Difference between revisions of "Special:Badtitle/NS100:Algorithms"
| Line 1: | Line 1: | ||
{{Projects/NAMICWeb}} | {{Projects/NAMICWeb}} | ||
| − | + | = Algorithms = | |
[[Image:Big-Algorithm-Logo.png|150px|left]] | [[Image:Big-Algorithm-Logo.png|150px|left]] | ||
| − | + | NA-MIC is an interdisciplinary, multi-institutional center with the | |
| + | mission of creating, developing, integrating and deploying computation | ||
| + | tools for the analysis and visualization of medical image data. | ||
| + | Core 1, the Algorithms Core, focuses on the development of new methods | ||
| + | for medical image analysis, new computational strategies for | ||
| + | implementing certain classes of algorithms, the deployment of these | ||
| + | algorithms within the software framework of the NAMIC Kit (developed | ||
| + | in Core 2), and the | ||
| + | application of these algorithms to clinical or scientific questions, | ||
| + | such as those presented by the driving biological problems in Core 3. | ||
| + | Also important are the dissemination of this work to the broader | ||
| + | medical community and the training of medical, clinical, and | ||
| + | scientific professionals in the use of these algorithms and the | ||
| + | associated software. | ||
| − | + | ==Philosophy and Priorities== | |
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | + | Several priorities or goals underly this work. First, is that | |
| + | algorithms should be fundamentally motivated and sufficiently general to be | ||
| + | used in a wide range of applications in biology and medicine. In this | ||
| + | way algorithms, and the associated software modules, will | ||
| + | be reconfigurable to be useful in a variety of software | ||
| + | applications. | ||
| + | This generality will allow the software associated with this center to | ||
| + | address image data: | ||
| + | *at a range of scales, starting at the scale of organs, and ranging down to the scale of cells; | ||
| + | *across a range of modalities, starting with structural, functional and diffusion MRI, quantitative EEG, and metabolic and receptor PET, and extends to include microscopic, genomic and other image data; | ||
| + | * from subjects taken across a wide range of time scales; | ||
| + | * from both from individual patients as well as studies executed across large populations; and | ||
| + | * for a broad range of diseases, in a broad range of organs. | ||
| + | A second goal is that algorithms should be | ||
| + | developed to the state where they are demonstrably useful for | ||
| + | applications represented by our driving biological problems (in Core | ||
| + | 3). This utility is demonstrated through either systematic validation | ||
| + | or through production of new clinical results. This advanced | ||
| + | development and validation is supported within the project by | ||
| + | validation software, data, and benchmarks, | ||
| + | organized events for comparisons and validation (e.g. see the the DTI | ||
| + | Validation Projection and associated Santa Fe Meeting [link]). It is | ||
| + | also fostered by ongoing interactions and collaborations with Core 3 | ||
| + | partners (DBPs), ancillary projects, and | ||
| − | To focus the technical development of these tools, as well as to address critical health care issues, the initial driving biological projects (DBPs) for the development and deployment of these computational methods came from the study of schizophrenia | + | A third priority is the dissemination of these algortihms to the |
| + | broader scientific and biomedical community. This is achieved | ||
| + | through conventional academic channels, such as journal publications | ||
| + | and conference presenations, and also through | ||
| + | integration of these algorithms into software applications, direct | ||
| + | collaboration with DBPs and other clinical collaborators, and | ||
| + | participation in Core 5 training sessions. | ||
| + | |||
| + | ==Technical Focus== | ||
| + | |||
| + | To focus the technical development of these tools, as well as to | ||
| + | address critical health care issues, the initial driving biological | ||
| + | projects (DBPs) for the development and deployment of these | ||
| + | computational methods came from the study of schizophrenia. | ||
| + | Subsequent DPBs address problems in lupus, autism, velocardiofacial | ||
| + | syndrome (VCSF), and prostate cancer, as detailed in | ||
| + | [[Projects/NAMICWeb:Driving Biological Projects|'''Core 3''']]. These | ||
| + | tools are being designed to apply, with minimal amplification or | ||
| + | modification, to other critical medical needs including multiple | ||
| + | sclerosis, Alzheimer’s disease, epilepsy, prenatal alcohol | ||
| + | syndrome. Moreover, we expect to continue to develop other potential | ||
| + | collaborations in other diseases as our center evolves. | ||
| + | |||
| + | Within this scope of applications Core 1 partners have several areas technical focus. These are: | ||
| + | * __Shape Analysis:__ The development of statistical methods and tools for quantitative analysis of anatomical shapes across subjects | ||
| + | * __Diffusion Weighted (Diffusion Tensor) Images:__ Methods for filtering, visualizing, and analyzing DWI and DTI datasets | ||
| + | * __Functional and Multimodal Image Analysis:__ Methods for the statistical analysis of fMRI and MEG and multimodal analysis of functional and structural data. | ||
| + | * __Partial Differential Equations:__ Applications of PDEs to image filtering and analysis and new algorithms for efficiently solving PDEs on conventional and specialized hardware. | ||
| + | |||
| + | Core 1 Partners: | ||
| + | # Massachussets Institute of Technology | ||
| + | # Georgia Institute of Technology | ||
| + | # The University of North Carolina | ||
| + | # The University of Utah | ||
| + | # University of California, Los Angelos | ||
<gallery Caption="Algorithm Core Members" widths="100px" heights="75px" perrow="5"> | <gallery Caption="Algorithm Core Members" widths="100px" heights="75px" perrow="5"> | ||
Revision as of 00:18, 2 December 2007
Algorithms
NA-MIC is an interdisciplinary, multi-institutional center with the mission of creating, developing, integrating and deploying computation tools for the analysis and visualization of medical image data. Core 1, the Algorithms Core, focuses on the development of new methods for medical image analysis, new computational strategies for implementing certain classes of algorithms, the deployment of these algorithms within the software framework of the NAMIC Kit (developed in Core 2), and the application of these algorithms to clinical or scientific questions, such as those presented by the driving biological problems in Core 3. Also important are the dissemination of this work to the broader medical community and the training of medical, clinical, and scientific professionals in the use of these algorithms and the associated software.
Philosophy and Priorities
Several priorities or goals underly this work. First, is that algorithms should be fundamentally motivated and sufficiently general to be used in a wide range of applications in biology and medicine. In this way algorithms, and the associated software modules, will be reconfigurable to be useful in a variety of software applications. This generality will allow the software associated with this center to address image data:
- at a range of scales, starting at the scale of organs, and ranging down to the scale of cells;
- across a range of modalities, starting with structural, functional and diffusion MRI, quantitative EEG, and metabolic and receptor PET, and extends to include microscopic, genomic and other image data;
- from subjects taken across a wide range of time scales;
- from both from individual patients as well as studies executed across large populations; and
- for a broad range of diseases, in a broad range of organs.
A second goal is that algorithms should be developed to the state where they are demonstrably useful for applications represented by our driving biological problems (in Core 3). This utility is demonstrated through either systematic validation or through production of new clinical results. This advanced development and validation is supported within the project by validation software, data, and benchmarks, organized events for comparisons and validation (e.g. see the the DTI Validation Projection and associated Santa Fe Meeting [link]). It is also fostered by ongoing interactions and collaborations with Core 3 partners (DBPs), ancillary projects, and
A third priority is the dissemination of these algortihms to the broader scientific and biomedical community. This is achieved through conventional academic channels, such as journal publications and conference presenations, and also through integration of these algorithms into software applications, direct collaboration with DBPs and other clinical collaborators, and participation in Core 5 training sessions.
Technical Focus
To focus the technical development of these tools, as well as to address critical health care issues, the initial driving biological projects (DBPs) for the development and deployment of these computational methods came from the study of schizophrenia. Subsequent DPBs address problems in lupus, autism, velocardiofacial syndrome (VCSF), and prostate cancer, as detailed in Core 3. These tools are being designed to apply, with minimal amplification or modification, to other critical medical needs including multiple sclerosis, Alzheimer’s disease, epilepsy, prenatal alcohol syndrome. Moreover, we expect to continue to develop other potential collaborations in other diseases as our center evolves.
Within this scope of applications Core 1 partners have several areas technical focus. These are:
- __Shape Analysis:__ The development of statistical methods and tools for quantitative analysis of anatomical shapes across subjects
- __Diffusion Weighted (Diffusion Tensor) Images:__ Methods for filtering, visualizing, and analyzing DWI and DTI datasets
- __Functional and Multimodal Image Analysis:__ Methods for the statistical analysis of fMRI and MEG and multimodal analysis of functional and structural data.
- __Partial Differential Equations:__ Applications of PDEs to image filtering and analysis and new algorithms for efficiently solving PDEs on conventional and specialized hardware.
Core 1 Partners:
- Massachussets Institute of Technology
- Georgia Institute of Technology
- The University of North Carolina
- The University of Utah
- University of California, Los Angelos
- Algorithm Core Members
Ross Whitaker, Guido Gerig, SCI Institute, U of Utah
David Kennedy, Bruce Fischl, Martinos Center, MGH
Polina Golland, Eric Grimson, Csail, MIT
Martin Styner, UNC, UNC
Allen Tannenbaum, BME, Georgia Tech
