Slide 1Slide 2Slide 3SHARING NEUROIMAGING DATAHow can computing help understand how the human brain functions?Slide 6Comparing, correlating neuroscience dataSlide 8Slide 9Slide 10Slide 11Slide 12Slide 13What if people do not wish to share original data? SCENS: Secure Content Exchange Negotiation System A Metadata-Based Negotiation System For Sharing Diverse Research ResourcesHuman-Computer Interface Design Example Negotiation Conditions RegistrationNew NCR InterfaceSCENS Recommendation SystemRecommendation Problem FormulationOutput: Recommendations PresentationUsability Evaluation QuestionsUsability AnalysisUsability Evaluation: Collect data and analyze them properly, defend your analysisSlide 23Next : Let us now review CSESlide 25Engineering DepartmentsEngineering FacultyCollege of EngineeringBuilding PlansELB Expansion and ERBCSE Directions in EducationSlide 32COE Fall Enrollment Trends (Headcount)Fall Enrollment Trends (Semester Credit Hours)External Funding (BER)Department of Computer Science and EngineeringWho we areSlide 38Degree Programs OfferedResearch Areas of ExpertiseFillia MakedonNew Faculty MembersSlide 43Slide 44Multi-View Video: The Next Generation TelevisionUnmanned Aerial Vehicle TeamSlide 47Slide 48Analysis of Nodule Characteristics using Dynamic Contrast-enhanced MRISlide 50Slide 51Heracleia - Human Centered Computing LabSlide 53Slide 54Slide 55Slide 56Slide 57Nikola Stojanovic Bioinformatics, Genomics Algorithms for DNA sequence analysis http://ranger.uta.edu/~nick Research supported by National Institutes of Health (National Library of Medicine)Slide 59Slide 60Slide 61Summary and conclusionsHandling a Complex Digital World: Information Access in the 21st Century Modeling, Integrating, Mining and Sharing Heterogeneous Resources for Distributed ApplicationsDisparate disciplines coming togetherObservations - unmet existing demandsData Access in the 21st Century => Secure AccessComputer science is everywhereFilia MakedonProfessor and Department [email protected] Meaning of ComputingWhy study computer science?• Solving problems• Role of Algorithms• Examples• Airline scheduling• Searching government documents• Analysis of sensor data• Design of cars• Testing software• Brain Imaging analysis• Genome Research• Understanding human motion• Security• Weather analysis• Predicting Structural health• Art (Art Computing and Story Telling (ARCS course)• Is there a problem where computer science cannot help?Samples of Brain Activation Data: fMRI Motor MappingSHARING NEUROIMAGING DATA•Benefits?–Discovery (early detection) –Consolidation (data fusion) –Cost reduction (expensive data, scarcity of equipment)–Democratizing knowledge•Obstacles –Technical:•Heterogeneous data –2D images, 3D images, audio, video,text, animation, other• Voluminous and multidimensional•Sensitive (protected)–Non Technical- human•Owners’ inhibitions and lack of incentives•Security, privacy, liability•Ownership and digital rights management support1: Background 5How can computing help understand how the human brain functions?•What are these data?–Datasets and Methods (tools)–Multi-spectral •(MRI, PET, SPECT, Methods, fMRI, etc.)–Voluminous •50 MB of raw scanner data per 5 minute scan;•Hundreds of MB of processed data per subject•Computational needs?–Processing Massive amounts of data–Modeling Disparate data: raw data, methods, observations, protocols–Understanding the domain and working with domain expertsExamples of two modalities: PET (left) and MRI (right) imagesResting Brain Glucose Metabolism(FDG)Corresponding T1 Weighted Anatomic MRIAtlas ScanResting Brain Glucose Metabolism(FDG)Corresponding T1 Weighted Anatomic MRIAtlas ScanHow can we make sense out of different types (MODALITIES) of brain scans?Comparing, correlating neuroscience data19Semantic Activation: Category Function TaskP<0.005, k=0ControlsAlzheimer17Figure 1: Human brain data example. PET (left) andMRI (right) imagesResting Brain Glucose Metabolism(FDG)Corresponding T1 Weighted Anatomic MRIAtlas ScanResting Brain Glucose Metabolism(FDG)Corresponding T1 Weighted Anatomic MRIAtlas Scan18fMRI Motor MappingNeed to model Data and Methods;Need to model Data and Methods; MultispectralMultispectral (MRI, PET, SPECT, Methods, fMRI); (MRI, PET, SPECT, Methods, fMRI); VoluminousVoluminous :50 MB of raw scanner data per 5 minute :50 MB of raw scanner data per 5 minute scan;Hundreds of MB of processed data per subjectscan;Hundreds of MB of processed data per subject•Sequencing of human genome = 3 billion nucleotides!–How does code translate into structure & function?–What parts of the code are switched “on” or “off” and when?–Algorithms to interpret the information of microarrays describing gene expression84OutlineNeed to also compare with genomic data to find which Gene is responsible: Going from phenotype to genomics and back you need tons of samples!Drug design = > data interoperability and sharing(e.g.,molecular biology + pharmacology)Multidisciplinary compute- intensive problems requiring deep domain knowledgeDoug Brutlag 2002Data Collection and workflow management - The Multiple Sclerosis Cataloguer Tool PrototypeShows two tasks for current user: user uploads requested file to finish task; below: overview of available data (subset of subject scan/trace database)Can show additional information by pointing at item…Point at specific scan to get more info on that scan……New window with thumbnails of scans organized by date and type of scan…Simple comparison between different sessions and scan typesWhat if people do not wish to share original data?SCENS: Secure Content Exchange Negotiation System A Metadata-Based Negotiation System For Sharing Diverse Research Resourcesquery metadatametadata resultsrequest for datapreset conditionsaccept as-issuggest new conditionsUSER NEGOTIATION SYSTEMDATA OWNERaccept as-issuggest new conditionsnotify of agreementaccept agreementaccept agreementnotify of data exchangesend datasend feedback,usage reportsend new track record(update record)Human-Computer Interface Design ExampleNegotiation Conditions RegistrationSCENS 0.9: Data owners and requesters can only negotiate on two conditions: Price and Usage Period.Can not meet users’ other requirements or conditions (e.g., copyrights, etc.).DollarsDaysNew NCR Interface•SCENS 1.0: –Users can now define their own