to hypothesized BOLD signals and functional connectivity, which are themselves indirectly related to the underlying neuronal activities and interactions (Hutchison et al. 2013), but also provide a modeled ground truth for validation of fMRI data, including in potential clinical applications in the future. Finally, recent work [Liu, Nutter, Ao, Mitra, 2011] in alternative coding of collected MRI signals, specifically wavelet encoding, suggests that significant efficiencies in data collection time may be achievable. A realistic fMRI simulator would provide a platform for further exploring and evaluating this technology. Proposed work The proposed work is to create an fMRI simulator with more realistic signal modeling, as compared to current simulators, a user-friendly interface for non-specialists in MRI/fMRI technology and physics, and capability to model wavelet-encoded signal acquisition. A description of the proposed methodology and validation methods are given below. Methodology First, a baseline brain model will be constructed from accurate proton density, T1 and T2* parameter maps, as well a fuzzy mask of the cortex to properly localize neural activity (see flowchart below). One requirement for an fMRI simulator are realistic 4D T2* maps. All changes in the magnetic field, ΔB, can be added to the baseline by the formula (Edelman et al., 2006): 1/(T_2^* (t))=1/(T_2 (t))+1/2 γ∆B(t) The various sources of magnetic deviations (off resonance terms) can then be added separately to the baseline. For example, BOLD effect activation blobs can be defined in the MNI152 space using existing probabilistic brain atlas masks, each masked to the cortex. Multiple activation blobs can be used to define activation clusters and saved as templates. Templates of various resting state networks are available (from back-calculating from group ICA experimental results) or can be constructed from models. A time series is defined for each activation template as well as activation statistical parameters. The time series is convolved with the regional HRF to generate the changes in the T2* baseline. A breathing pattern is generated and a typical ΔB-amplitude is plotted across the volume. Cardiovascular noise is also added with the heart-rate specified and areas affected most by vasculature masked. Other known artifacts such as eye movement can be added. Hence a combination of synthetic neuronal BOLD signals, physiological noise and other artifacts can be used to construct a synthetic 4D T2* map. This allows for a rigorous evaluation of existing analysis technique, since the hypothetical BOLD signal ground truth as modeled is assumed to be known. The resulting 4D T2* map along with sequence information could then be inputted into a Bloch equation solver where additional MRI-related artifacts, such as head motion and eddy current effects can be simulated. In this way synthetic yet realistically modeled 4D fMRI data can be generated in a NIFTI format. The experimentally acquired fMRI data can be analyzed and compared to the modeled ground truth using a variety of statistical optimization techniques. The proposed fMRI simulator will also have the capability of investigating the use of novel pulse sequences for wavelet encoding of the signal as opposed to traditional Fourier encoding (k-space) for high quality reconstruction of MR images at a significantly reduced sampling rate (Liu et al., 2013, Xie et al., 2015) using a new under sampling scheme, thus allowing faster and multi-resolution MR data acquisition. Approach to validation The need for proper objective functions in validity measure of fMRI data is still an open research topic. There are two main data structures to be compared when comparing simulated and real data: the 3D activation maps and their associated time series. Typical statistically derived voxel-wise comparisons (i.e. MSE) are noisy and error prone, especially for inter-subject comparisons (G.A. Cecchi et al., 2009). This is because voxel-wise comparison suffers from a lack of functional co-registration as well as other sources of variability. A better approach is to use cluster based methods, such as weighted cluster coverage, to compare two activation regions (R. Heller, et al., 2006). An information theoretic clustering approach (Hill et al. 2014) may also prove to be quite valuable in developing such validity measures. In general, the minimal achievable distortion, dK, in misclassification of clusters is defined as the mean squared Mahalanobis distance of every point in the cluster to the nearest cluster's centroid (or cluster representative), i.e. where xi is the ith data point that belongs to the kth cluster, ck is the centroid or representative point of the kth cluster,  is the inverse covariance matrix for each cluster (assumed to be approximately equal for all clusters), and K is the number of clusters considered. The details of this information theoretic clustering approach can be found in reference Hill et al., 2014. Figure 1. A flowchart for generating realistic 4-D fMRI data in a NIFTI format to provide an iteratively modeled ground truth. Time series comparison should also be done on the cluster level, and can also suffer from a phase offset due to variability of the HRF. Standard time series comparison techniques include auto-regressive (AR) spectral analysis and coherence, and must allow for a phase offset to be present. A useful statistical comparison is the comparison of the predicted error or residuals. Comparison on the group level can be done spatially by simple comparison of the mean activation amplitude of the activation regions; temporal comparisons may be accomplished by using a t-statistic that tests the difference in frequency bins versus the null hypothesis of no difference. Projected project timeline Because of the high research component of this work, it will be carried out in stages, with careful evaluations and feasibility assessments at each stage. Continuing the work to a subsequent stage will be dependent on and must be justified by favorable feasibility assessment of the work of the current stage. The expected work progression is as follows: (1) year one; simulate 3D fMRI including task-related and resting state, single subject, with Gaussian and Rician noise; (2) year two; simulate full 4D fMRI including task-related and basic resting state, single subject, and multiple noise sources including physiological noise; and (3) year three; simulate full 4D fMRI including task-related and resting state, and group subjects including subject variability and noise. Task Description The following tasks shall be completed. All deliverables shall include commented source code, documentation, sample program demonstrating capabilities, and appropriate performance metrics, It is the goal of the government to interact with the contractor at monthly intervals to stay informed of technical options that arise during the work, assess technical progress, provide project guidance, and exploit opportunities for early implementations of the research work for testing and feedback to the contractor. To this end, the contractor shall discuss progress by Web-based demonstrations, teleconference, or site visit, as appropriate and as negotiated with the government. Task 1: Design and carry out initial implementation of a simulator for 3D fMRI including task-related and resting state, single subject, with Gaussian and Rician noise. Task 2: Design and carry out detailed implementation of a simulator for 3D fMRI including task-related and resting state, single subject, with Gaussian and Rician noise, and carry out test and evaluation of the simulator. The contents of the deliverables shall be as follows: Deliverable 1: Detailed technical progress report and source and executable code for the initial (3D) fMRI simulator. Deliverable 2: Detailed technical progress report, including test and evaluation results, and source and executable code for the detailed (3D) fMRI simulator. Notice of Government Unlimited Rights to Work First Produced Under This Contract Government rights to work first produced under this contract are established by Federal law including, but not limited to, this specific reference: FAR 42.227-14, Rights in Data - General, (b) (1). Requirement to Notify Government of Proprietary Work Dependencies The Contractor is required to notify the Government in writing of any dependencies of the deliverables under this contract on proprietary, copyrighted, or patented work that potentially inhibits, restricts, or requires permission for the dissemination of the deliverables to the public, other governmental agencies or research groups, or to any other parties whatsoever. ANTICIPATED PERIOD OF PERFORMANCE: It is anticipated that the period of performance shall be for a 12 month base year with two (2) 12 month option periods. An award is anticipated to be made on or around September 2015. It is anticipated that the contract will be a Firmed-Fixed price type. Interested firms responding to this Sources Sought Notice must adhere to the following: (a) Provide a capability statement demonstrating relevant experience, skills, and ability to fulfill the Government's requirement. The capability statement should be complete and contain sufficient detail for the Government to make an informed decision regarding capabilities; however, the statement should not exceed 10 pages. (b) The capability statement must identify the responder's business type and size; DUNS number; NAICS code, and technical and administrative points of contact, including names, titles, addresses, telephone and fax numbers, and e-mail addresses. (c) The National Library of Medicine (NLM) requires proposals to be submitted via eCPS.: 1) Electronic copy via the NLM electronic Contract Proposal Submission (eCPS) website at https://ecps.nih.gov/nlm. All submissions must be submitted by 1:00pm, Local Prevailing Time, on August 18, 2015. For directions on using eCPS, go to https://ecps.nih.gov/nlm/home/howto and click on "How to Submit." NOTE: To submit your electronic proposal using eCPS, all offerors must have a valid NIH External Directory Account, which provides authentication and serves as a vehicle for secure transmission of documents and communication with the NLM. The NIH External Directory Account registration process may take up to 24 hours to become active. Submission of proposals by facsimile or e-mail is not accepted. EMAILS AND FACSIMILES WILL NOT BE ACCEPTABLE. Disclaimer and Important Notes: This notice does not obligate the Government to award a contract or otherwise pay for the information provided in response. The Government reserves the right to use information provided by respondents for any purpose deemed necessary and legally appropriate. Any organization responding to this notice should ensure that its response is complete and sufficiently detailed to allow the Government to determine the organization's qualifications to perform the work. Respondents are advised that the Government is under no obligation to acknowledge receipt of the information received or provide feedback to respondents with respect to any information submitted. After a review of the responses received, a pre-solicitation synopsis and solicitation may be published in Federal Business Opportunities. However, responses to this notice will not be considered adequate responses to a solicitation. Confidentiality. No proprietary, classified, confidential, or sensitive information should be included in your response. The Government reserves the right to use any non-proprietary technical information in any resultant solicitation(s).
- Web Link
-
FBO.gov Permalink
(https://www.fbo.gov/spg/HHS/NIH/OAM/NIHLM2015604/listing.html)
- Place of Performance
- Address: Bethesda, Maryland, 20894, United States
- Zip Code: 20894
- Zip Code: 20894
- Record
- SN03835237-W 20150814/150812235514-21694b4780129af7754aba2bb65b8cc7 (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 |