>>THANK YOU ACD MEMBERS FOR BEING WILLING TO GIVE US AN HOUR AND A HALF OF YOUR TIME ON THIS BEAUTIFUL AFTERNOON. AND WE NEED YOUR INPUT ABOUT SOMETHING THAT IS QUITE TIME SENSITIVE BECAUSE IT IS FUNDING THAT THE CONGRESS HAS PROVIDED FOR US TO WORK IN THE AREA OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING BUT IT'S FUNDING THAT NEEDS TO BE ALLOCATED BEFORE OCTOBER 1st. THIS BEING MAY WE COULD NOT WAIT UNTIL JUNE TO SEEK YOUR INPUT ON THIS WHICH IS WHY WE ASKED YOU TO JOIN THIS SPECIAL HOUR AND A HALF MEETING. THIS WAS ADVERTISED IN THE FEDERAL REGISTER. THIS IS A PUBLIC MEETING AND WE WILL HAVE A PRESENTATION FROM DR. TABAK ABOUT THE WORKING GROUP THAT IS PUTTING TOGETHER AN IDEA. THERE WILL BE DISCUSSION AND THEN A CONCEPT CLEARANCE PRESENTED BY DR. DINA. THAT IS WHAT WE'RE AIMING TO ACCOMPLISH. THIS IS PART OF THE INTENSE EXPERIENCES THAT WE'RE HAVING THAT HAS NOT LESSENED. AND I'M SURE IT HAS NOT FOR YOU EITHER. MANY OF US WORKING FLAT OUT BUT NOT NEGLECT OTHER THINGS. THIS IS AN EFFORT TO BRING TOGETHER OPPORTUNITIES WITH AIML ELECTRONIC HEALTH RECORDS AND HEALTH DISPARITIES AND PUT THAT INTO SOMETHING THAT WE THINK COULD BE PRETTY EXCITING BUT WE NEED YOUR INPUT. IN TERMS OF EVERYTHING ELSE GOING ON AT NIH I THINK WE'RE LOOKING AT WHAT IS HAPPENING IN THE VACCINE ARENA. FEELING ACTUALLY PRETTY GOOD ABOUT THE WAY IN WHICH SCIENCE HAS COME TO THE FLOOR HERE ESPECIALLY WITH THE MESSENGER RNA VACCINES WHICH WE BELIEVE ARE ABOUT TO BE APPROVED FOR INDIVIDUALS AS YOUNG AS 12 AS SOON AS THE FDA COMES FORWARD WITH THEIR DECISION ON THAT. REALLY ENCOURAGING REAL WORLD DATA FROM BOTH RITZ AND QATAR ABOUT HOW THE MRNA VACCINES ARE PERFORMING IN THOSE SETTINGS GIVING OVER 90% PROTECTION WHERE QATAR HAD ENOUGH CASES TO BE ABLE TO ASSESS THAT. SO FAR THE VACCINES ARE HOLDING THEIR OWN. WE DO WORRY ABOUT THE VARIANCE THAT ARE APPEARING AND WE WANT TO BE SURE WE DO EVERYTHING POSSIBLE TO REDUCE THEIR LIKELIHOOD BY REDUCING THE NUMBER OF PEOPLE WHO GET NEWLY INFECTED. WE'RE DEEPLY TROUBLED ABOUT WHAT IS HAPPENING IN INDIA AND THE NEW HORRIBLE RECORDS THAT ARE BEING BROKEN THERE IN TERMS OF OVER 400,000 CASES IN A SINGLE DAY AND OVER 4,000 DEATHS IN A SINGLE DAY. DOING EVERYTHING THAT THE GOVERNMENT CAN IN TERMS OF SHIPPING OXYGEN AND VENTILATORS AND PPE AND HOPEFULLY SOON VACCINES AS WELL. TO TRY TO HELP. AND OF COURSE THAT IS NOT THE ONLY HOT SPOT IN THE WORLD. -- REMINDED US AT OUR DIRECTOR'S MEETING THAT SOUTH AMERICA IS SUFFERING GREATLY AND PROBABLY READ MORE ABOUT BRAZIL BUT IT'S NOT THE ONLY SOUTH AMERICAN COMPANY THAT IS HAVING TROUBLE. MEXICO AS WELL. SO EVEN AS WE CAN SEE IN THE UNITED STATES SOME IMPROVEMENT IN OUR CIRCUMSTANCES AND PEOPLE ARE EVEN PREDICTING BY JULY WE MIGHT HAVE REALLY SEEN THESE CURVES ABOUT CASES AND DEATHS GOING DOWN QUITE LOW. THE REST OF THE WORLD IS NOT IN SUCH A POSITIVE PLACE AND AS CITIZENS OF THE WORLD WE NEED TO TAKE THAT PART ON WITH THE GREATEST SERIOUSNESS WORKING WITH OUR COLLABORATORS AND COLLEAGUES AND TONY AND I ARE SPENDING TIME IN THAT GLOBAL HEALTH ARENA AS WELL. OTHER THINGS TO MENTION, WE ARE INTO THE SEASON OF PRESIDENTIAL BUDGET ANNOUNCEMENTS AND CONGRESSIONAL APPROPRIATIONS HEARINGS. SO OUR HEARING IS COMING UP IN ABOUT THREE WEEKS IN THE SENATE AND THE HOUSE BACK TO BACK TO DAYS MAY 25-26. AND YOU'VE PROBABLY SEEN THE BIG BUZZ ABOUT NIH'S PRESIDENTIAL BUDGET. THERE IS A NEW ENTITY. A NEW DIVISION. NOT AN INSTITUTE. WE'LL HAVE TO CALL IT A DIVISION OR AUTHORITY OR SOMETHING LIKE THAT WHICH IS CALLED ARPAH. WHICH BOROUGHS FROM THE DEFENSE DEPARTMENT ALL OF THE MYSTIQUE AND PRODUCTIVITY. TO GIVE US HERE AT NIH THE KIND OF RAPIDLY MOVING NIMBLE EFFORT WHICH WE'VE LEARNED CAN BE INCREDIBLY VALUABLE FOR THINGS LIKE VACCINES AND DIAGNOSTICS. BUT WITH THE PROPOSED $6.5 BILLION BUDGET IN FY22 AND BEYOND. WE'RE EXCITE ABOUT THIS. WE HAVE ALL KIND OF IDEAS ABOUT WHAT WE CAN DO WITH THAT NIMBLE AUTHORITY BUT OF COURSE WE WANT TO BE CLEAR THAT MOST OF WHAT WE DO WOULD NOT FIT WELL INTO THIS. THIS IS MORE FOR PARTICULAR PROJECTS THAT ARE TECHNOLOGICALLY CHALLENGING AND HAVE TIMETABLES AND MILESTONES THAT CAN BE DEFINED. WE'LL SEE HOW THIS PLAYS OUT AS CONGRESS GETS THEIR MIND AROUND THIS POTENTIAL AND WHAT THEY WANT TO DO TO SUPPORT IT. WHICH WE HOPE THEY WILL WANT TO DO. OTHER THINGS WE'VE BEEN TALKING TO CONGRESS ABOUT AND I HAD A MARATHON HEARING LAST WEEK IS ABOUT THE LONG COVID SITUATION WITH INCREASING RECOGNITION THAT AS MANY AS 30% OF PEOPLE WHO GET INVESTED DO NOT GET BETTER IN A TWO WEEK PERIOD. SO HERE IS A BIG CHALLENGE FOR US AT NIH TO TRY TO PUT TOGETHER SOME VERY LARGE COHORT STUDIES. WHO IS AT RISK AND HOW DO WE INTERFERE WITH THAT OUTCOME AND HOW TO TREAT FOLKS WHO MAY BE WEEKS OR MONTHS LATER STILL NOT BACK TO NORMAL. I'M HEARING A GOOD DEAL OF BACKGROUND NOISE AND I'M HOPING THAT IS LARRY TABAK. >> HE IS HAVING AUDIO ISSUES. >> YOU'RE REELING BREAKING UP. I DON'T KNOW WHAT KIND OF MICROPHONE YOU'RE USING. [ AUDIO DIFFICULTIES ] >> LARRY YOUR AUDIO IS SO FRACTURED IT'S REALLY HARD TO HEAR. IS SOMEBODY SENDING LARRY THE PHONE NUMBER? >> JOHN IS DOING IT. SO I UNDERSTOOD THAT COURTNEY MIGHT HAD BEEN READY TO GO THROUGH SOME OF THE LOGISTICS ABOUT GETTING THIS STARTED IF LARRY WAS NOT ABLE TO JOIN US AT THE BEGINNING. COURTNEY ARE YOU ON? >> I AM, FRANCIS. LET ME START AND BUY LARRY A LITTLE BIT MORE TIME TO GET THE TECHNOLOGY FIGURED OUT. GOOD AFTERNOON, AND THANK YOU FRANCIS. I'LL COVER A FEW OF THE ADMINISTRATIVE ITEMS. AGAIN THANK YOU ALL FOR BEING HERE TODAY. I REALLY APPRECIATE THAT YOU'RE HERE. AS FRANCIS MENTIONED THIS IS AN OPEN MEETING SO THERE ARE NO ISSUES IN TERMS OF YOUR CONFLICT OF INTEREST FORM. THOSE WON'T NEED TO BE COMPLETED. I'LL START WITH ROLL CALL SO WE CAN HAVE THAT FOR THE RECORD AND IF YOU CAN JUST SAY HERE AND SHOW YOUR FACE THAT WOULD BE WONDERFUL. SHELLEY BURGER. >> I'M HERE. >> ROBERTA DIAZ. >> HERE. >> WENDY CHAPMAN. >> I'M HERE. >> ANNIE -- ANN CHURCH LAND. FRANCIS CUSS. >> I'M HERE. >> REBECCA. >> I'M HERE. >> MARK. >> HERE. >> DAVID BLAZER. >> HERE. >> JAMES HILL DRESS. >> PRESENT. >> CHRISTINA JOHNSON. >> HERE. DANA KATABI. >> HERE. >> ALEXA KIMBALL. >> HERE. >> JUDITH KIMBALL. >> HERE. >> SPERO MANSON. >> HERE. >> LORI WILSON. >> I'M HERE. THANK YOU. >> BARBARA WALLS. OKAY. THANK YOU VERY MUCH. SO I'M GOING TO GO AHEAD AND -- >> SORRY. >> THANK YOU BARBARA. GLAD TO HAVE YOU ON-BOARD. >> BEFORE MOVING ON LET'S TALK ABOUT A FEW MORE ADMINISTRATIVE OPPORTUNITIES. FIRST BECAUSE THIS IS AN OPEN MEETING IT IS BEING ACCESSIBLE BY THE NIH VIDEO CAST. THIS CAN BE FOUND ON -- [ AUDIO DIFFICULTIES ] AND IS BEING RECORDED. ON THE MEETING PAGE. A NOTICE HAS BEEN PUBLISHED IN THE FEDERAL REGISTER AS OF APRIL 26th AND THIS IS WHERE WE WERE GOING TO PAUSE TO GIVE LARRY AN OPPORTUNITY TO START HIS SLIDE. FRANCIS I DON'T KNOW IF YOU HAVE ANY MORE UPDATES YOU WANT TO PROVIDE TO THE GROUP. >> NO. I THINK I WAS JUST TRYING TO SET THE TABLE HERE WHILE WE ARE WAITING FOR LARRY TO APPEAR. THANK YOU COURTNEY. LARRY, HAVE YOU BEEN ABLE TO CONNECT UP IN A FASHION WHERE WE CAN HEAR YOUR AUDIO? THANK YOU ALL VERY MUCH FOR PARTICIPATING TODAY AND AS I HOPE YOU HAVE BEEN TOLD THIS IS SORT OF TWO PARTS. THE FIRST IS TO FORMALLY PROVIDE THE COMMUNITY AND OF COURSE THE MEMBERS OF THE ACD WITH A REPORT FROM OF A WORKING GROUP OF THE ACD ON ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING APPROACHES TO WORKING WITH ELECTRONIC MEDICAL RECORDS FOR RESEARCH PURPOSES AND THEN IN THE SECOND PART AFTER WE HAVE DISCUSSION ABOUT WHAT I'M GOING TO SHARE WITH YOU IN THE SECOND PART WE WILL HAVE A CONCEPT CLEARANCE WHICH THOSE OF YOU WHO HAVE BEEN ON THE NATIONAL ADVISORY COUNCILS OF INSTITUTES AND CENTERS KNOW IS THE VERY FIRST STEP IN THE WAY NIH TELLS THE COMMUNITY AND STAKEHOLDERS OF AN INTEREST AND A POTENTIAL FUTURE FUNDING ACTIVITY. SO WITH THAT LET ME HOPEFULLY ADVANCE THE NEXT SLIDE. THIS IS GOING TO FIGHT ME ALL THE WAY. THERE WE GO. YOU'RE ADVANCING THE SLIDES FOR ME. THANK YOU. SO WE COULD JUST TO GIVE BACKGROUND BACK IN 2019 WE CONVENED THE FIRST WORKING GROUP OF THE ACD ON ARTIFICIAL INTELLIGENCE AND AS YOU CAN SEE FROM THIS DISPLAY WITH THE EXCEPTION OF ME EVERYBODY ELSE ON THIS COMMITTEE IS REALLY A TRUE EXPERT IN THE FIELD OF AI. IF I COULD HAVE THE NEXT SLIDE, PLEASE. AND THE SUMMARY OF WHAT THAT WORKING GROUP THE FIRST WORKING GROUP REPORTED OUT ON WAS THAT THERE IS ENORMOUS OPPORTUNITY FOR USE OF ARTIFICIAL INTELLIGENCE IN BIOMEDICAL RESEARCH APPLICATIONS. 7 AND THAT TO REACH THE FULL POTENTIAL WE CANNOT DO THINGS THE WAY WE'VE ALWAYS DONE THEM AND ONE KEY POINT WAS THAT WE NEED NEW DATA GENERATION PROJECTS. AND THOSE IN THE FIELD OF ARTIFICIAL INTELLIGENCE MACHINE LEARNING REITERATED OVER AND OVER AGAIN THAT THE BEST WAY TO ATTRACT THE RIGHT PEOPLE IS WITH THE RIGHT DATA. AND THEN FINALLY ALTHOUGH A CRUCIALLY IMPORTANT POINT AND ONE THAT WE'LL COME BACK TO LATER ON IS THE TIME TO INVEST IN ETHICS IS NOW. BECAUSE THE LONGER WE WAIT THE LONGER WE IGNORE THE ETHICAL DIMENSIONS OF THIS THE BIGGER THE HOLE WE WILL HAVE DONE FOR OURSELVES. SO IT FALLS INTO THREE SPEARS. * DATA COLLECTION AND WHEN TO USE IT AND NOT TO. A WAY TO ATTRACT THE BEST PEOPLE TO THE FIELD AND THEN AN ETHICAL OVERLAY FOR ACCOUNTABILITY AND INFORMED REPRESENTATION IN ALL OF THE EFFORTS THAT WE DO. OKAY. SO NOW I'M GOING TO TAKE YOU TO THE SECOND ACD WORKING GROUP ON ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING. AND AS YOU WILL SEE FROM LOOKING AT THE MAINS THERE WAS SOME CARRY OVER FROM THE FIRST WORKING GROUP INCLUDING MEMBERS OF THIS NATIONAL ADVISORY COUNCIL BUT WE DID ADD ADDITIONAL EXPERTS BECAUSE OF OUR PARTICULAR INTEREST IN EXPLORING THE SPACE OF HOW AI AND ML APPROACHES COULD BE USED TO ENHANCE THE ABILITY OF USING ELECTRONIC MEDICAL RECORDS FOR RESEARCH PURPOSES AND IN ADDITION TO THESE EXTERNAL MEMBERS THEY WERE JOINED BY A NUMBER OF USG INDIVIDUALS -- YOU WILL SEE DISPLAYED THE NIH MEMBERS OF THIS TEAM INCLUDING OF COURSE THE NIH DIRECTOR. IF I COULD HAVE THE NEXT SLIDE PLEASE. AND SO MY PURPOSE IS TO QUICKLY SUMMARIZE WHAT THIS SECOND WORKING GROUP CAME UP WITH AND AS WE HAVE MEMBERS OF OUR ADVISORY COUNCIL WHO ARE ON THIS WORKING GROUP I WILL CALL THEM AT THE END TO FILL IN ANY GAPS OR MAKE ANY MODIFICATIONS. THE CHARGE TO THE WORKING GROUP WAS TO IDENTIFY THE UNIQUE RESEARCH OPPORTUNITIES FOR NIH TO APPLY RESOURCES IN A PRACTICAL WAY TO MAKE ELECTRONIC HEALTH RECORDS USABLE IN THE RESEARCH SETTING. AND THAT TO DO THIS WE NEEDED TO IDENTIFY THE ELECTRONIC HEALTH RECORD RESEARCH CHALLENGES THAT THESE APPROACHES OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING COULD HAVE THE GREATEST IMPACT ON. WE ALSO WANTED TO DETERMINE THE BARRIERS TO THE WIDESPREAD USE AND DEPLOYMENT OF AI AND ML CAPABILITY AND WHAT COULD NIH DO TO SUPPORT THESE TO OVERCOME THE BARRIERS AND WE WANTED TO IDENTIFY THE PARTNERS THAT WOULD BE NEEDED TO HELP SCALE ANY OF THESE CAPABILITIES TO THE POINT WHERE THEY WOULD BE PRACTICAL -- AND USABLE. AMONG THE CHALLENGES THAT PEOPLE THOUGHT WERE VERY, VERY IMPORTANT WAS FIRST AND THIS -- OR SEVERAL MEMBERS -- CERTAINLY WEIGHED IN ON THIS. PARTICULARLY -- BB BEGAN THE LINE OF DISCUSSION. TO START BY ADDRESSING THE NEEDS OF MARGINALIZED COMMUNITIES. IF THE COMMUNITIES REMAIN AN AFTER THOUGHT THE SAME INEQUITIES AND DISEASE BURDEN AND HEALTH CARE WILL BE REPEATED. AND IN PARTICULAR THERE WAS A VERY, VERY STRONG RECOMMENDATION TO ADDRESS THE SIGNIFICANT ERRORS, THE SIGNIFICANT GAPS AND THE RACIAL AND GENDER INEQUITIES IN ELECTRONIC HEALTH RECORD DATA IN GENERAL AND OF COURSE MANY OF YOU UNDERSTAND THAT IT IS NECESSARY BY USING PROBLEMATIC DATA FOR MODELS WHEN YOU THEN AMPLIFY THE PROBLEM. YOU WOULD AMPLIFY THE CONCERN. AND SO THE FOUNDATION THAT THE MODELS ARE BEING BUILT MUST BE ADDRESSED. AND SO JUST ONE EXAMPLE THAT ONE OF OUR MEMBERS INDICATED WAS THAT THERE WAS ANALYSIS OF 50,000 COVID-19 PATIENT HALF OF WHICH WERE AFRICAN-AMERICAN BUT NO RIGOROUS METHODS TO ANALYZE THE DATA. THAT WAS LOST AND WE WERE NEVER ABLE TO CONSIDER THAT. SO THE CHALLENGES AND I'M SURE SOME OF YOU WILL RESONATE WITH THESE IS THAT WE NEED TO FIGURE OUT COMMUNITY ENABLED REAL WORLD EFFORTS TO ENCOURAGE SOLICITATIONS OF CLINICS AND SYSTEMS. WE CANNOT GO TO THE TERTIARY CARE CENTERS AND ENGAGE AMOUNT THE COMMUNITY LEVEL. WE HAVE TO THINK ABOUT DIFFERENT GROUPS OF PEOPLE THAT HERETOFORE HAVE BEEN LEFT OUT. WE NEED TO CREATE CONSORTIAN TO MAKE SURE THAT THEY PARTICIPATE TO ENRICH AND SHARE THE DATA. IT WOULD BE HELPFUL IF WE HAD A TRIALS NETWORK -- INDIVIDUALS POINTED OUT THAT THERE ARE CERTAIN DENSE DATA FOR EXAMPLE MEDICAL IMAGES. THAT WE NEED TO CONSIDER AS WELL. AND THEN FINALLY THAT WE HAVE GOT TO FIGURE OUT HOW TO IMPUTE OR FILL IN MISSING DATA AND WE HAVE TO FIGURE OUT HOW TO LINK SOCIAL DETERMINANTS OF HEALTH WITH ACTUAL OUTCOMES. SINCE RARELY DOES AN ELECTRONIC HEALTH RECORD ADDRESS THAT TYPE OF INFORMATION. NEXT PLEASE. SO THE BARRIERS ARE MANY. AND THIS IS NONTRIVIAL. THESE ARE REAL SIGNIFICANT ISSUES. INFORMATION INFRASTRUCTURE THAT IS NEEDED INCLUDES NOT ONLY ELECTRONIC HEALTH RECORDS BUT ALSO THE NETWORK OF LARGE SCALE DATA REPOSITORIES AND IMAGING STUDIES, CLINICAL DATA CAPTURE IS HUMAN INTENSIVE. IT'S BIASED. IT'S INCOMPLETE AND SPARSE. ELECTRONIC HEALTH RECORDS ALONE ONLY PROVIDE A PARTIAL PICTURE OF HEALTH AND INDEED THE HEALTH CARE EXPERIENCE. THERE IS CLEARLY AN INCONSISTENT QUALITY OF THE CONTENT INTEGRITY OF DATA AND THEREFORE THE RESEARCH USEFULNESS OF ELECTRONIC HEALTH RECORDS AND NONTRIVIAL BUT PRACTICAL ACCESSIBILITY IS A VERY DIFFICULT AND SERIOUS ISSUE. NEXT PLEASE. AND SO TO ADDRESS THESE BARRIERS THIS ACD WORKING GROUP SUBJECTED THAT ONE APPROACH WOULD BE TO CREATE A BODY THAT COULD BE THE TRANSLATOR, TAKE IN DATA, IDENTIFY IT AND THEN PROVIDE THE DATA TO END USERS. ANOTHER SUGGESTION MADE WAS TO CREATE AN INFRASTRUCTURE TO FACILITATE THE DESIGN AND TESTING OF ALGORITHMISMS AND THEN USE THEM TO COMPARE ACROSS INSTITUTIONS. A THIRD EXAMPLE TO ADDRESS THE BARRIERS WAS TO INVEST IN EARLIER STAGE RESEARCH. THE TYPE OF STUFF THAT PRIVATE INDUSTRY IS LOATHED TO FUND AND THEN TO ADDRESS THE ISSUE OF OWNERSHIP. CLARITY IS NEEDED. WITH REGARD TO OWNERSHIP AND SHARING OF THE ML PRODUCED AFTER PUBLIC SHARED DATA ARE EMPLOYED. AND TO DO THIS REGARDLESS OF WHAT SUITE OF APPROACHES THAT WE ADOPT WE'VE GOT TO DEFINE WHO THE PARTNERS WOULD BE THAT WOULD HELP US. SCALE THESE CAPABILITIES. AND THIS IS GOING TO INCLUDE THE AI INDUSTRIES, COMPUTING PARTNERS. RESEARCHERS BUT IT IS ABSOLUTELY REQUIRED THAT WE LOOK AT NONTRADITIONAL PARTNERS. PEOPLE AT THEIR COMMUNITY LEVEL. PEOPLE ATKINS STEWSES THAT SERVE THE UNDERSERVED. PEOPLE IN MARGINALIZED PARTS OF OUR SOCIETY. NOW THE NIH OF INVESTMENT ALSO NEEDS TO HAD MEANINGFUL COMPLIMENTS THAT WHAT IS ALREADY BEING DONE AND THEN FINALLY THE NOTION THAT WHATEVER WE DO WE HAVE TO THINK ABOUT SUSTAINABILITY THROUGH STRATEGIC PUBLIC/PRIVATE PARTNERSHIPS. SO HERE ARE SOME OF THE IDEAS THAT EMERGED AND YOU WILL SEE THAT IN THE CONCEPT CLEARANCE WE HAVE TAKEN ADVANTAGE OF SOME OF THESE IDEAS. SO FIRST WAS TO SUPPORT AIML WORK TO ADDRESS THE CHALLENGES OF HEALTH DISPARITIES AND HEALTH INEQUITY AND MINORITY HEALTH. THE SECOND WAS TO DEVELOP THE INFRASTRUCTURE AND TRAINING OF DIVERSE COMMUNITIES. THE THIRD WAS TO CATALYZE ACCESS TO HIGH QUALITY DIVERSE DATASETS. THEIR QUALITY OF COURSE DEPENDED UPON THE DIVERSITY OF THE DATASET AND THEN LAST BUT VERY, VERY IMPORTANT, THE NEEDS TO ADDRESS THE ERRORS AND BIASES IN ELECTRONIC HEALTH DATA AND TO TACKLE THE PROBLEM OF HOW BEST TO LINK EHR DATA WITH DATA THAT DEFINES AND DESCRIBE SOCIAL DETERMINANTS OF HEALTH. SO A SERIES OF THINGS THAT NEED TO BE IMPROVED AND HOW CAN WE HELP BEST DO THIS. HOW DO WE IMPROVE THE QUALITY OF THE CONTENT IN EHRs. THROUGH AI STRATEGIES AND OTHER COMPUTATIONAL APPROACHES TO DEBUT ASKING DATA TO AVOID THE PERPETRATION OF BIASED UNDERSTANDING OF THE PATIENT EXPERIENCE. HOW DO WE INCREASE RESEARCH USEFULNESS BY IDENTIFYING THE HIGH PRIORITY PROBLEMS BY DOING MODELING BY DOING VALIDITY TESTING, VERIFICATION AND OF COURSE INTEGRAL IS THE SECURITY AND EQUITY AND PRIVACY ISSUE. WHAT ABOUT THE INTEGRITY OF THE DATA. THE NEED TO USE THE STANDARDS AND THE CAPTURE OF CLINICAL DATA AND TO DEPLOY METHODS. AGAIN A NONTRIVIAL ISSUE AND IN TERMS OF CLINICAL CARE THINKING ABOUT THE INTEGRATED DATA MODELS THAT WOULD ENRICH OUR UNDERSTANDING OF HEALTH BEYOND THE DATA THAT IS JUST FOUND IN THE BASIC ELECTRONIC HEALTH RECORD A METHODOLOGY THAT SUPPORTS TRANSPARENCY AND TRANSLATION AND WHATEVER MODEL THAT IS ACAPTAINED WILL FIT FOR THE PURPOSE REQUIRED. SO WITH THAT I WOULD LIKE TO TURN TO THOSE MEMBERS OF THE ACD WHO SERVED ON THIS ACD WORKING GROUP AND MY MEMORY WILL FAIL ME BUT I KNOW DAVID GLACIER AND DINA KATABI WERE AMONG THOSE THAT DID. SO I DON'T KNOW IF EITHER OF WOULD LIKE TO START AND IF I'VE LEFT ANYBODY OUT PLEASE DO REMIND ME. DINA, WOULD YOU LIKE TO START? >> SURE. THANKS, LARRY FOR THE GREAT DESCRIPTION. SO FIRST I WANT TO START WITH A COMMENT BECAUSE IT COMES UP QUITE OFTEN WHEN WE TALK ABOUT DENSE DATA AND WHAT DATA OTHER THAN EHR. PEOPLE SAY IMAGES AND MEDICAL IMAGES ARE ONE THING BUT THERE ARE MANY OTHER TYPES OF DATA SUCH AS EEG DATA -- -- PSG DATA. THERE ARE MANY OTHER TYPES OF DENSE DATA THAT EXIST OUT THERE IN THE WORLD NOT JUST MEDICAL IMAGES. I JUST DON'T WANT THAT TO BE MISSED. SO OF COURSE YOU CAPTURED IN YOUR PRESENTATION THE VAST MAJORITY AND ALMOST TOUCHED ON EVERY SINGLE SUBJECT AND DISCUSSION THAT WE HAD IN BOTH CASES. JUST ONE THING THAT I WANT TO EMPHASIZE IS IT IS REALLY IMPORTANT ALSO NOT TO MISS THE INTEGRATION OF MACHINE LEARNING WITHIN THE HEALTH UNIT ALSO AND THE DISEASE AREAS THAT ARE UNDER NIH. AND NOT TO TAKE THAT OUT OF THE CONTEXT TO ENSURE THAT WE HAVE VERY -- INTERACTION BETWEEN MACHINE LEARNING AND RESEARCHERS AND COLLABORATION ON COMPREHENSIVE AND COMPLIMENTARY ASPECTS THAT DIRECTLY ADDRESS THAT. >> THANK YOU SO MUCH. YOU SAID THAT FAR BETTER THAN I DID OR COULD HAVE. LET ME TURN TO DAVID GLACIER WHO IS ALSO A MEMBER OF THIS WORKING GROUP AND WE DO BEGIN TO HAVE SOME QUESTIONS AS WELL FROM OTHER MEMBERS OF THE ACD. DAVID, WOULD YOU LIKE TO ADD ANYTHING PLEASE. >> YES, THANK YOU AND SECOND WHAT DINA SAID ABOUT THE SLIDES AND THE PRESENTATION. AND I THINK THAT IT IS IMPORTANT TO THINK ABOUT ALL OF THOSE AND THEN AND THIS IS THE PROCESS THAT WE'RE IN THE MIDDLE OF RIGHT NOW AND THEN TO DECIDE WHICH PARTS OF THOSE ARE MOST AMENABLE TO NIH LEVERAGE POINTS. VERSUS ENABLING. AND I LIKE THE DIRECTIONS THAT ARE BEING HIGHLIGHTED. I THINK THAT ONE OF THE INTERESTING THINGS WILL BE REALIZING WHICH OF THESE OPPORTUNITIES TO MAKE THINGS BETTER ARE GOOD ENOUGH, WHICH ONES ARE URGENTLY BLOCKING PROGRESS. RATE LIMITING IN THE NEAR TERM AND WHICH THINGS WILL BE RATE LIMITING AND THEREFORE ARE GOOD OPPORTUNITIES FOR MULTI-YEAR INVESTMENT WHERE WE SHOULD BE LOOKING FOR OPPORTUNITIES TO FUND THINGS THAT WILL MATTER IN THREE OR FIVE YEARS VERSUS WHERE WE NEED TO BE FUNDING THINGS THAT WILL MATTER IN THREE OR FIVE MONTHS BECAUSE IT WILL ALLOW MANY MORE PEOPLE TO TAKE PART AND MY LAST THOUGHT WILL BE THE FIRST THOUGHT THAT YOU SUMMARIZED FROM THE 2019 REPORT WITH DINA AND ME AND OTHERS WHICH IS SHOW ME THE DATA. AND I LIKE THAT THAT COMES THROUGH LOUD AND CLEAR HERE AS WHAT DOES IT TAKE TO WHEN WE'RE TALKING ABOUT DATA FROM THE EHR WHICH IS A VERY BURBELOY SET OF TRIBUTARIES OF DATA THAT WE WANT TO COLLECT. HOW DO WE BUILD THAT COLLECTION SYSTEM TO EQUITABLY AND WITH GOOD ENOUGH FIDELITY OF CAPTURE START THAT FLOWING. SO THAT OVER TIME WE CAN IMPROVE THE FIDELITY OF CAPTURE SO I LIKE THAT EMPHASIS. I THINK THAT IS ONE OF THE NEAR TERM ONES THAT I'M HAPPY TO BE SEEING PUSHED ON. >> THANK YOU SO MUCH. NOW HAVE I LEFT ANY MEMBER OF THE ACD WHO WAS ON THE WORKING GROUP OUT. IF I DID I DO APOLOGIZE BUT IDENTIFY YOURSELF NOW AND I WILL CALL ON YOU NEXT. GOOD. SO I DID GET IT RIGHT. LET'S GO TO THE CHAT BOX THEN. SHELLEY YOU HAVE A QUESTION, PLEASE. >> YES. THANK YOU. THAT WAS WONDERFUL AND I THINK THIS IS AN AMAZING EFFORT FOR THE ELECTRONIC HEALTH RECORDS. I WAS WONDERING I DIDN'T SEE ANY EXPLICIT MENTION OF BIOBANKING AND HOW IMPORTANT IT'S GOING TO BE TO HAVE TO SUPPORT INCREASED BIOBANKING OF DIVERSE POPULATIONS. THAT IS GOING TO FEED INTO THE REALLY IMPORTANT BASIC RESEARCH THAT NEEDS TO BE DONE TO UNRAVEL SPECIFIC UNDERLYING HEALTH -- GENETIC EPIGENETIC AND OTHER EXPLANATIONS FOR HEALTH DISPARITIES. I WONDER IF YOU COULD SPEAK TO THAT A BIT. >> SO YOU ARE ABSOLUTELY CORRECT IN POINTING OUT THE IMPORTANCE. THIS PARTICULAR WORKING GROUP REALLY DID FOCUS ON HOW ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING CAN BE EMPLOYED BEING ABLE TO ADDRESS SOME OF THE ISSUES. SO I VIEW THE BIOBANKING ISSUES AS COMPLIMENTARY TO AND CERTAINLY IN OTHER NIH ACTIVITIES FOR EXAMPLE ALL OF US THERE IS A CONCERTED EFFORT IN THIS SPACE. FRANCIS COLLINS, I DON'T KNOW IF YOU WOULD LIKE TO COMMENT ABOUT THE ALL OF US -- APPROACH TO BIOBANKING BECAUSE THEY HAVE SUCH A GREAT DIVERSE POPULATION. BUT I KNOW THEY HAVE THOUGHT ABOUT THIS A GREAT DEAL. >> THEY HAVE. AND WHAT WE'RE TALKING ABOUT HERE IN TERMS OF THE ARTIFICIAL INTELLIGENCE MACHINE LEARNING APPROACH TO ELECTRONIC HEALTH RECORDS IS ALSO NEAR AND DEAR TO ALL OF US. I KNOW JOSH DENNY IS LISTENING IN TO THIS CONVERSATION. I SAW HIS ICON APPEAR EARLIER AND WITH ALL OF US NOW ROUGHLY 300,000 FULLY ENROLLED PARTICIPANTS ALL OF WHO HAVE BIOSAMPLES IN THIS REMARKABLE BIOBANK THAT THE MAYO CLUB IS RUNNING. THIS IS AN OPPORTUNITY TO DO ALL OF THE THINGS THAT I THINK SHELLEY IS TALKING ABOUT. MAKING SURE THAT THE BIOBANKING FITS TOGETHER. BUT WHAT WE'RE TRYING TO AND LARRY CORRECT ME IF YOU DON'T AGREE. WHAT WE'RE TRYING TO DO IS WE'RE ASKING YOU TO LOOK AT IT IS MUCH MORE FOCUSED ON THE ELECTRONIC HEALTH RECORD AS THE ENTITY THAT DESPERATELY NEEDS MORE INNER OPERABILITY AND MORE OPPORTUNITIES FOR MACHINE LEARNING TO HAPPEN BUT TRYING TO BE CAREFUL THAT WE DON'T GO WITH THE PLACES THAT HAVE ALL OF THE RESOURCES THAT HAVE ALREADY DONE SOME WORK IN THE SPACE AND THEN CONTINUE THE PRACTICE OF HAVING HEALTH DISPARITIES IGNORED OR MAYBE MADE WORSE. HENCE THE FOCUS THAT LARRY PUT FORWARD IN THE PRESENTATION AND TO THE DEGREE THAT THAT LINKS UP WITH BIOBANKING IS GREAT BUT THE FOCUS INTENDS TO BE ON THE ELECTRONIC HEALTH RECORD PROBLEM WHICH IS MASSIVE ENOUGH FOR OUR $50 MILLION. >> THANK YOU. >> THANK YOU. OKAY. THE NEXT UP IS WENDY CHAPMAN WHO'S A BONE IDENTIFIE A -- BONA FIDE EXPE RT. WENDY, YOU HAVE THE NEXT QUESTION, PLEASE. >> THANKS, LARRY. SO THE FOCUS SEEMS RIGHTLY TO BE ON THE RESEARCH USING DATA WHICH IS THE FIRST PART OF SOLVING THE PROBLEM OF AI. AS ALGORITHMS ARE TRAINED THERE ARE OPPORTUNITIES TO IMPLEMENT AI INTERVENTIONS. WILL THERE BE ANY FUTURE WORK ON HOW TO IMPLEMENT AI IN DIVERSE SETTINGS BECAUSE THAT IS A LARGELY UNSOLVED PROBLEM AND CRITICAL TO THE USEFULNESS OF AI. >> THANK YOU. THIS WAS NOT A SET UP FOLKS. STAY TUNED FOR THE SECOND PART OF WHAT WE'RE GOING TO TALK ABOUT THIS AFTERNOON. AS I'VE SHARED WITH YOU WHAT THIS WORKING GROUP PRODUCED AND I'VE TRIED AGAIN WITH THE HELP OF MY COLLEAGUES TO PORTRAY WHAT THE DELIBERATIONS WERE ABOUT IN THE SECOND PART OF TODAY'S MEETING WE WILL MY COLLEAGUE DR. DINA -- WILL PRESENT TO YOU A CONCEPT CLEARANCE OR AN INITIATIVE WHICH I THINK WILL VERY MUCH SPEAK TO THE ISSUE THAT YOU JUST RAISED. OKAY. BUT THANK YOU. 7 OKAY. NEXT UP IS ROBERTA. >> OKAY. THANKS, LARRY. QUICK TWO POINTS. ONE QUESTION IS IT SOUNDS LIKE YOU HAVE THIS CUED UP. PLANS FOR A PROOF OF CONCEPT TO ADDRESS THESE BARRIERS WHETHER THERE ARE THOSE IDENTIFIED AND STRATEGIES. >> YES. WE'RE CUEING THIS UP FOR TWO PURPOSES. ONE, TO GET THE FEEDBACK FROM OUR NATIONAL ADVISORY COUNCIL. AND SECONDLY WE DON'T OFTEN USE THIS COUNCIL FOR CONCEPT APPROVAL BUT IT WAS JUST A VERY CONVENIENT TIMING ISSUE AND BECAUSE OF SOME OF THE EXPERTISE ON THE COUNCIL WE THOUGHT IT WOULD BE A PERFECT OPPORTUNITY BUT YES. WE WILL IN THE NEXT PART HOPEFULLY SKETCH OUT FOR YOU SOME WAYS OF APPROACHING THIS AND WOULD WELCOME YOUR FEEDBACK ON IT BECAUSE WE ARE AT THE ABSOLUTE -- STAGE OF THINKING AND WE WANT TO REFINE AS WE LEARN FROM YOU AND OTHER EXPERTS GOING FORWARD. >> THANK YOU AND I WAS VERY EXCITED TO SEE THAT THE TOPIC IS INCLUDING AI ON ELECTRONIC MEDICAL RECORDS BECAUSE AT THE RECENT ALZHEIMER'S SUMMIT ONE OF THE CONSISTENT GAPS IN OPPORTUNITIES REALLY REVOLVED AROUND ACCESS TO DATA AND THE ABILITY TO INTERROGATE AND LEARN FROM THAT DATA AND PARTICULARLY IN DIVERSE POPULATIONS SO IT'S A MAJOR EMPHASIS FOREGOING FORWARD IN THE ALZHEIMER'S SUMMIT SPACE. >> INDEED. THANK YOU FOR SHARING THAT. THANK YOU. OKAY. NEXT FRANCIS CUSS PLEASE. >> THANK YOU VERY MUCH. REALLY FASCINATING LARRY. THANK YOU. IS IT LIKE LIKE OUT OF THIS OR SOME OTHER PROJECT THAT YOU COULD DEVELOP A SET OF STANDARDS THAT COULD BE APPLIED TO ANY DATASET THAT WOULD ENSURE IT WOULD FIT FOR PURPOSE PARTICULARLY WHEN THE DATA GETS PUBLISHED. IT MIGHT BE QUITE HARD TO LOOK AT A PAPER OR ANY OUTPUT OF THIS AND REALLY UNDERSTAND WHETHER THERE ARE GAPS OR BIASES OR WHATEVER. AND COULD YOU APPLY ARTIFICIAL INTELLIGENCE TO THE ARTIFICIAL INTELLIGENCE TO SEE IF IT ACTUALLY DOES HAVE THOSE ELEMENTS THAT WOULD MAKE IT A BONA FIDE DATABASE. >> SO MY VERY PRIMITIVE KNOWLEDGE COULD SAY YES. THIS IS AN AS PERFECT RACIAL GOAL AND * AND THAT EFFORTS LIKE THIS WOULD BE FIRST STEPS BUT I WOULD CERTAINLY WELCOME SOME OF THE EXPERTS ON THE ACD IF THEY WOULD LIKE TO ADD MUCH MORE KNOWLEDGEABLE INSIGHT INTO THIS VERY IMPORTANT QUESTION. SO I DON'T KNOW IF DAVID OR DINA OR WINDY IF YOU WOULD LIKE TO WEIGH IN. >> I CAN PROVIDE AN ANSWER TO FRANCIS. SO ACTUALLY IN AI THERE ARE METHODS TO TRY TO MAKE THE DATA MORE -- MORE USEFUL FOR MATCH -- MACHINE PROCESSING. -- RANDOMLY -- AND THEN EXTRACT FROM IT THE INFORMATION THAT IS RELEVANT AND THEN YOU CAN HAVE SOMETHING THAT EVEN LOOKS LIKE A DATABASE. THE VARIOUS FIELDS AND JUST PUT THEM IN A VERY CATEGORIZED WAY. THERE IS ALSO A METHOD THAT ALLOWS US TO TAKE MULTIPLE DATASETS AND ELIMINATE DIFFERENCES LIKE -- TO MAKE THEM MORE SIMILAR IN NATURE. TO MAKE THE DISTRIBUTIONS CLOSER SO THEY CAN ANALYZE THEM TOGETHER. THAT SAID, EVERY TIME YOU DO SOMETHING LIKE THIS YOU MAKE A DECISION ABOUT WHAT INFORMATION TO KEEP AND WHAT INFORMATION TO LOSE IN ORDER TO MAKE THINGS MORE STRUCTURED IN THE WAY YOU EXPECT THEM. OR THE MACHINE EXPECTS THEM TO BE. SO IT IS BETTER DONE AS AN INTERACTIVE PROCESS. WE NEED ENOUGH DATA THAT EVEN WHEN WE DESIGN AND APPLY THE METHODS WE WE CAN SEE THAT IT'S STILL CAPTURING THE INFORMATION BUT REMOVE THE NOISE IF YOU LIKE. OR THE SPECIFIC NOISE TO EVERY DATA SITE. SO I WOULD SAY FIRST THE DATA HAS TO COME AND THEN THE METHOD AND THEN AGAIN ITERATING ON THESE THINGS. BUT SCIENTIFICALLY THERE IS ALSO A METHOD THAT CAN BE USED. >> THANK YOU SO MUCH. THAT IS VERY HELPFUL. >> JUST MAKE A QUICK COMMENT. 7 I REALLY LIKE THE REPORT. ONE OF THE MAIN PROBLEMS WITH THE DATA THAT WE HAVE IS THAT IF THERE IS BIAS IN IT. AND ITS NOT JUST FROM MISSING DATA FROM PEOPLE. IT'S FROM DECISIONS THAT CLINICIANS MAKE THAT ARE BIASED. AND THAT IS THE GOLD STANDARD. HOW CAN WE ADDRESS -- THERE WAS A GREAT PAPER LOOKING AT RADIOLOGISTS AS THE GOLD STANDARDS AND USING CONSUMERS AND PATIENTS TO REDEFINE THAT STANDARD AND HAVE A DIFFERENT OUTCOME. >> THANK YOU SO UP IN. THAT IS GREAT. JAMES, YOU HAVE A COMMENT, PLEASE. >> THANKS, LARRY AND THANKS FOR THE GREAT PRESENTATION. 7 I'M ON THE PRESIDENT'S COVID-19 HEALTH EQUITY TASK FORCE. ONE FOCUS IS DATA AND WE QUICKLY DISCOVERED THAT UNFORTUNATELY FOR A LOT OF PEOPLE FROM MINORITY COMMUNITIES THERE IS NO DATA ON THEIR LEVEL OR QUALITY OF CARE. SO I WONDER WHETHER OR NOT THAT WE'RE GOING TO WIDEN THE CHASM BY HAVING SOME FOR WHOM AI AND MACHINE LEARNING CAN BE APPLIED TO THEIR HEALTH STATUS AND OTHERS THAT WILL BE LEFT OUT. IS THAT SOMETHING THAT WE'RE GOING TO HEAR ABOUT IN THE NEXT PART? >> YOU WILL JAMES. WHAT THE NEXT PRESENTATION WILL I HOPE SHOW EACH OF YOU IS THAT OUR GOAL IS TO PLACE THESE CAPABILITIES CAPACITIES IN COMMUNITIES THAT HAVE NOT TYPICALLY ENJOYED THEM. TO BUILD INFRASTRUCTURE TO SUPPORT TRAINING AND TO ENABLE PEOPLE TO ASK QUESTIONS THAT ARE RELEVANT TO THEM AND THEIR COMMUNITIES. >> EXCELLENT. THANK YOU. >> GETTING AWAY FROM THE MORE TRADITIONAL MODEL OBVIOUSLY. THANK YOU. OKAY I'M GOING TO SAY ROBERTA WILL HAVE THE LAST WORD IF WE CAN AND THEN WE'LL GO ON TO THE NEXT PRESENTATION. >> JUST A QUICK PLEA. THE VA DATA IS A RICH DATASET DESPERATELY IN NEED OF DATA MANAGEMENT AND DATA ACCESS. AND IT PROVIDES A GREAT RESOURCE IN TERMS OF DIVERSITY PARTICULARLY FOR BLACK MEN. AND SO I OFFER UP A PLEA TO PUT THE VA DATA AS A PRIORITY. THANK YOU. >> THANK YOU. AND DAVID GLAZ ER IS POINTING OUT -- WHAT WINDY INDICATED HIGHLIGHTING THE 2019 REPORT. SO THANK YOU DAVID. FRANCIS IF IT'S ALL RIGHT MAY WE PROCEED TO THE NEXT PART OF THE SESSION? >> I THINK THAT SEEMS HIGHLY APPROPRIATE. SO DO YOU WANT TO INTRODUCE DINA. >> I THINK MANY OF THE COUNCIL MEMBERS KNOW DINA. SHE GRACIOUSLY AGREED TO TAKE ONE OF THE LEADING ROLES. SHE IS AT THE NATIONAL HEART, LUNG AND BLOOD INSTITUTE AND I WILL TURN IT OVER TO HER. PLEASE, DINA. >> GOOD AFTERNOON. CAN YOU HEAR ME? >> YES, WE CAN. 7 I THINK SOMEONE WAS GOING TO PULL UP THE SLIDES FOR ME. IF NOT I CAN -- PULL THEM UP. OKAY. GREAT. THANK YOU FOR THE INTRODUCTION. MY NAME IS DINA, ASSISTANT DIRECTOR FOR SCIENTIFIC STRATEGY AND INNOVATION IN THE IMMEDIATE OFFICE OF THE DIRECTOR. THANK YOU FOR THE OPPORTUNITY TODAY TO PRESENT A DRAFT CONCEPT ON DIGITAL HEALTH EQUITY AND TRAINING AND RESEARCH CONSORTIUM. IN THE EFFORT TO REDUCE HEALTH INEQUITIES AND ENHANCE DIVERSITY OF THE AIML WORKFORCE. I WOULD LIKE TO UP FRONT THANK THEM FOR THEIR CONTRIBUTION. SO BASED ON THE EFFORTS OF THE AD HOC WORK GROUP OF ACD AND FURTHER INTERNAL DELIBERATESES MAJOR CHALLENGES HAVE BEEN IDENTIFIED WITH RELATION TO AIML INCLUDING CAPABILITIES KNOWING THAT THEY CAN BE COSTLY, DIFFICULT AND VERY TIME-CONSUMING TO IMPLEMENT AND LEARN. THE FIELD ALSO RUNS A RISK OF PERPETUATING HARMFUL BIASES AS WELL AS OUTCOMES WITH MORE DIVERSE DIVERSITY IN BOTH THE DATA AND THE RESEARCHERS AS WE'VE HEARD IN THE PREVIOUS DISCUSSION. BIOMEDICAL STUDIES AND DATASETS ALSO LACK DIVERSE REPRESENTATION WHICH CAN LEAD TO INADEQUATE UNDERSTANDING. ADDITIONALLY MANY UNDER REPRESENTED COMMUNITIES ALSO HAVE THE POTENTIAL TO CONTRIBUTE DATA, DIVERSE RECRUITMENT, CUTTING EDGE SCIENCE AS WELL AS OTHER EXPERTISE BUT THEY MAY LACK FINANCIAL AND INFRASTRUCTURE AND TRAINING SUPPORT. EHRs CAN BE A GREAT PROVING GROUND TO BEGIN TO BUILD CAPACITY AND LEARNING BUT WE ALSO NEED TO HAVE A PATH OVER TIME TO ADD IN ALL OF THESE OTHER DATA TYPES WHICH WERE MENTIONED PREVIOUSLY IN OUR MEETING TODAY. INCLUDING SOCIAL DETERMINANTS OF HEALTH, GET THEMIC AND GENOMIC DATA, IMAGING AND OTHER DATA TYPES. SO AS YOU MAY KNOW -- THERE WAS A PROVISION FOR ADVANCING LIFE SCIENCE AND THAT IS DEPENDENT ON DATA COMPUTATION AND MACHINE LEARNING. THIS ALSO INCLUDED $50 MILLION TO THE NIH OD TO EXPAND THE NUMBER OF ML FOCUSED GRANTS AND THERE WAS A RECOMMENDATION TO SUPPORT EXPANDED TRAINING INCLUDING FOR UNDER REPRESENTATIVE AND UNDERSERVED GROUPS. SO THE NIH HAS BEEN DISCUSSING A STRATEGY AS TO WHAT TO DO WITH THESE FUNDS AND WE APPRECIATE THE AD HOC WORKING GROUP. WE'RE LOOKING AT A MULTI-YEAR PROGRAM BEGINNING WITH FISCAL YEAR 21 APPROPRIATED FUNDS TO DEVELOP A SUPPORT A RESEARCHER OF DATA NETWORK. THESE INSTITUTIONS WILL NOT ONLY CONTRIBUTE DATA BUT THEY WILL ALSO LEVERAGE AIML INFRASTRUCTURE AND TRAINING AND KNOW HOW AND CONDUCT BIOMEDICAL RESEARCH THAT IS MOST IMPORTANT TO THEM AND ALSO IN LINE WITH THE MISSION. NIH WILL ALSO ACTIVELY SEEK PUBLIC/PRIVATE PARTNERSHIPS TO ACHIEVE THESE OF AMBITIOUS GOALS OF THE PROGRAM. SO THE OVER-ALL GOAL OF THE DIGITAL HEALTH EQUITY TRAINING AND RESEARCH CONSORTIUMS TO ESTABLISH MUTUALLY BENEFICIAL AND COORDINATED PARTNERSHIPS TO INCREASE THE PARTICIPATION AND REPRESENTATION OF RESEARCHES AND COMMUNITIES THAT ARE CURRENTLY UNDER REPRESENTED IN THE DEVELOPMENT OF AIML MODELS AND ALSO ENHANCE THE CAPABILITIES OF THIS EMERGING TECHNOLOGY BEGINNING WITH THE EHR DATA. THE FOCUS OF THE CONSORTIUM IS TO ESTABLISH A COORDINATED DATA AND COMPUTING INFRASTRUCTURE. TO ENHANCE THE INCLUSION OF GROUPS THAT ARE UNDER REPRESENTED IN THE AIML RESEARCH WORKFORCE AND PARTICIPATION ACROSS ALL SCIENTIFIC SUBJECT AREAS. TO REDRESS THE CHALLENGES OF HEALTH DISPARITIES AND HEALTH INEQUITIES AND MINORITY HEALTH. AS WE MENTIONED CONNECT OTHER DATA SUCH AS SOCIAL DETERMINANTS OF HEALTH DATA TO ADDRESS BIASES AND MISSING AND THE LACK OF DATA TO DEVELOP PREDICTIVE MODELS. AND ALSO TO CATALYZE ACCESS TO HIGH QUALITY DIVERSE DATASETS. SO WE'RE LOOKING AT A THREE PRONG APPROACH TO THIS RESEARCH CONSORTIUM. THE FIRST IS INFRASTRUCTURE. WE'RE LOOKING AT A FEDERATED DATA NETWORK WHEREBY DATA CAN BE MINED AND MANAGED AND MAINTAINED BY THE INDIVIDUAL MEMBER INSTITUTIONS SO THAT THE DATA CAN BE -- THE PRIVACY OF THE DATA CAN BE PRESERVED AS WELL AS THE INSTITUTIONS WOULD HAVE AUTONOMY BUT ENSURING THAT THE DATA WOULD BE INNER OPERABLE ACROSS THE NETWORK. THE SECOND IS TRAINING AND PARTNERSHIP. WE'RE LOOKING TO FACILITATE PIPS TO CREATE A NETWORK OF NETWORKS TO INTEGRATE DATA SCIENCE WITH COMMUNITY ENGAGEMENT WITH CLINICAL RESEARCH NETWORKS TO FORM COLLABORATIONS THAT CAN SUPPORT THE ENGAGEMENT OF UNDER REPRESENTED SCIENTISTS ACROSS THE CAREER PIPELINE. AND WE'RE ALSO LOOKING TO SUPPORT RESEARCH QUESTIONS. WE WANT TO FOSTER TRANS DISCIPLINARY PARTNERSHIPS THAT CAN BUILD OR LEVERAGE DATASETS TO DEVELOP AND ENHANCE AIML ALGORITHMS TO BE ABLE TO APPLY APPROACHES TO ADDRESS HEALTH INEQUITIES AND DISPARITIES BUT ALSO TO BE ABLE TO USE THE TECHNOLOGIES AND APPROACHES TO IMPROVE HEALTH CARE, PREVENTION AND DIAGNOSIS AND TREATMENT STRATEGIES AS WELL AS IMPLEMENTATION STRATEGIES. SO LOOKING AT THE STRATEGY FOR INFRASTRUCTURE WE'RE LOOKING AT THREE PARTS OF A SYSTEM. FIRST WOULD BE A COLLABORATIVE NETWORK OF FEDERATED DATA WHERE THE DATA WOULD REMAIN UNDER THE CONTROL OF THE INSTITUTION EITHER ON THE PREMISES OR IN THE CLOUD. PART TWO WOULD INCLUDE AIML APPLICATIONS AND TOOLS THAT CAN RUN OVER THIS STRUCTURE. THE SITES ARE THE ONES CONTROLLING THIS BUT THAT THE MODELS CAN BE TESTED ON DATA COMING FROM THE INSTITUTIONS. AND THEN PART THREE WOULD INCLUDE INFRASTRUCTURE THAT IS SEAMLESS TO THE DATA USER. IF SOMEONE COMES IN ACCESS THESE DATA OR ACCESS THE TOOLS OR APPROACHES IT WOULD BE A SEAMLESS INFRASTRUCTURE TO THEM. FOR TRAINING AND PARTNERSHIPS WE'RE LOOKING AT TWO APPROACHES. WE WANT TO BE ABLE TO CONVENE REGIONAL NETWORKS TO BRING TOGETHER DATA SCIENCE, COMMUNITY ENGAGEMENT AND CLINICAL RESEARCH. WE WANT TO FACILITATE THE INTEGRATION OF UNDER REPRESENTED RESEARCHERS ACROSS THE CAREER PIPELINE TO FOSTER A DEVELOPMENTAL NETWORK SO A LEARNING NETWORK AND ALSO TO BE ABLE TO PRIORITIZE EQUITY IN FUNDING. FOR TRAINING WE'RE LOOKING TO IDENTIFY BOTH GENERAL AND SPECIALIZED CURRICULUM. BUILD ON EXISTING KNOWLEDGE OF EACH STAKEHOLDER TO PIVOT INTO NEW AREAS SO THAT THE DATA SCIENTISTS CAN DELVE INTO HEALTH DISPARITIES RESEARCH AND THE INSTITUTIONS CAN ALSO DELVE INTO DATA SCIENCE AND THIS WOULD HELP THE TRAIN THE TRAINER APPROACH. FOR THE AIML RESEARCH QUESTIONS WE'RE LOOKING TO FORM THESE TRANS DISCIPLINE THEORY PARTNERSHIPS THAT CAN BUILD DATASETS BUT ALSO LEVERAGE EXISTING RESOURCES AND THAT CAN ALSO DEVELOPMENTAL GO RHYTHMS AND APPROACHES. FOR EXISTING RESOURCES WE'RE LOOKING AT EXISTING DATASETS BUT ALSO NIH SUPPORTED PROGRAMS AND INITIATIVES AND STUDIES AND PLATFORMS AND REPOSITORIES THAT ARE HOLDING DATA AND THIS WILL HELP TO ENSURE EQUITABLE ACCESS ACROSS AND SHARING OF BEST PRACTICES. WE WANT TO BE ABLE TO ENGAGE STAKEHOLDERS TO IDENTIFY USE CASES PILOT PROJECTS THAT ARE MOST IMPORTANT TO THEM. POTENTIAL RESEARCH AREAS COULD INCLUDE DETECTION OF BIASES AND OTHER DATASETS OR EVEN HUMAN FACTOR BIASES. OTHER AREAS INCLUDE A CRITERIA FOR WHAT WE'RE DEFINING AS SUCCESS IN AIML AND THE ROLE OF IMPACT OF SOCIAL DETERMINANTS OF HEALTH AND METRICS THAT CAN BE DEVELOPED TO BOTH MEASURE HEALTH DISPARITIES AND INEQUITIES. THE DEVELOPMENT OF PREDICTIVE MODS. SO THOSE ARE JUST SOME. 7 SO THE NEXT STEPS AFTER TODAY -- IF ACD APPROVES THE CONCEPT CLEARANCE THEN WE WOULD LOOK INTO HOLDING STAKEHOLDER ENGAGEMENT. AS SOON AS POSSIBLE. WE WANT TO BE ABLE TO HEAR FROM THE COMMUNITY. WE WANT TO HEAR FROM THE INSTITUTIONS AS TO THEIR INTEREST AND THEIR NEEDS IN BOTH INFRASTRUCTURE, TRAINING AND PARTNERSHIPS AND POTENTIAL RESEARCH AREAS. WE WILL USE THAT INFORMATION TO REFINE THE INITIATIVE. WE WOULD PUBLISH THE RESEARCH OPPORTUNITY ANNOUNCEMENTS LATER IN THE SUMMER AND THEN TRY TO PROVIDE THE AWARDS BY THE ENDS OF SEPTEMBER. AND ONCE AGAIN I WOULD LIKE TO THANK THE CONCEPT TEAM WHO HAS PLAYED A KEY ROLE IN BRINGING ALL OF THIS TOGETHER. AND I GUESS AT THE POINT WHERE WE'RE READY FOR DISCUSSION WE CAN ASK OTHER MEMBERS OF THE TEAM OR THE WORK STREAM LEADS TO CHIME IN -- IN THE DISCUSSION AS NECESSARY. THANK YOU. >> THANK YOU DINA. THAT WAS VERY, VERY CLEAR AND WELL DONE. 7 REALLY APPRECIATE US ACCURATELY REPRESENTING THE MANY CONVERSATIONS THAT YOU AND YOUR COLLEAGUES HAVE HAD ABOUT THIS. SO THANK YOU. SO THIS IS NOW OPEN FOR DISCUSSION. FOR MEMBERS OF THE ADVISORY COMMITTEE. ROBERTA, WOULD YOU LIKE TO BEGIN PLEASE. >> THIS IS VERY EXCITING AND VERY CONSISTENT WITH A NUMBER OF PRIORITIES FOR THE ALL-TIMERS COMMUNITY. I WANTED TO ASK A QUESTION AROUND FOR EXAMPLE HOW DO PEOPLE ENGAGED WITH THIS? WE HAVE A NUMBER OF PROJECTION ONGOING NOW WITH THE NATIVE PEOPLE IN ARIZONA FROM THE NAVAJO NATION TO THE TONA. ACTUALLY TRAINING YOUNG FOLKS IN DATA SCIENCE AND IS THIS SOME PLACE WHERE WE CAN JUMP IN AND CONTRIBUTE? >> DANA, WOULD YOU LIKE TO START AND I WILL BE HAPPY TO WEIGH IN AS WELL. >> SURE. WE ARE LOOKING TO -- FIRSTLY AS PART OF OUR STAKEHOLDER ENGAGEMENT WE WILL BE HOPEFULLY CONVENING PEOPLE VERY SOON TO LEARN WHAT THEIR NEEDS AND INTEREST ARE AND POTENTIAL RESEARCH QUESTIONS AND WE DO SEE THE TRAINING AND PARTNERSHIP COMPONENT OF THIS AS FACILITATING SOME OF THE LEARNING. AND HELPING TO DEVELOP ANY CURRICULUM AND MENTORING AND PARTNERSHIPS THAT COULD ARISE. >> I REALLY DON'T HAVE ANYTHING TO ADD. I THINK THAT DOES REPRESENT THINGS VERY WELL. >> THANK YOU. OTHER QUESTIONS. DINA, YOU HAVE A QUESTION, PLEASE. >> I SEE THAT YOU HAVE MANY DINAS IN THIS MEETING. I'M JUST JOKING BECAUSE TYPICALLY I DON'T FIND IN OTHER MEETINGS I'M THE ONLY DINA SO I'M HAPPY TO SEE A PRESENTATION HERE. I HAVE TWO QUESTIONS. ONE IS RELATED TO HOW -- YOU'RE TALKING ABOUT DEVOTING EFFORT TO CREATE THIS INFRASTRUCTURE AND WHAT CAME INTO THE DISCUSSION ABOUT THE MEDICAL -- THE EHR -- THE RECORD AND WITH LARRY HAS ALSO POINTED OUT. IT'S VERY DISTRIBUTED AND EVEN WITHIN THE INSTITUTION IT IS FRACTURED AND ONE OF THE ISSUES THAT CAME OUT IS ALSO THAT SOME OF THESE INSTITUTIONS TRADITIONALLY DID NOT HAVE INCENTIVES TO SHARE THE DATA. SO I WANT TO JUST UNDERSTAND HOW YOU'RE ENVISIONING IT. LET'S SAY I'M A RESEARCHER AND INTERESTED IN DOING A STUDY THAT WOULD REQUIRE LOOKING AT LET'S SAY MAMMOGRAM IMAGES. AND ALSO I WANT TO KNOW ALL OF THE CO-MORBIDITIES THAT THOSE PATIENTS HAVE WHEN I LOOK AT THAT. SO HOW DO I GET THIS AND HOW DO YOU IMAGINE GETTING THIS INFORMATION FROM THE HOSPITALS AND GETTING THE MEDICAL RECORDS. GETTING SOME OTHER ALSO MAYBE I'M INTERESTED IN LOOKING -- WHEN I LOOK AT CO-MORBIDITIES -- OTHER TYPES OF SIGNALS. LET'S SAY IF THEY HAVE AN EEG SIGNAL TO CONSIDER. 7 HOW -- DO I ACCESS ONE SIDE? DO I GET -- LIKE IF I NEED -- 10,000 RECORDS FOR EXAMPLE. JUST GUIDE ME THROUGH THE PROCESS. >> SO MAYBE DINA -- IF IT'S ALL RIGHT. MAYBE WE CAN TURN TO -- WHO'S A MEMBER OF OUR GROUP WHO HAS AN EXAMPLE THAT HE MIGHT SHARE. SO PARTHA, PLEASE. >> I'LL GIVE YOU THE EXAMPLE THAT -- BEGAN TO MONITOR COVID VACCINATION ACROSS U.S. NURSING HOME AND THE CHALLENGE WAS THERE WAS NOT ENOUGH DATA. THAT IS THE CHALLENGE. AND THE CLAIMS IF WE MONITOR THEM THEY COME WITH A PLOT OF FLAGS. SO THEY PARTNERED WITH FOUR VENDORS ALONG WITH AN ADVISORY BOARD IN ORDER TO CREATE PARTNERSHIPS WHICH WILL ENABLE DATA SHARING AND IT'S A PUBLIC/PRIVATE PARTNERSHIP AKIN TO THE NURSING HOME GETS FEEDBACK ON HOW MANY INDIVIDUALS GOT VACCINATED. RIGHT. AND HOW THEY ARE PERFORMING ACROSS OTHER GROUPS. IN THEIR REGION OR ACROSS THE REGION. SO THERE IS SOME GOOD BACK AND THEY ARE GAINING SOMETHING IN RETURN AND WHAT THEY ARE PROVIDING IS THE EHR DATA WHICH WILL BE ANALYZED. THERE IS A LOT OF MISSING GAPS IN THIS EHR DATA. WE'RE TRYING TO COMPLIMENT THAT BY LINKING IT. THAT WILL PICK UP SOME SIGNAL BUT IT'S NOT GOING TO BE COMPLETE ACROSS ALL TYPES OF POSITIONS. I HOPE THAT IS JUST -- THAT IS JUST ONE EXAMPLE. >> THANK YOU SO MUCH. >> SO THIS IS HELPFUL BUT MOST OF THE MEDICAL -- RECORDS AS YOU KNOW ARE NOT ALWAYS NURSING HOMES -- AND RETIREMENT HOMES, IT'S THE CONCERNS WITH THE HOSPITALS AND THE BIG MEDICAL INSTITUTIONS WHICH HAVE A PROCESS. >> NO QUESTION. BUT PART OF WHAT DINA LAID OUT IS THIS NOTION OF LOCAL CONTROL. WHICH WE THINK IS REALLY, REALLY IMPORTANT TO BUILD TRUST IN COMMUNITIES THAT HAVE NOT HAD THE VERY BEST OF EXPERIENCES WITH RESEARCH EFFORTS AS I'M SURE YOU APPRECIATE. 7 AND SO IT COULD VERY WELL BE THAT BY BUILDING THAT TRUST AND BY PROVIDING THE RESOURCES THAT ARE NEEDED TO HELP FORTIFY WHATEVER EXISTING INFRASTRUCTURE IS AVAILABLE THAT YOU ARE ABLE TO OVER TIME BUILD THE TYPE OF CONSORTIUM MODEL THAT YOU'D NEED TO REALLY TO BUILD LARGE DATASETS FROM MULTIPLE HOSPITALS. THEY MAY NOT BE THE TERRIFIC RARELY CARE -- * HOSPITALS THAT WE'RE FAMILIAR WITH. THEY MAY BE HOSPITALS THAT ARE COMMUNITY-BASED. THEY MAY BE HOSPITALS ASSOCIATED WITH TRIBAL NATIONS FOR EXAMPLE. AND IF THE LARGER ENTITIES WANT TO PARTICIPATE OF COURSE HOPEFULLY THERE IS THE OPPORTUNITY FOR THAT AS WELL. SO AGAIN IT'S A DIFFERENT MODEL. WE'RE TURNING IT ON ITS HEAD. WE'RE NOT GOING TO THE USUAL PLACES THAT WE ALWAYS GO. BUT WE'RE NOT -- WE'RE NOT MINIMIZING THE CHALLENGE. I THINK YOU ARE QUITE RIGHT IN POINTING THAT OUT TO US. >> LIKE CAN I JUST FOLLOW WITH ONE POINT. >> OF COURSE. YES, PLEASE. >> SO TRUST -- LIKE WHAT ARE THE INCENTIVES FOR PEOPLE WHO HAVE THE DATA? TRUST OF COURSE ALL OF THESE ARE MEDICAL INSTITUTIONS. THEIR FIRST PRIORITY IS TO ENSURE THAT EVERYTHING IS ETHICAL AND PRIVACY IS PROTECTED. ALL OF THAT. THAT IS THE FIRST THING BUT THE SECOND ONE THAT COMES AFTER THAT IS THIS NOTION WHICH HAS BEEN AROUND FOR SOME TIME WHICH IS DATA IS OIL OR THE NEXT OIL WHICH IS THE VALUE OF DATA AND WE'RE LIKE WHO GETS WHAT RETURN ON THAT VALUE? AND THAT IS ALSO PART OF THE INCENTIVE AND I DON'T HEAR PEOPLE TALKING ABOUT IT AT ALL WHEN WE HAVE OUR CONVERSATION BUT AT THE END OF THE DAY IT IS PLAYS AN IMPORTANT ROLE. >> AND I AGREE THERE ARE FINANCIAL ASPIRATIONAL GOALS. THIS MAY AFFORD INDIVIDUALS OR INSTITUTIONS AN OPPORTUNITY TO BEGIN TO DO THAT FOR THEMSELVES WITH THEIR OWN DATA. AND SO MY HOPE IS THAT -- AT LEAST FOR SUBSET OF INSTITUTIONS THAT WILL BE A POWERFUL INCENTIVE. ONE OF THE REASONS WHY WE NEED THESE STAKEHOLDER ENGAGEMENT. ACTIVITIES IS TO TEST THAT POSSIBILITY. TO SEE IF PEOPLE -- METAPHORICALLY RAISE THEIR HANDS AND SAY THIS WOULD BE SOMETHING THAT WE WOULD BE END IN ENGAGING IN BECAUSE OF THE POTENTIAL BENEFITS. ALTERNATIVELY WE MAY HEAR SILENCE AND -- BUT THAT IS WHY THE ENGAGEMENT IS SO VERY IMPORTANT. DAVID YOU ARE MAKING A POINT ABOUT A DISCUSSION -- WOULD YOU LIKE TO RAISE THAT WITH THE WHOLE GROUP? >> YES. SECONDING THE CHALLENGE THAT DINA YOU'RE HIGHLIGHTING AND THAT WAS ALSO DONE IN THE MARCH MEETING. BUT AN ENCOURAGING SIGN IS ALL OF US AND MC3 HAVE WITH DIFFERENT APPROACHES AND CONTACTS HAVE HAD SUCCESS IN ALIGNING THE INCENTIVES AND THERE ARE THINGS TO LEARN FROM THAT. THIS IS DIFFERENT BUT RELATED ENOUGH. AND I THINK THE FQACs ENGAGEMENT IS PROBABLY WHERE THERE IS THE MOST TO LEARN FROM TAKING UNDER RESOURCED INSTITUTIONS WHERE IT WOULD NORMALLY BE NOT ON THE TOP 10 PRIORITY LIST AND FINDING A WAY TO LINE IT UP WITH A COMBINATION OF OUTSIDE EXPERTISE AND MISSION ALIGNMENT AND THE TRUST OF BEING PART OF AN NIH PROGRAM. A WHOLE SET OF THINGS. AND THE GOOD NEWS IS THAT IF I THINK ABOUT AND WITH ALL OF US ACCESS TO EHR DATA THERE IS AN EXTRA STEP WHICH IS HAVING THE PROVIDER FILTER THE RECORDS DOWN TO THE PARTICIPANTS THAT ARE ENROLLED AND IT WOULD BE THE OPPORTUNITY TO SKIP THAT STEP BUT USE ALL OF THE REST OF THE PIPES -- THE SAME SHAPE PIPES THAT WERE BUILT. SO THERE IS PROMISING POINTS OF LIGHT TOWARDS TACKLING THESE HARD PROBLEMS. >> THANK YOU VERY MUCH FOR ADDING THAT AND AMPLIFYING. OTHER QUESTIONS OR COMMENTS THAT THE MEMBERS OF THE ACD HAVE PLEASE? I KNOW WE'RE AT TIME -- PRETTY CLOSE. BUT WOULD LIKE ANY ADDITIONAL COMMENTS OR QUESTIONS THAT PEOPLE HAVE. I'M NOT ABLE TO SEE RAISING OF HANDS BUT I DO SEE THINGS IN THE CHAT BOX. SO IF ANYBODY IS RAISING THEIR HAND LET ME KNOW. OKAY. >> LARRY I THINK YOU'RE CAUGHT UP WITH ACD QUESTIONS BOTH CHAT BOX AND OTHERWISE. >> OKAY. GREAT. OKAY. AND RYAN IS PUTTING IN -- INFORMATION ABOUT HOW BLOCKING RULES COULD REVOLUTIONIZE ACCESS TO EHR DATA. THANK YOU, RYAN. HE WAS A MEMBER OF OUR GROUP. THANK YOU. FRANCIS AT THIS POINT I WONDER IF YOU WOULD LIKE TO INDICATE TO THE ACD IF THEY ARE WILLING TO GIVE THEIR SUPPORT FOR US TO MOVE FORWARD. ON THIS -- >> I HOPE YOU'VE GOTTEN A SENSE. WHERE WE WANT TO GO WITH THIS. THE CONCEPT CLEARANCE PRESENTED VERY NICELY BY DINA. THANK YOU, DINA. IT IS BROAD IN ITS DESCRIPTION AND WE HAVE THESE ADDITIONAL STEPS YET TO COME PARTICULARLY STAKE HOLIER ENGAGEMENTS -- STAKEHOLDER ENGAGEMENT. THIS IS PRETTY BOLD. THE IDEA OF * TAKING ON SOMETHING AS CHALLENGING AS AIML AND TO DO IT IN A FASHION THAT PUTS HEALTH DISPARITIES AT THE VERY BEGINNING. AND SO WE'RE GOING TO HAVE AN IMPLEMENTING EXPERIMENT TO SEE WHAT KIND OF RESEARCH COMES FORWARD WITH THIS. AS YOU CAN TELL WE'LL NEED TO DO THAT STAKEHOLDER ENGAGEMENT. I DON'T KNOW IF WE TALKED MUCH LARRY GIVEN THE FACT THAT THIS IS MAY AND WE'RE DOING THE ENGAGEMENT IN JUNE AND YOU WENT OVER A QUICK TIMETABLE. THAT MEANS WE'RE GOING TO NEED TO FUND THIS EFFORT OTHER THAN A TRADITIONAL GRANT CYCLE BECAUSE THESE DOLLARS NEED TO BE ALLOCATED BY SEPTEMBER. SO WE WILL BE USING OUR OTHER TRANSACTION AUTHORITY WHICH GIVES US FLEXIBILITY IN HOW TO MOVE QUICKLY AND ALSO MAKE IT POSSIBLE FOR INSTITUTIONS THAT MAY NOT BE TRADITIONAL HIGH POWERED GRANTEES OF THE NIH TO BE ABLE TO COME AND JOIN THIS PARTICULAR ENTERPRISE AND WE'RE VERY SENSITIVE TO THAT. BUT A LOT OF THE PARTNERS THAT WE WOULD LIKE TO BRING ON-BOARD MAY NOT BE ONES THAT HAVE A HUGE GRANTS DEVELOPMENT OFFICE TO ASSIST THEM AND WE'RE GOING TO THINK OF ALL OF THE WAYS TO PUT TOGETHER APPLICATION OPPORTUNITIES THAT ARE WELCOMING. TO THOSE NONTRADITIONAL PARTNERS. THAT IS GOING TO BE A BIG PART OF THIS. SO I THINK THANK YOU ROBERTA FOR THE POSITIVE COMMENT HERE. SO EXCITING AND ENORMOUS POTENTIAL. YOU ARE SERVING ACD LIKE AND ADVISORY COUNCIL WOULD. AND THAT MEANS WE NEED TO GET YOUR VOTES ABOUT WHETHER OR NOT YOU ARE IN APPROVAL OR MODE FOR THIS CONCEPT CLEARANCE BECAUSE THAT THEN GIVES US THE GREEN LIGHT TO GO TO THE NEXT STEP. WE'RE AT A PRETTY EARLY STEP BUT IT'S A VERY IMPORTANT ONE. IN THIS PUBLIC MEETING TO SEEK YOUR GUIDANCE ABOUT WHETHER WE SHOULD GO FORWARD WITH THIS OVER-ALL PLAN THAT YOU'VE HEARD ABOUT FROM DINA. THE FLOOR IS OPEN IF ANYONE WOULD LIKE TO MAKE A MOTION. >> SO MOVED. >> SO MOVED. >> SECOND. >> OKAY. SOMEBODY IS RECORDING ALL OF THIS I PRESUME. IS THERE FURTHER DISCUSSION ABOUT THE CONCEPT CLEARANCE. OKAY. HEARING NONE LARRY DO WE NEED TO DO AN ORAL ROLL CALL OR CAN WE HAVE PEOPLE SIGNIFY BY RAISING THEIR ELECTRONIC HANDS AND/OR THEIR VISIBLE HANDS AND WE'LL CALL IT A VOTE. >> I THINK A VOICE VOTE IS APPROPRIATE GIVEN THE REALITY THAT WE'RE IN. >> EVERYBODY PLEASE UNMUTE YOURSELVES SO THAT YOUR VOICE VOTE CAN BE APPROPRIATELY RECORDED. ALL IN FAVOR PLEASE SAY AYE. >> AYE. >> THAT WAS REALLY WONDERFULLY COORDINATED. ANY OPPOSED. ANY ABSTENTIONS. I BELIEVE WE HAVE UNANIMITY. LET IT BE SO NOTED THAT THE ACD HAVING DISCUSSED THIS CAME FORWARD WITH A VERY STRONG SUPPORT AND I DO SEE BARBARA'S COMMENT ABOUT THE ASPIRATIONS TERRIFIC BUT THE BUDGET SEEMS SMALL AGAINST THE GOAL. WE AGREED. THAT IS ALWAYS TRUE. AND OF COURSE IT'S A BUDGET WHICH IS GOING INTO THE BASE WHICH WON'T HELP. SO WE'RE TALKING ABOUT A $50 MILLION START POINT BUT OVER TIME THAT WILL BE SOMETHING MORE SUBSTANTIAL. YES AND OF COURSE WE WILL HAVE TO SEE WHAT ELSE HAPPENS IN OUR BUDGETARY CIRCUMSTANCES THAT MIGHT BE ABLE TO COME ALONG SIDE WITH THIS BUT IT'S A PLACE TO START. DAVID GLAZER I TAKE YOUR POINT. IT WOULD BE GOOD IF I GAVE YOU A PERFECT MATCH FOR THE KEY OF F. BUT WE GOT CLEAR. THE SIGNAL WAS CLEAR. ARE THERE ANY OTHER COMMENTS? LARRY THINGS THAT YOU MIGHT WANT TO SAY SINCE YOU'VE SHOULDER SO MUCH OF THE LEADERSHIP FROM THE DIRECTOR'S OFFICE. WE ALL WANT TO SAY THANK YOU OVER THIS PERIOD OF TIME WORKING ON THE AIML EFFORT GOING BACK TO THE ORIGINAL WORKING GROUP AND NOW THIS NEW ONE AND NOW BRINGING FORWARD TO THE ACD THIS VERY SPECIFIC KIND OF PROPOSAL. ANY COMMENT YOU MIGHT WANT TO MAKE BEFORE WE DECLARE VICTORY. >> JUST FRANCIS TO THANK ALL OF THE MEMBERS OF THE ACD. ALL OF THE FOLKS WHO PARTICIPATED IN BOTH WORKING GROUPS AND THE GREAT TEAM THAT WE'VE ASSEMBLED AT NIH TO WORK THROUGH THIS. A LOT OF WORK AHEAD OF US BUT IT'S BEEN WONDERFUL TO WORK WITH ALL OF YOU. THANK YOU. >> THANK YOU. SO ACD MEMBERS, WE ARE GOING TO BRING YOU BACK VIRTUALLY HOWEVER ON JUNE 10-11. I HOPE THAT IS ON YOUR DIARY BECAUSE IT'S NOT THAT FAR AWAY AND WE'RE GOING TO HAVE A VERY INTERESTING AND FULL COUPLE OF DAYS. WE WILL TRY BECAUSE IT IS VIRTUAL AND ZOOM FATIGUE IS A REAL THING TO LIMIT THE NUMBER OF HOURS EACH OF THOSE TWO DAYS TO MAYBE FOUR OR FOUR AND A HALF WITH MAYBE A BREAK IN BETWEEN SO THAT WE CONTINUE TO TAP INTO YOUR BRIGHTEST AND MOST ENERGETIC AND CREATIVE RESPONSES. THERE WILL BE COOL STUFF TO TALK ABOUT. SOME COVID OF COURSE BUT OTHER THINGS AS WELL INCLUDING LARRY TABAK AND JORDAN GLAD MAN, THE PRESENTATION ON THE RESULTS OF THE WORKING GROUP OF ANIMAL MODELS. MUCH ANTICIPATED OUT POURING OF WISE ADVICE AND THE NIH IS VERY INTERESTED IN TO ACHIEVE RECOGNITION OF. SO WE'RE READY FOR THAT IN JUST ABOUT A MONTH AND FOUR DAYS. BARBARA IS SMILING. I DON'T KNOW IF SHE IS GRIMACING. BUT WE'RE GOING TO GET THERE. >> IT WAS WHAT I WAS WORKING ON BEFORE THIS MEETING BEGAN. I'LL RETURN TO IT SHORTLY. >> VERY GOOD. >> FRANCIS BEFORE WE SIGN OFF THIS IS SPERO IT WOULD BE GREAT TO INCLUDE AN UPDATE REGARDING THE UNITE INITIATIVE AND YOUR STRUCTURE RACISM STATEMENT. I WOULD WELCOME THAT OPPORTUNITY TO GET A BRIEF UPDATE. >> YOU WILL GET IT. WE ABSOLUTELY WILL HAVE AN OPPORTUNITY THERE TO TELL YOU WHERE WE ARE WITH UNITE. LARRY, YOU WANT TO SAY MORE ABOUT IS THAT? >> WE'VE HAD A VERY EXTENSIVE ENGAGEMENT THROUGH THE RFI AND WE'LL HAVE IT ANALYZED BY THEN AND BE ABLE TO REPORT OUT SOME OF THE TRENDS AND THEMES AND WE ALSO HAVE SOME INTERNAL INFORMATION THAT WE MIGHT BE ABLE TO SHARE AS WELL AS WE WORK WITH OUR OWN COMMUNITY HERE AT NIH. THAT WILL BE VERY MUCH PART OF THE JUNE MEETING. >> THANK YOU, SPERO. HEARING NO OTHER COMMENTS AND I'M RECOGNIZING THAT AT LEAST IN SOME TIME ZONES IT'S GETTING A BIT INTO HAPPY HOUR. LET ME THANK ALL OF YOU FOR THE OPPORTUNITY TO LEARN FROM YOU AND WE WILL LOOK FORWARD TO TALKING AGAIN IN JUNE. DINA, ARE YOU WAVING GOOD-BYE. >> ONE LAST THING. I WANTED TO ASK BECAUSE WE'RE TALKING ABOUT ML AND AI IN HEALTH TODAY. IS THIS THE LAST TIME WE'RE GOING TO VISIT THIS TOPIC? BECAUSE WE WENT THROUGH THE RECOMMENDATION AS BOTH WORKING GROUPS OR ARE WE GOING TO REVISIT AGAIN? >> WE WILL BE REVISITING IT IN JUNE WHEN A COMMON FUND EFFORT WILL BE DESCRIBED MORE FULLY THAT DIRECTLY ADDRESSES THE NUMBER OF RECOMMENDATIONS MADE BY THE FIRST WORKING GROUP. SO A LITTLE OUT OF SYNC BUT THE DIFFERENT RESOURCE PLOTS MADE IT SO TO DO IT THIS WAY. YES, WE WILL BE REVISITING WHAT THE FIRST WORKING GROUP HAD TO SAY WITH A COMMON FUND INITIATIVE THAT WILL BE DISCUSSED IN JUNE. >> WE ARE NOT DONE WITH SEEKING YOUR ADVICE. NOT EVEN CLOSE. THANKS FOR THE QUESTION. ALL RIGHT. WITH THAT MANY THANKS TO ALL OF YOU AND HAVE A LOVELY AFTERNOON OR EVENING. AND WE WILL SEE YOU IN A LITTLE OVER A MONTH. >> BYE, STAY WELL.