Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
High Throughput Pattern Matching of fMRI Maps and Molecular Features
We develop high-level fMRI pipelines with the goal of increasing automation and throughput in the field of animal imaging. For this project we are looking to leverage information theoretical metrics and large scale databases in order to better relate fMRI data to underlying molecular features.
Keywords: fMRI, neuroscience, biomedical engineering, data analysis, histology, psychopharmacology, free and open source, FOSS, Python, Git, neurosynth, high throughput, animal imaging, pattern matching, information theory, multimodal, pattern recognition, molecular biology, cell biology, histology
You will address a data management and information theoretical challenge with ample relevance for functional imaging and neuroscience. You will use state-of-the art software and data produced by cutting edge biomedical imaging. You will be given abundant resources in both operational costs and tutoring in order to fulfill your goal.
.
**This project affords numerous opportunities:**
* Become familiarized with modern data analysis environments in both human and animal neuroimaging.
* Understand the logical complexities associated with regularizing data for preclinical research.
* Gain insight into biomedical workflows and the software engineering opportunities available therein.
* Utilize and enhance your practical abilities with Python.
* Develop a free and open source project, and become accustomed to Git and the world of collaborative coding.
* Receive proficient tutoring in Python, Linux, and Git, as needed.
.
**To adequately tackle the challenges of this project you should have:**
* *some* previous experience coding in Python.
* *some* previous experience organizing high volumes of data.
.
**Though not mandatory or assumed, the following would be a significant plus:**
* prior experience working on Linux.
* prior experience with web API usage and design.
.
**Research Context and Technical Details**
A common end result of fMRI experiments is a statistical map. This 3D data structure represents the modulatory/effect on brain activity of a stimulation, intervention, state, or an interaction thereof.
This is a compact data representation, which can easily be shared and processed in order to integrate findings from multiple experiments.
Information metrics allow such data not only to be meaningfully compared between fMRI modalities (e.g. BOLD and CBV), but also between fMRI and entirely different modalities with molecular and cellular specificity. These comparisons are key for establishing better relationships between molecular structure and brain function.
The Allen Brain Institute (ABI) publishes a large library of semiquantitative antigen density and cell type projection maps (www.brain-map.org) regularized to a standard template an accessible via a complex but open API. The available data fulfills all the requirements for constructing a lookup library (e.g. in order to rank structural maps for any given functional map in similarity metric order).
Our in-house data analysis suite, SAMRI (github.com/IBT-FMI/SAMRI), follows a standardized functional brain imaging organisation scheme (BIDS, bids.neuroimaging.io), and thus generates suitable substrates for high-throughput library lookup. Additionally, there exists software providing automatic query and 3D model synthesis for the ABI human brain data (neurosynth, github.com/neurosynth/neurosynth), which can be used as inspiration for an ABI-to-SAMRI similarity ranking library.
You will address a data management and information theoretical challenge with ample relevance for functional imaging and neuroscience. You will use state-of-the art software and data produced by cutting edge biomedical imaging. You will be given abundant resources in both operational costs and tutoring in order to fulfill your goal.
.
**This project affords numerous opportunities:**
* Become familiarized with modern data analysis environments in both human and animal neuroimaging. * Understand the logical complexities associated with regularizing data for preclinical research. * Gain insight into biomedical workflows and the software engineering opportunities available therein. * Utilize and enhance your practical abilities with Python. * Develop a free and open source project, and become accustomed to Git and the world of collaborative coding. * Receive proficient tutoring in Python, Linux, and Git, as needed.
.
**To adequately tackle the challenges of this project you should have:**
* *some* previous experience coding in Python. * *some* previous experience organizing high volumes of data.
.
**Though not mandatory or assumed, the following would be a significant plus:**
* prior experience working on Linux. * prior experience with web API usage and design.
.
**Research Context and Technical Details**
A common end result of fMRI experiments is a statistical map. This 3D data structure represents the modulatory/effect on brain activity of a stimulation, intervention, state, or an interaction thereof. This is a compact data representation, which can easily be shared and processed in order to integrate findings from multiple experiments.
Information metrics allow such data not only to be meaningfully compared between fMRI modalities (e.g. BOLD and CBV), but also between fMRI and entirely different modalities with molecular and cellular specificity. These comparisons are key for establishing better relationships between molecular structure and brain function.
The Allen Brain Institute (ABI) publishes a large library of semiquantitative antigen density and cell type projection maps (www.brain-map.org) regularized to a standard template an accessible via a complex but open API. The available data fulfills all the requirements for constructing a lookup library (e.g. in order to rank structural maps for any given functional map in similarity metric order).
Our in-house data analysis suite, SAMRI (github.com/IBT-FMI/SAMRI), follows a standardized functional brain imaging organisation scheme (BIDS, bids.neuroimaging.io), and thus generates suitable substrates for high-throughput library lookup. Additionally, there exists software providing automatic query and 3D model synthesis for the ABI human brain data (neurosynth, github.com/neurosynth/neurosynth), which can be used as inspiration for an ABI-to-SAMRI similarity ranking library.
Design a workflow which can programmatically fetch, reformat, and register Allen Brain Institute mouse data, and which calculates relevant similarity metrics for given fMRI statistical maps against the aforementioned library.
Design a workflow which can programmatically fetch, reformat, and register Allen Brain Institute mouse data, and which calculates relevant similarity metrics for given fMRI statistical maps against the aforementioned library.