FP00043869SUB1801S
DATA MANAGEMENT DATA COLLECTION THE DATABASE HAS ALREADY BEEN BUILT IN REDCAP BY CHOP. ALL UPDATES AND MAINTENANCE OF THE DATABASE WILL BE MADE BY UTAH DCC PERSONNEL. UTAH DCC PERSONNEL WILL REQUIRE APPROPRIATE PERMISSIONS WITHIN THE CHOP REDCAP INSTANCE. THE DATABASE WILL BE MAINTAINED IN A CHOP- HOSTED INSTANCE OF REDCAP. DATA VALIDATION THE UTAH DCC WILL BE RESPONSIBLE FOR SETTING UP AND MANAGING THE DATA VALIDATION SYSTEM. VALIDATION MAY BE ACCOMPLISHED THROUGH THE USE OF THE REDCAP DATA VALIDATION SYSTEM, THROUGH THE USE OF THE UTAH QUERY MANAGER APPLICATION, OR A COMBINATION OF THE TWO. THE INITIAL APPROACH WILL BE TO MANAGE DATA VALIDATION ENTIRELY THROUGH REDCAP BECAUSE 1) IT WILL NOT REQUIRE FREQUENT TRANSFERS OF DATA FROM CHOP TO UTAH, AND 2) IT WILL PROVIDE RESEARCH COORDINATORS WITH REAL-TIME FEEDBACK REGARDING INVALID, INCONSISTENT, OR INCOMPLETE DATA WHEN THE DATA IS ENTERED. IF THE INITIAL APPROACH IS INSUFFICIENT, OTHER APPROACHES, INCLUDING USE OF QUERY MANAGER, WILL BE CONSIDERED TO SUPPLEMENT THIS APPROACH. DEPENDING ON THE COMPLEXITY OF OTHER APPROACHES, ADDITIONAL EFFORT MAY BE REQUIRED. IT DATA TRANSFER PERIODIC TRANSFERS OF THE REDCAP DATA FROM CHOP TO THE UTAH DCC WILL BE NEEDED TO FACILITATE STATISTICAL ANALYSES. IF QUERY MANAGER IS USED FOR DATA VALIDATION, MORE FREQUENT TRANSFERS WILL BE NEEDED. THE UTAH DCC WILL IMPLEMENT A SECURE INFRASTRUCTURE FOR CHOP TO UPLOAD DATA EXPORTED FROM REDCAP. THE REDCAP DATA SHOULD BE EXPORTED IN FLAT FILE FORMAT, AND SUBSEQUENTLY TRANSFERRED TO THE UTAH DCC VIA SECURE WEB-BASED TRANSFER. IN THE EVENT THAT CHOP WOULD LIKE TO AUTOMATE THE DATA UPLOADING, A SECURE FTP {SFTP) WILL BE PROVIDED. THE UTAH DCC IT WILL OFFER FULL TECHNICAL SUPPORT AND PARTNERSHIP WITH CHOP IT TO OPERATIONALIZE THIS PROCESS. BI TEAM REPORTING, SUCH AS THE CREATION OF REPORT CARDS, WILL BE DONE BY CHOP. BIOSTATISTICS MANUSCRIPTS THE UTAH DCC WILL PROVIDE STATISTICAL SUPPORT FOR UP TO 3 MANUSCRIPTS IN THE FIRST YEAR OF THE PROJECT AND UP TO 5 MANUSCRIPTS IN SUBSEQUENT YEARS. IN COLLABORATION WITH CHOP, UTAH WILL CREATE AND MANAGE A PROCESS FOR EFFICIENT COLLABORATION. PROPOSALS FOR MANUSCRIPTS WILL BE SUBMITTED USING THE UTAH DCC MARF TEMPLATE OR SIMILAR. UTAH WILL SCHEDULE CALLS TO DISCUSS AND REFINE PLANNED ANALYSES AND WRITE A DETAILED ANALYSIS PLAN FOR EACH MANUSCRIPT. UTAH WILL CONDUCT PLANNED ANALYSES AND PREPARE TABLES AND FIGURES FOR MANUSCRIPTS. UTAH WILL DRAFT THE STATISTICAL METHODS SECTION OF MANUSCRIPTS, AND WILL REVIEW THE FINAL MANUSCRIPT. UTAH WILL ALSO PROVIDE STATISTICAL QUALITY CONTROL FOR EACH MANUSCRIPT; THIS INCLUDES A SECOND STATISTICIAN INDEPENDENTLY PROGRAMMING DATASETS AND REVIEWING CODE USED FOR ANALYSIS. PROJECT MANAGEMENT PROJECT MANAGEMENT WILL BE PERFORMED BY CHOP. CHOP WILL SCHEDULE AND HOST MEETINGS BETWEEN THE UTAH AND CHOP TEAMS AS NEEDED FOR NON-STATISTICAL COLLABORATION. UTAH WILL SCHEDULE THEIR OWN INTERNAL MEETINGS AS NEEDED. UTAH WILL SCHEDULE AND MANAGE MANUSCRIPT DEVELOPMENT CALLS. CHOP WILL BE RESPONSIBLE FOR ALL SITE TRAINING AND ALL COMMUNICATION WITH SITES, INCLUDING ANY COMMUNICATION RELATED TO THE RESOLUTION OF DATA QUERIES. ALL SITE AND REMOTE MONITORING, IF ANY, WILL BE CONDUCTED BY CHOP. CHOP IS THE SINGLE IRB {MOST SITES RELY ON CHOP). CHOP PROJECT MANAGEMENT WILL HELP SITES WITH IRB SUBMISSIONS AND AMENDMENTS. UTAH WILL BE RESPONSIBLE FOR GETTING IRB APPROVAL FROM THE UTAH IRB FOR THEMSELVES, BUT WILL NOT BE INVOLVED WITH HELPING SITES WORK WITH THE CHOP SINGLE IRB OR THEIR LOCAL IRBS. CHOP WILL WRITE AND MAINTAIN THE STUDY PROTOCOL. REVIEW OF DATA USE AGREEMENTS THE UTAH DCC WILL SET UP A DATA USE AGREEMENT {DUA) WITH CHOP. UTAH WILL NOT BE REQUIRED TO SET UP DUAS WITH OTHER PARTICIPATING SITES. AS UTAH DCC PERSONNEL WILL BE ACCESSING PHI WITHIN CHOP'S REDCAP INSTANCE, AND AS THESE DATA WILL BE DOWNLOADED TO THE UTAH SERVERS, CHOP MAY NEED TO AMEND THEIR DUAS WITH ENROLLING SITES. DETERMINATION OF WHETHER AMENDMENTS TO CHOPS DUAS WITH EN
University Of Utah
Project Grant R01HL175433
$216.9k 10/1/24 FP00043869SUB1901S
DR. CHANDRASEKHAR NATARAJ, CO-INVESTIGATOR AT VILLANOVA UNIVERSITY, WILL BE PRIMARILY RESPONSIBLE FOR SUCCESSFUL EXECUTION OF AIM 3. HE WILL MENTOR AND WORK CLOSELY WITH HIS POSTDOCTORAL FELLOW, DR. DIETER BENDER AND A PH.D. STUDENT (TBD) TO DEVELOP INNOVATIVE AND EFFECTIVE ALGORITHMS THAT ENCAPSULATE NONLINEAR DYNAMICS, ADVANCED SIGNAL PROCESSING AND INNOVATIVE MACHINE LEARNING. DR. NATARAJ WILL ALSO BE RESPONSIBLE FOR COORDINATING PUBLICATIONS AND REPORTS. DR. DIETER BENDER, CURRENTLY SERVING AS A POSTDOCTORAL RESEARCHER AT VILLANOVA UNIVERSITY, HAS BEEN DESIGNATED TO WORK 50% ON THIS PROJECT. DRAWING UPON HIS EXTENSIVE BACKGROUND CHARACTERIZED BY NUMEROUS YEARS OF INDEPTH EXPERIENCE, DR. BENDER IS ENTRUSTED WITH THE FORMULATION AND DEVELOPMENT OF NOVEL COMPUTATIONAL ALGORITHMS AND TO CARRY OUT ALL DEVELOPMENT BY WORKING CLOSELY WITH A PH.D. STUDENT WITH A BS/MS DEGREE IN ENGINEERING. THESE SOPHISTICATED ALGORITHMS ARE INTENDED TO INTEGRATE NONLINEAR DYNAMICAL SYSTEM SIGNAL PROCESSING WITH CUTTING-EDGE MACHINE LEARNING METHODOLOGIES. IN ADDITION, DR. BENDER WILL SERVE AS THE WEEKLY POINT OF CONTACT FOR COMMUNICATION AND COLLABORATION WITH CHOP. THE WHOLE TEAM INCLUDING PI MORGAN AND CO-I NATARAJ WILL MEET FORTNIGHTLY THE RESEARCH TEAM AT THE VILLANOVA CENTER FOR ANALYTICS OF DYNAMIC SYSTEMS (VCADS) AT VILLANOVA UNIVERSITY WILL CONCENTRATE THEIR EFFORTS ON AIMS 1C AND 3. AIM 1C IS TO EVALUATE THE PHYSIOLOGIC RESPONSE TO EPINEPHRINE AS AN OUTCOME DISCRIMINATOR IN PEDIATRIC IHCA. OUR WORK WILL FOCUS ON DETERMINING THE CAPACITY FOR PULSE OXIMETRY PLETHYSMOGRAPHIC WAVEFORM (PPG) CHARACTERISTICS TO ACCURATELY DETECT DBP RESPONSIVENESS TO EPINEPHRINE DURING CPR. AIM 3 IS TO DEVELOP MACHINE LEARNING (ML) ALGORITHMS TO PREDICT EPINEPHRINE RESPONSE AND SUBSEQUENT OUTCOMES. WE WILL COLLECT EXTENSIVE DEMOGRAPHIC AND CLINICAL CHARACTERISTICS TO BUILD ML MODELS TO: 1) PREDICT THE PHYSIOLOGIC RESPONSE TO EPINEPHRINE AND THE ABILITY TO MEET DBP THRESHOLDS BASED ON PATIENT AND ARREST CHARACTERISTICS AND 2) PREDICT EVENT OUTCOMES AND RESPONSE TO FURTHER DOSES OF EPINEPHRINE AND OTHER THERAPIES BASED ON PATIENT AND ARREST CHARACTERISTICS AND THE RESPONSE TO THE INITIAL DOSE OF EPINEPHRINE. THE SUCCESSFUL COMPLETION OF THESE AIMS WILL GREATLY EXPAND OUR UNDERSTANDING OF THE PHYSIOLOGY OF PEDIATRIC IHCA AND HOW EPINEPHRINE DOSING AND RESUSCITATION IN GENERAL CAN BE TAILORED TO INDIVIDUAL PATIENTS AND WILL FACILITATE THE DESIGN OF FUTURE INTERVENTIONAL TRIALS OF PHYSIOLOGYDIRECTED CPR. OUR EXTENSIVE BACKGROUND SPANS OVER FOUR DECADES, DURING WHICH WE HAVE CONSISTENTLY INNOVATED ALGORITHMS TO UNCOVER HIDDEN RELATIONSHIPS IN COMPLEX SIGNALS BY ENCOMPASSING SEVERAL AREAS OF EXPERTISE, SUCH AS SIGNAL PROCESSING, NONLINEAR SIGNAL ANALYSIS, ADVANCED MACHINE LEARNING, STATISTICAL LEARNING, AND OPTIMIZATION. WE WILL LEVERAGE THIS RICH BACKGROUND IN CONJUNCTION WITH THE COMPREHENSIVE PEDIRES-Q DATASET OBTAINED FROM OUR EXTENSIVE HOSPITAL NETWORKS. OUR MAIN GOAL IS TO DEVELOP GROUNDBREAKING ALGORITHMS TO DETERMINEÂ FOR THE FIRST TIME- OPTIMAL VENTILATION PARAMETERS TO DRAMATICALLY IMPROVE CPR OUTCOMES FOR CHILDREN. MACHINE LEARNING WILL BE COAXED TO REVEAL COMPLEX MULTIMODAL AND DYNAMIC RELATIONSHIPS, WHICH ARE OUT OF REACH OF TRADITIONAL STATISTICAL METHODS. THE UNPARALLEL DEPTH OF KNOWLEDGE IN THIS DOMAIN, COMBINED WITH EXCLUSIVE AND UNIQUE DATA, PROVIDES A SIGNIFICANT ADVANTAGE IN THE PURSUIT OF THESE CHALLENGING AIMS. IN ADDITION TO THE USUAL COMPUTING EQUIPMENT, SUCH AS POWERFUL ENGINEERING WORKSTATIONS, WE ARE PRIVILEGED TO HAVE FULL ACCESS TO AUGIE (DR. NATARAJ, CO-I ON NSF GRANT THAT SECURED AUGIE), A RECENTLY INSTALLED STATE-OF-THEART HIGH-PERFORMANCE CLUSTER (HPC) WITH EXCEPTIONAL COMPUTATIONAL CAPABILITIES SUITED FOR ADVANCED RESEARCH INVOLVING COMPUTATIONALLY EXTENSIVE MACHINE LEARNING AND OPTIMIZATION ALGORITHMS. AUGIE HAS 13,3 TB RAM, 292 TB STORAGE SPACE, 1,856 TOTAL AMD EPYC SERIES CPU CORES, 2X TESLA A100 / 40 GB HBM2 MEMORY (GPU NODES
Villanova University In The State Of Pennsylvania
Project Grant R01HL175433
$101.1k 9/17/24