#JACR Transcript
Healthcare social media transcript of the #JACR hashtag.
– ().
See #JACR Influencers/Analytics.
| Profile | Tweet |
|---|---|
 | JACR @JACRJournal Welcome to our #JACR tweet chat on #DataScience moderated by @cekahn and @radhelper Let's start with intros. https://t.co/Aix7WjkIEd |
 | Radhelper PLLC @radhelper RT @cekahn: T minus 1... and counting... #JACR tweetchat -- comin' at ya https://t.co/Bvem8FsytS |
 | Jenny Jones @Jenny_Writer Started 1 minute ago! Join the conversation about data science #jacr https://t.co/JY8nWORGGe |
| JACR @JACRJournal All tweets are assumed to be your own #JACR |
| Radhelper PLLC @radhelper @JACRJournal @cekahn #JACR hi everyone! thanks for joining this tweetchat. i'm the medical director of radiology informatics at stanford children's health and co-editor of the special #JACR issue focused on big data, AI and machine learning |
 | Charles Kahn, MD @cekahn Chuck Kahn here. Radiologist (abdominal CT + US) at @PennRadiology. Have worked in #AI and decision support. #JACR https://t.co/8clh0OcLxc |
| JACR @JACRJournal Be sure to include T1, T2, etc in your tweets so we know which questions you're responding to. Don't forget to include #JACR https://t.co/Z71OJMq250 |
| JACR @JACRJournal RT @radhelper: @JACRJournal @cekahn #JACR hi everyone! thanks for joining this tweetchat. i'm the medical director of radiology informatics at stanford children's health and co-editor of the special #JACR issue focused on big data, AI and machine learning |
| JACR @JACRJournal RT @cekahn: Chuck Kahn here. Radiologist (abdominal CT + US) at @PennRadiology. Have worked in #AI and decision support. #JACR https://t.co/8clh0OcLxc |
| Charles Kahn, MD @cekahn @JACRJournal @radhelper Thanks for joining us! As @radhelper mentioned, he, @ruthcarlosmd and I are guest eds. of an upcoming special issue of #JACR on #AI. |
 | PF Anderson @pfanderson@disabled.social @pfanderson RT @JACRJournal: Welcome to our #JACR tweet chat on #DataScience moderated by @cekahn and @radhelper Let's start with intros. https://t.co/Aix7WjkIEd |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @RadiologyBoy1: T minus 10 minutes until this Month's #JACR tweetchat. Topic is #DataScience including #ArtificialIntelligence #MachineLearning #DeepLearning and #radiology https://t.co/Y7o0BEwhAJ |
| JACR @JACRJournal Let's get started! T1: Why is there so much hype about #AI and #datascience in medical imaging? Is the hype warranted? #JACR https://t.co/StmUbF0MUU |
| PF Anderson @pfanderson@disabled.social @pfanderson Oh, this is going to be fun ... maybe. Love #DataScience and #AI #ArtificialIntelligence but lacking confidence in claims that Watson can read scans as well as a radiologist #JACR |
| Jenny Jones @Jenny_Writer RT @JACRJournal: Let's get started! T1: Let's get started! T1: Why is there so much hype about #AI and #datascience in medical imaging? Is the hype warranted? #JACR https://t.co/StmUbF0MUU |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: @JACRJournal @radhelper Thanks for joining us! As @radhelper mentioned, he, @ruthcarlosmd and I are guest eds. of an upcoming special issue of #JACR on #AI. |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @JACRJournal: Let's get started! T1: Let's get started! T1: Why is there so much hype about #AI and #datascience in medical imaging? Is the hype warranted? #JACR https://t.co/StmUbF0MUU |
| Jenny Jones @Jenny_Writer RT @pfanderson: Oh, this is going to be fun ... maybe. Love #DataScience and #AI #ArtificialIntelligence but lacking confidence in claims that Watson can read scans as well as a radiologist #JACR |
 | Radiology Partners @Rad_Partners T1: The concept of augmenting radiologists with technology that could make radiologists more effective is exciting, particularly for algorithms that identify conditions radiologists don’t typically look for and for things that humans are not optimized to do. #JACR https://t.co/DmUlQuAfTR |
 | Rich Heller @reh3md T1 I think part of the hype is related to peoples idea of possibility. Something I recently learned about is the Gartner Hype Cycle. #jacr |
| Radhelper PLLC @radhelper T1 where do people think we are on the hype cycle? #jacr |
| Radhelper PLLC @radhelper T1 https://t.co/nBMOHBj2Ik #jacr |
 | Y, like yellow, E @jackyeclayton RT @JACRJournal: Let's get started! T1: Let's get started! T1: Why is there so much hype about #AI and #datascience in medical imaging? Is the hype warranted? #JACR https://t.co/StmUbF0MUU |
 | Bernadette Keefe MD @nxtstop1 The #JACR tweet chat on #datascience and #AI in the practice of #imaging #radiology is beginning now ~ do join the conversation. T1 is next |
| PF Anderson @pfanderson@disabled.social @pfanderson Hoping from #HITsm and #hcldr folk pop over into the #JACR chat on #AI hype in medical imaging |
| Charles Kahn, MD @cekahn @JACRJournal T1 Are folks familiar with the @Gartner "hype cycle"? It's often the case that new technologies get promoted (to raise venture capital, etc.) before solid results are in. #JACR |
 | Eileen O'Brien @EileenOBrien RT @nxtstop1: The #JACR tweet chat on #datascience and #AI in the practice of #imaging #radiology is beginning now ~ do join the conversation. T1 is next |
| Bernadette Keefe MD @nxtstop1 T1 Via @radhelper >> T1 Where do people think we are on the hype cycle? #jacr |
| Rich Heller @reh3md T1 - I think we are still on upswing of peak of inflated expectations. Got to be near peak though. Right? #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson @reh3md @RadHelper I've been tracking the Gartner Hype Cycle(s) closely for over a decade. Please note, they don't actually track the viability of the tech, but media response #JACR |
| Radhelper PLLC @radhelper T1 i agree @reh3md that we are near the peak. the reality regarding the "radiologist killer" algorithm is a tall order and elusive #jacr |
 | Penn Radiology @PennRadiology Happening right now, @PennRadiology's @cekhan is moderating @JACRJournal's tweetchat from 12 noon - 1pm EST. Follow the conversation: #JACR https://t.co/eDN4c8zmGN |
 | Becky Haines @grebeccahaines RT @cekahn: @JACRJournal T1 Are folks familiar with the @Gartner "hype cycle"? It's often the case that new technologies get promoted (to raise venture capital, etc.) before solid results are in. #JACR |
| Radiology Partners @Rad_Partners T1: The concept of augmenting radiologists with technology that could make radiologists more effective is exciting, particularly for algorithms that identify conditions radiologists don’t typically look for and for things that humans are not optimized to do. #JACR |
| Jenny Jones @Jenny_Writer RT @cekahn: @JACRJournal T1 Are folks familiar with the @Gartner "hype cycle"? It's often the case that new technologies get promoted (to raise venture capital, etc.) before solid results are in. #JACR |
| Jenny Jones @Jenny_Writer RT @PennRadiology: Happening right now, @PennRadiology's @cekhan is moderating @JACRJournal's tweetchat from 12 noon - 1pm EST. Follow the conversation: #JACR https://t.co/eDN4c8zmGN |
| Bernadette Keefe MD @nxtstop1 T1> 'Hype'= The Gartner Hype Cycle https://t.co/QUFVfLOzSH #JACR https://t.co/tOiYqh8HXR |
| Radhelper PLLC @radhelper T1 @pfanderson @reh3md @RadHelper very good point. the media hype is intense. the reality on the ground is a different story. #jacr |
| Charles Kahn, MD @cekahn @reh3md T1 Yup! #AI #DeepLearning #DataScience are right up there... #JACR https://t.co/bSffOhtDxL |
 | Lindsey Shea Johnstone, MD MS @lagshea RT @cekahn: @JACRJournal @radhelper Thanks for joining us! As @radhelper mentioned, he, @ruthcarlosmd and I are guest eds. of an upcoming special issue of #JACR on #AI. |
| Lindsey Shea Johnstone, MD MS @lagshea RT @JACRJournal: Welcome to our #JACR tweet chat on #DataScience moderated by @cekahn and @radhelper Let's start with intros. https://t.co/Aix7WjkIEd |
| Becky Haines @grebeccahaines T1: I think one of the reasons the hype is so high is that folks are excited by the potential for improved care and greater satisfaction for radiologists who can gain > confidence in their decisions #JACR |
| Radhelper PLLC @radhelper T1 rad workflow augmentation is not as "sexy" as a fully automated image diagnosis and interpretation. a lot of the #AI claims fail #jacr |
| Charles Kahn, MD @cekahn T1 -- Is the #AI hype justified?? #JACR https://t.co/KPY1xPSxx5 |
| Lindsey Shea Johnstone, MD MS @lagshea RT @JACRJournal: Let's get started! T1: Let's get started! T1: Why is there so much hype about #AI and #datascience in medical imaging? Is the hype warranted? #JACR https://t.co/StmUbF0MUU |
| Bernadette Keefe MD @nxtstop1 T1 Granted there is a ton of hype around #artificialintelligence and it's imminent earth-shaking effects, RE: all industries, but..... #jacr |
| JACR @JACRJournal T2: Will computer algorithms replace radiologists? Or, should I switch to a new medical specialty? #JACR https://t.co/BEYr4ef5Xp |
 | Penn Medicine MD @PennMDForum RT @PennRadiology: Happening right now, @PennRadiology's @cekhan is moderating @JACRJournal's tweetchat from 12 noon - 1pm EST. Follow the conversation: #JACR https://t.co/eDN4c8zmGN |
 | Brett Hansen @RadiologyAlly T1: Everyone has their own idea of the future, but it never comes true exactly how you think it will. #jacr #AI #ThursdayThoughts https://t.co/Wi6ngMHn9I |
| PF Anderson @pfanderson@disabled.social @pfanderson @reh3md Yep. A group I'm in for librarians working with emerging tech had a member this week make the claim that Watson can read x-rays as well as a radiologist. She's pretty politically powerful, so I cdnt say what I really thought about that. #jacr |
| Lindsey Shea Johnstone, MD MS @lagshea RT @JACRJournal: Just one hour until the #JACR tweet chat. on Data Science. The JACR has some great #AI #machinelearning articles in print https://t.co/dAuGwIbaNc https://t.co/K7d0el1W1P |
| Radhelper PLLC @radhelper T1 @grebeccahaines very important point. #AI companies seem to be more focused at replacing docs versus improving patient care. #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson Watson Health: Medical Imaging https://t.co/LAovWcWIKQ #JACR |
| Bernadette Keefe MD @nxtstop1 T1 Re: application in #imaging #radiology specifically> IF #radiologists want to be in on the ground floor (have a say)>now's the time #jacr |
 | Nicole Racadag @NRacadag If you're not on this #jacr tweet chat on data science, you're missing out! Great convo going on #AI and #DeepLearning! #radiology @JACRJournal |
 | Ed Mariano, MD, MAS, FASA, FASRA 🇺🇸🇵🇭 @EMARIANOMD RT @nxtstop1: T1> 'Hype'= The Gartner Hype Cycle https://t.co/QUFVfLOzSH #JACR https://t.co/tOiYqh8HXR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: T1 i agree @reh3md that we are near the peak. the reality regarding the "radiologist killer" algorithm is a tall order and elusive #jacr |
| Becky Haines @grebeccahaines T1 but workflow automation, even small changes, can add up to > time saved < cost & > satisfaction. Not sexy but what a relief! #jacr incremental change is good https://t.co/Lgj13cQDk7 |
 | Saurabh Jha @RogueRad Ahh the existential question! No it won't replace, but could chew a few FTEs. Some workforce planning might be in order #JACR https://t.co/mUbBM4EktD |
| Penn Radiology @PennRadiology Happening right now, @PennRadiology's @cekahn is moderating @JACRJournal's tweetchat from 12 noon - 1pm EST. Follow the conversation: #JACR https://t.co/eDN4c8zmGN |
| Penn Medicine MD @PennMDForum RT @PennRadiology: Happening right now, @PennRadiology's @cekahn is moderating @JACRJournal's tweetchat from 12 noon - 1pm EST. Follow the conversation: #JACR https://t.co/eDN4c8zmGN |
| Radhelper PLLC @radhelper T1 @RogueRad could it also eat into RVU reimbursement? if rads are more efficient, will take less time to perform and interpret exams #jacr |
| Charles Kahn, MD @cekahn T1 - Agree! There's a lot of gee-whiz excitement about "superhuman" #AI abilities. It's new, and shows the potential to use the latest parallel-processing (GPU) technologies. #JACR https://t.co/frtxtHnGy6 |
| Rich Heller @reh3md T1 Totally agree. Analytics & workflow are areas that are impactful but not as obvious. Computers / #AI to help analyze reports. #JACR https://t.co/ffi5kBm2rK |
| PF Anderson @pfanderson@disabled.social @pfanderson T1: my expectation is that #ai in medical imaging will follow same pattern as medical AI diagnosis, where it becomes a tool for decision support rather than supplanting the human expert. Shifts skills & competencies needed to be expert #JACR |
| Becky Haines @grebeccahaines T2: https://t.co/FwKZftsAJS #JACR |
| Radiology Partners @Rad_Partners T2: No, humans & computer learning algorithms have different strengths that complement each other. AI won't replace radiologists, rather a #radiologist who embraces #AI will replace one who does not; and practices who embrace AI will replace those who do not. #JACR |
| Charles Kahn, MD @cekahn @RogueRad T2 - How do we re-align? Which areas might be affected first? #JACR |
| Jenny Jones @Jenny_Writer T1 what role can radiologists play in taming the hype cycle/steering the hype cycle to focus on patient care? #JACR https://t.co/y4iuBDG4jf |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: T1 rad workflow augmentation is not as "sexy" as a fully automated image diagnosis and interpretation. a lot of the #AI claims fail #jacr |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: @reh3md T1 Yup! #AI #DeepLearning #DataScience are right up there... #JACR https://t.co/bSffOhtDxL |
| PF Anderson @pfanderson@disabled.social @pfanderson The real strength of #AI in supporting medical diagnosis and decision making, is managing implicit bias reduction on the part of the human expert — what was overlooked, what assumptions were made #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @Rad_Partners: T2: T2: No, humans & computer learning algorithms have different strengths that complement each other. AI won't replace radiologists, rather a #radiologist who embraces #AI will replace one who does not; and practices who embrace AI will replace those who do not. #JACR |
| Saurabh Jha @RogueRad Yes. "AI will scale our productivity" is a trap. Don't fall for it #JACR https://t.co/JmcbzDiQBs |
| PF Anderson @pfanderson@disabled.social @pfanderson @Rad_Partners Bingo. You nailed it. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @grebeccahaines: T2: T2: https://t.co/FwKZftsAJS #JACR |
| Charles Kahn, MD @cekahn T1 Definitely. The most productivity may come from the least sexy applications, at least at first. #JACR https://t.co/JRhEKeZDdP |
| Charles Kahn, MD @cekahn T1 - Education. The hype blurs out the real truths about #AI strengths + weaknesses. #JACR https://t.co/4jltxwGepK |
| Rich Heller @reh3md @radhelper @RogueRad T1 Which has to be balanced with the practice expense of investing in AI. #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson I spend a lot of time in #JACR chats, but don't usually trot out what I really do in my day job. I'm an emerging technologies informationist who has been both a medical librarian & engineering librarian (where I was the artificial intelligence & neural networks liaison). |
| Saurabh Jha @RogueRad @cekahn Give unto AI what is unto AI. Lung nodules, head CTs to R/O bleed, PE studies. All sub specialists should prepare to downsize #JACR |
 | CancerGeek @CancerGeek T2: No; They will enable #radiologists to focus on more difficult cases as well as providing more one on one with patients & other Md #jacr |
| Brett Hansen @RadiologyAlly Agree Historically technology has raised expectations in increased productivity rather than provided individuals with "more time" #JACR https://t.co/a2KV3xP2NM |
| Bernadette Keefe MD @nxtstop1 An adjustment in alignment/FTEs has been suggested MT @cekahn: @RogueRad T2 -How do we re-align? Which areas might be affected first? #jacr |
| Becky Haines @grebeccahaines Agree 100% Digital anything (publishing in my case) seems to accelerates and transforms workload. Work is not less just different #jacr https://t.co/xIznGVX842 |
| Charles Kahn, MD @cekahn T2 - Yep. Most of us believe Rad + AI will be better than either alone. #JACR https://t.co/AFvDvnKKV0 |
| CancerGeek @CancerGeek RT @cekahn: T2 - Yep. Most of us believe Rad + AI will be better than either alone. #JACR https://t.co/AFvDvnKKV0 |
| Rich Heller @reh3md RT @cekahn: T2 - Yep. Most of us believe Rad + AI will be better than either alone. #JACR https://t.co/AFvDvnKKV0 |
| Bernadette Keefe MD @nxtstop1 @nxtstop1 @cekahn @RogueRad Any plain films used for #binary type questions /diagnosis i.e. Yes or No fracture. #mskrad #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson @cekahn @RogueRad T2: new competencies to build and require are 1) lifelong learning skills; 2) general tech competencies; 3) plain language & storytelling skills to communicate; 4) being able to customize algorithms & software & DataViz; 5) bias/stigma reduction #jacr |
| Becky Haines @grebeccahaines RT @cekahn: T2 - Yep. Most of us believe Rad + AI will be better than either alone. #JACR https://t.co/AFvDvnKKV0 |
| CancerGeek @CancerGeek T2: Like all #tech its only as good as the #humans using it. #AI has ability to improve the work if leveraged correctly #jacr |
| Radhelper PLLC @radhelper @nxtstop1 @cekahn @RogueRad T2 Rads and groups that don't provide value or efficiency will be the first to get hit if #ai tools aren't utilized #JACR |
| Charles Kahn, MD @cekahn T2 - As for workload: do we see that going DOWN anytime soon as our popluation ages? #AI will shift the work in #radiology. #jacr |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: @nxtstop1 @cekahn @RogueRad T2 Rads and groups that don't provide value or efficiency will be the first to get hit if #ai tools aren't utilized #JACR |
 | Vasanth @vasanthdrv T1 the hype is amplified by oversimplification of radiology as a field of reading images. Until data scientists realize the fine print, the expectations will be simplistic and unrealistic #JACR |
| Rich Heller @reh3md @nxtstop1 @cekahn @RogueRad Or based on accepted reference data pools like #boneage #jacr |
| Bernadette Keefe MD @nxtstop1 MT @cekahn: T1 Definitely. The most productivity may come from the least sexy applications, at least at first https://t.co/CQAt8jewTu #jacr |
| Charles Kahn, MD @cekahn Agree! Some combo of "liberal arts" and "techie" skills, then... #jacr https://t.co/CrWtnOChtg |
| Bernadette Keefe MD @nxtstop1 Yes>> precisely! RT @reh3md: @nxtstop1 @cekahn @RogueRad Or based on accepted reference data pools like #boneage #jacr |
| CancerGeek @CancerGeek T2: I think the other thought is will #AI + #Radiologist eliminate need of other #physicians in healthcare? #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson @cekahn Exactly. As it takes up some burdens, it will create new ones. Work and workflows shift, but the overall burden stays the same or increases #JACR |
 | Seth M Hardy, MD MBA exFACR @sethmhardy Everyone thinks the other person’s job is easy. #jacr https://t.co/JYnyXULdu7 |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @vasanthdrv: T1 the hype is amplified by oversimplification of radiology as a field of reading images. Until data scientists realize the fine print, the expectations will be simplistic and unrealistic #JACR |
| Radhelper PLLC @radhelper @CancerGeek T2 it depends on whether or not we get reimbursed for value we provide over volume we interpret and bill for. i'm also an optimist but the realist in me says that rads will try to preserve the status quo in reimbursement #JACR |
 | Syam Reddy MD @sreddyimaging RT @Rad_Partners: T1: T1: The concept of augmenting radiologists with technology that could make radiologists more effective is exciting, particularly for algorithms that identify conditions radiologists don’t typically look for and for things that humans are not optimized to do. #JACR |
| Becky Haines @grebeccahaines Great recording of @ACRRFS journal club pointing out challenges w #AI & describing specific issues, how to review critically https://t.co/5yxZITEplj #JACR |
| Bernadette Keefe MD @nxtstop1 MT @cekahn: T2 As for workload: do we see that going DOWN anytime soon as our pop. ages?{No}#AI will shift the work in #radiology{Yes} #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson @vasanthdrv Love this. Thank you for saying it so well, that the human in the radiologist contributes value beyond simply looking at pictures & saying what they are. #jacr |
| CancerGeek @CancerGeek RT @pfanderson: I spend a lot of time in #JACR chats, but don't usually trot out what I really do in my day job. I'm an emerging technologies informationist who has been both a medical librarian & engineering librarian (where I was the artificial intelligence & neural networks liaison). |
| Saurabh Jha @RogueRad Pre-TAVR measurements will drive even AI to despair #JACR https://t.co/GNpeYwYfg5 |
| Charles Kahn, MD @cekahn T2 Some examples: ECGs are read by machines ("AI"), but cardiologists still bill for them (and overread them). Lab medicine docs don't measure serum creatinine in testtubes, they QC the machines that process the samples #jacr |
| Radhelper PLLC @radhelper @grebeccahaines @ACRRFS T2 there will definitely be a whole industry on trying to validate the #AI tools which will take domain expertise #JACR |
| Rich Heller @reh3md @pfanderson @cekahn Agree. A point driven home in the movie #HiddenFigures when machine-based computers first move in. #jacr |
| Becky Haines @grebeccahaines Volume reimbursement likely a basis for value $$ per @zekesilva3 so important to advocate strategically to ensure patients get care they need. #jacr https://t.co/1iIpAkkyxk |
| PF Anderson @pfanderson@disabled.social @pfanderson @CancerGeek I don't see how. It isn't providing care, it's supporting diagnosis. Even now, radiology reports are integrated w/ other information into a broad picture to support diagnosis for the individual patient. But may make for interesting #PrecisionHealth #PrecisionMedicine stuff #JACR |
| Bernadette Keefe MD @nxtstop1 RT @grebeccahaines: T2: T2: https://t.co/FwKZftsAJS #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: T2 Some examples: T2 Some examples: ECGs are read by machines ("AI"), but cardiologists still bill for them (and overread them). Lab medicine docs don't measure serum creatinine in testtubes, they QC the machines that process the samples #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: Agree! Some combo of "liberal arts" and "techie" skills, then... #jacr https://t.co/CrWtnOChtg |
| Becky Haines @grebeccahaines Indeed @RadiologyACR Data Science Institute https://t.co/A89jS9JeQw teady to do just that #jacr https://t.co/NMKF1eyZzv |
| JACR @JACRJournal T3: Where will data science and #AI impact #medicine and specifically #radiology the most? #JACR https://t.co/jWRVuIIqsg |
| Jenny Jones @Jenny_Writer T2 @UVARadiology are using an #AI app to help them detect incidental findings on CT scans. It's even helping them detect low bone density before symptoms are visible to them. It's a tool, not replacement https://t.co/gPhHXF8AOS #JACR #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @cekahn Combined with the EXISTING requirements of medical knowledge and infrastructure understanding #JACR |
| Jenny Jones @Jenny_Writer RT @grebeccahaines: Indeed @RadiologyACR Data Science Institute https://t.co/A89jS9JeQw teady to do just that #jacr https://t.co/NMKF1eyZzv |
| Radhelper PLLC @radhelper T2 out of curiosity, if you could have 1 #AI tool at your practice, what would it be? smart worklist? better hanging protocol? automation of protocoling? detection of lung nodules on CT? #JACR |
| Bernadette Keefe MD @nxtstop1 RT @grebeccahaines: Great recording of @ACRRFS journal club pointing out challenges w #AI & describing specific issues, how to review critically https://t.co/5yxZITEplj #JACR |
| CancerGeek @CancerGeek RT @JACRJournal: T3: T3: Where will data science and #AI impact #medicine and specifically #radiology the most? #JACR https://t.co/jWRVuIIqsg |
| Charles Kahn, MD @cekahn T2 Agree. Which means rads will need to maintain clinical + image-interpretation skills, PLUS gain new skills to oversee new tech. #JACR https://t.co/bVBevwl4iF |
| Bernadette Keefe MD @nxtstop1 RT @pfanderson: @cekahn @RogueRad T2: @cekahn @RogueRad T2: new competencies to build and require are 1) lifelong learning skills; 2) general tech competencies; 3) plain language & storytelling skills to communicate; 4) being able to customize algorithms & software & DataViz; 5) bias/stigma reduction #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @JACRJournal: T3: T3: Where will data science and #AI impact #medicine and specifically #radiology the most? #JACR https://t.co/jWRVuIIqsg |
 | Ellen Silveira @ellenbsilveira Good tweet chat happening now led by @JACRJournal on #AI in medical imaging + hype cycle of emerging tech https://t.co/AUTEXB5TCP |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @Jenny_Writer: T2 @UVARadiology are using an #AI app to help them detect incidental findings on CT scans. It's even helping them detect low bone density before symptoms are visible to them. It's a tool, not replacement https://t.co/gPhHXF8AOS #JACR #JACR |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: @grebeccahaines @ACRRFS T2 there will definitely be a whole industry on trying to validate the #AI tools which will take domain expertise #JACR |
| Charles Kahn, MD @cekahn T2. Classify thyroid nodules 😀 #JACR https://t.co/XkY5ngdMgR |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: T2 Some examples: T2 Some examples: ECGs are read by machines ("AI"), but cardiologists still bill for them (and overread them). Lab medicine docs don't measure serum creatinine in testtubes, they QC the machines that process the samples #jacr |
| Seth M Hardy, MD MBA exFACR @sethmhardy Which will be interesting given the verbal promise to be “Vanderbilt neutral”. #jacr https://t.co/2VUovIMvIX |
| PF Anderson @pfanderson@disabled.social @pfanderson @Jenny_Writer @UVARadiology Perfect example of how #AI and #DataScience can SUPPORT existing radiology work while improving patient care through early diagnosis. #JACR |
| Becky Haines @grebeccahaines Course recommended during @ACRRFS journal club is https://t.co/ulCOkmQCuP #JACR https://t.co/dhPZnKawue |
| Radhelper PLLC @radhelper T3 biggest impact will be population health. the normal/abnormal detection on imaging may be accessible to those that are not specialized to interpret imaging. so, they will know when something is wrong but may not know what it is #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy Which will be interesting given the promise to be “vendor neutral” that was made at ACR 2017. #jacr https://t.co/2VUovIMvIX |
| Jenny Jones @Jenny_Writer Agree! This is the future. #JACR https://t.co/zbpxLFRzSN |
| Radiology Partners @Rad_Partners A3: Data science & AI should be used to make it easier for radiologists to practice better. This goal can be met at any part of our imaging workflow from elevating critical cases on a worklist to providing relevant patient information for image interpretation/consultation. #JACR |
| Bernadette Keefe MD @nxtstop1 via @JACRJournal T3: Where will data science and #AI impact #medicine and specifically #radiology the most? #jacr |
| CancerGeek @CancerGeek T3: 1) Skip to best exam for clinical indication 2) Move from pop based models to #Nof1 for imaging & F/U 3) Find new correlations #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @grebeccahaines: Course recommended during @ACRRFS journal club is https://t.co/ulCOkmQCuP #JACR https://t.co/dhPZnKawue |
| Becky Haines @grebeccahaines Maybe I should clarify...algorithm validation #JACR https://t.co/Z3GeqCowAh |
| PF Anderson @pfanderson@disabled.social @pfanderson @Jenny_Writer @UVARadiology Case Study: AN Extra Set of Eyes: https://t.co/LHXe9JJnX6 #JACR |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: T2 Agree. Which means rads will need to maintain clinical + image-interpretation skills, PLUS gain new skills to oversee new tech. #JACR https://t.co/bVBevwl4iF |
| Bernadette Keefe MD @nxtstop1 RT @grebeccahaines: Course recommended during @ACRRFS journal club is https://t.co/ulCOkmQCuP #JACR https://t.co/dhPZnKawue |
| Charles Kahn, MD @cekahn @radhelper T3 - So what about limited applications, esp. in dev'ing nations? TB detection as a use case? Would nl/abnl be enough? #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @radhelper This makes me think of tools to help general practice docs ID appropriate referral, or supporting Dx in telemedicine. Mixed feelings, though #JACR |
| CancerGeek @CancerGeek T3: 4) Reduce "indeterminate results" 5) Create an Alexa Skills Dr. Jha for every reading room in the U.S. to keep everyone honest #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: T3 biggest impact will be population health. the normal/abnormal detection on imaging may be accessible to those that are not specialized to interpret imaging. so, they will know when something is wrong but may not know what it is #JACR |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: T2 out of curiosity, if you could have 1 #AI tool at your practice, what would it be? smart worklist? better hanging protocol? automation of protocoling? detection of lung nodules on CT? #JACR |
| Becky Haines @grebeccahaines RT @pfanderson: @Jenny_Writer @UVARadiology Case Study: AN Extra Set of Eyes: @Jenny_Writer @UVARadiology Case Study: AN Extra Set of Eyes: https://t.co/LHXe9JJnX6 #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy @grebeccahaines It’s a very slippery slope. #jacr |
| Radiology Partners @Rad_Partners T3 continued: AI can also help rads accomplish tasks from identifying conditions that are difficult to capture or tedious to measure to assisting the radiologist in creating the report to identifying patients who are due for follow-up. Much opportunity here. #JACR |
| Radhelper PLLC @radhelper @pfanderson T3 this may drive more business to radiologists to help provide expert opinion and follow-up recommendations #JACR |
| Jenny Jones @Jenny_Writer T3 While #AI won't replace radiologists, the radiologist's role will almost certainly change. #Imaging3 provides a guide for helping radiologists prepare for this future. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @Vilavaite @CancerGeek I would really like for specialists to able to ID when an AI algorithm is making assumptions that don't apply to the patient in front of them, and modify it appropriately #JACR |
| Becky Haines @grebeccahaines @sethmhardy Modeling Accreditation process gives me comfort. What do you think? #jacr |
| Radhelper PLLC @radhelper @CancerGeek T3 and reduce the variability in our follow-up recommendations or provide some potential outcomes stratification #JACR |
| CancerGeek @CancerGeek RT @radhelper: @CancerGeek T3 and reduce the variability in our follow-up recommendations or provide some potential outcomes stratification #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy @cekahn @radhelper Committing someone to TB therapy based on #AI is a big deal. #jacr |
| CancerGeek @CancerGeek @radhelper Yes!! #jacr |
| Jenny Jones @Jenny_Writer RT @pfanderson: @Vilavaite @CancerGeek I would really like for specialists to able to ID when an AI algorithm is making assumptions that don't apply to the patient in front of them, and modify it appropriately #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @radhelper It could, as long as primary care docs don't fall into the habit of trusting the machine when they shouldn't #JACR |
 | AI Today @IndustryAi RT @sethmhardy: @cekahn @radhelper Committing someone to TB therapy based on #AI is a big deal. #jacr |
| Bernadette Keefe MD @nxtstop1 Resource > MT @grebeccahaines: Indeed @RadiologyACR Data Science Institute https://t.co/qYw42pVJrE https://t.co/PLXDfHko6Q #jacr |
| CancerGeek @CancerGeek T3: Identify when #radiologist fatigue is setting in and remind them to take a break #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: @radhelper T3 - So what about limited applications, esp. in dev'ing nations? TB detection as a use case? Would nl/abnl be enough? #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @Rad_Partners: T3 continued: T3 continued: AI can also help rads accomplish tasks from identifying conditions that are difficult to capture or tedious to measure to assisting the radiologist in creating the report to identifying patients who are due for follow-up. Much opportunity here. #JACR |
| Rich Heller @reh3md RT @CancerGeek: T3: T3: 1) Skip to best exam for clinical indication 2) Move from pop based models to #Nof1 for imaging & F/U 3) Find new correlations #jacr |
| Charles Kahn, MD @cekahn @radhelper @pfanderson T3 If the test is how well it lets us care for the patient (in terms of both quality and cost), we will still be in the picture. It might redirect some care from specialists to less-expensive generalists. #JACR |
| Radhelper PLLC @radhelper @Rad_Partners T3 totally agree. if #AI tools can incorporate prior exams and EHR data, would be very powerful synthesis for providers and patients #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy @grebeccahaines I already perceive undisclosed COI today. Not optimistic. #jacr |
| Charles Kahn, MD @cekahn That's been a goal of some of my group's work #JACR https://t.co/VMQaVRPJa8 |
| PF Anderson @pfanderson@disabled.social @pfanderson I'm thinking about how in genetics, the BRCA samples became proprietary of a single company. Algorithms are likely to also be proprietary, not transparent. There are risks in that #JACR |
| Radhelper PLLC @radhelper @grebeccahaines @sethmhardy Can we rely on @US_FDA to keep up with the technology advancements? #JACR |
 | Deep Learning City @DeepLearningCty @reh3md T1 Yup! #AI #DeepLearning #DataScience are right up there... #JACR https://t.co/5hEsKlcva5 |
| Jenny Jones @Jenny_Writer T3 Triage cases, putting the most critical cases at the top of a radiologist's work list #JACR |
| CancerGeek @CancerGeek T3: Automatically query and retrieve all pertinent information & present in easy to see/hear while #radiologist is reading exam #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @sethmhardy: @grebeccahaines I already perceive undisclosed COI today. Not optimistic. #jacr |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @nxtstop1: Resource > MT @grebeccahaines: Resource > MT @grebeccahaines: Indeed @RadiologyACR Data Science Institute https://t.co/qYw42pVJrE https://t.co/PLXDfHko6Q #jacr |
| Jenny Jones @Jenny_Writer RT @radhelper: @Rad_Partners T3 totally agree. if #AI tools can incorporate prior exams and EHR data, would be very powerful synthesis for providers and patients #JACR |
| Bernadette Keefe MD @nxtstop1 @grebeccahaines @ACRRFS Thanks- NB: The #AI #imaging @ACRRFS @RadiologyACR journal club now has dedicated hashtag : #RadAIJC Cc #JACR |
| Rich Heller @reh3md @radhelper Assistance w/ evidence based guidelines. Cant memorize them & time consuming to look up (or I may not even know there is a guideline). #jacr |
| CancerGeek @CancerGeek T3: Scrape the findings/summary into a #patient friendly report and given to patient before they leave #jacr |
| Radhelper PLLC @radhelper @pfanderson T3 we are already seeing industry and health systems keep a closed lid on their data. everyone thinks their image data and rad reports are some kind of #bitcoin :) #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @Vilavaite @CancerGeek Can we NOW make a strong push for a Github / GitLab archive of AI algorithms used for medical imaging diagnosis? Proprietary / non-editable algorithms make me really nervous re. patient safety #jacr |
| Jenny Jones @Jenny_Writer T3 Create a patient-friendly report automatically from the radiologist's traditional report #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy And error detection, like why Mr. Jones suddenly has a uterus. #jacr https://t.co/acMuVyDroK |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: @radhelper @pfanderson T3 If the test is how well it lets us care for the patient (in terms of both quality and cost), we will still be in the picture. It might redirect some care from specialists to less-expensive generalists. #JACR |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: @Rad_Partners T3 totally agree. if #AI tools can incorporate prior exams and EHR data, would be very powerful synthesis for providers and patients #JACR |
| CancerGeek @CancerGeek RT @Jenny_Writer: T3 Create a patient-friendly report automatically from the radiologist's traditional report #JACR |
| Saurabh Jha @RogueRad Deep wisdom in this truism #JACR https://t.co/0cFVygGYe6 |
| Charles Kahn, MD @cekahn @pfanderson Unlike conventional software, the *data* generate the #AI tools. So large groups may be able to build their own systems (using published AI shells) from their own images. (It avoids FDA scrutiny...) #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @radhelper @grebeccahaines @sethmhardy @US_FDA No. Their testing and validation process is (of necessity) too slow. Also, new admin requirement for political oversight of projects is problematic #JACR |
| Jenny Jones @Jenny_Writer I can see how that would be helpful! :) https://t.co/oSSGVOTz1M |
| Ellen Silveira @ellenbsilveira Well said @Rad_Partners #JACR https://t.co/5ty78u7sZf |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @RogueRad: Deep wisdom in this truism #JACR https://t.co/0cFVygGYe6 |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: @radhelper @pfanderson T3 If the test is how well it lets us care for the patient (in terms of both quality and cost), we will still be in the picture. It might redirect some care from specialists to less-expensive generalists. #JACR |
| JACR @JACRJournal T4: How does someone learn more about #datascience, #AI and informatics? Is there additional training or certification? #JACR https://t.co/Vw3kF4YHgD |
| Radhelper PLLC @radhelper RT @ellenbsilveira: Well said @Rad_Partners #JACR https://t.co/5ty78u7sZf |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: @pfanderson Unlike conventional software, the *data* generate the #AI tools. So large groups may be able to build their own systems (using published AI shells) from their own images. (It avoids FDA scrutiny...) #JACR |
| Rich Heller @reh3md RT @ellenbsilveira: Well said @Rad_Partners #JACR https://t.co/5ty78u7sZf |
| Seth M Hardy, MD MBA exFACR @sethmhardy @RogueRad “Still waiting after all these years.” #jacr |
| Charles Kahn, MD @cekahn @ellenbsilveira @Rad_Partners And that's we we best use the technology! Computers + humans each do what they do best. #JACR |
 | Matthew Loxton is also on Mastodon @mloxton would love to be on #JACR Chat, but cleaning data and fixing reports instead |
| Radhelper PLLC @radhelper RT @RogueRad: Deep wisdom in this truism #JACR https://t.co/0cFVygGYe6 |
| Jenny Jones @Jenny_Writer T4 Lots of information out there, including on the #ACRDSI blog https://t.co/V8LDK0kxs3 #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @cekahn A big recent trend in #AI is machines that (I'm oversimplifying) write their own code & algorithms. Makes development fast, includes humans as guides, but makes it really hard to understand what they're doing & why #JACR |
| Radhelper PLLC @radhelper T4 lots of knowledge about #AI and #ML that is available. one example: https://t.co/Y7L9hApidY #JACR |
 | Boiarski @Boiarski The #Russians are building the Matrix. #tRump's cabinet all own #MatrixStock. https://t.co/nqkDhPRcGG |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @JACRJournal: T4: T4: How does someone learn more about #datascience, #AI and informatics? Is there additional training or certification? #JACR https://t.co/Vw3kF4YHgD |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: @pfanderson T3 we are already seeing industry and health systems keep a closed lid on their data. everyone thinks their image data and rad reports are some kind of #bitcoin :) #JACR |
| Radiology Partners @Rad_Partners T4: AI is a popular topic these days so there is plenty to read about it. @SIIM_Tweets holds an annual conference dedicated to AI, there are multiple lectures at radiology conferences (e.g., @RSNA), & much can be learned by speaking to the multitude of vendors in the space. #JACR |
| Jenny Jones @Jenny_Writer RT @radhelper: T4 lots of knowledge about #AI and #ML that is available. one example: T4 lots of knowledge about #AI and #ML that is available. one example: https://t.co/Y7L9hApidY #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @drsxr @radhelper It took a looooooong time to get a government requirement for open data & articles from govt-funded studies. Can we get in at the beginning with some sort of policy requirement related to AI in medical imaging? #JACR |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: @pfanderson Unlike conventional software, the *data* generate the #AI tools. So large groups may be able to build their own systems (using published AI shells) from their own images. (It avoids FDA scrutiny...) #JACR |
| Radhelper PLLC @radhelper T4 we need to embrace these tools starting in medical school to encourage the next generation of radioloigsts to lead these efforts. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @Jenny_Writer: T4 Lots of information out there, including on the #ACRDSI blog https://t.co/V8LDK0kxs3 #JACR |
| Radhelper PLLC @radhelper RT @Rad_Partners: T4: T4: AI is a popular topic these days so there is plenty to read about it. @SIIM_Tweets holds an annual conference dedicated to AI, there are multiple lectures at radiology conferences (e.g., @RSNA), & much can be learned by speaking to the multitude of vendors in the space. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: T4 lots of knowledge about #AI and #ML that is available. one example: T4 lots of knowledge about #AI and #ML that is available. one example: https://t.co/Y7L9hApidY #JACR |
| Jenny Jones @Jenny_Writer RT @radhelper: T4 we need to embrace these tools starting in medical school to encourage the next generation of radioloigsts to lead these efforts. #JACR |
 | Priscila Falla @prisFallaBC RT @radhelper: T3 biggest impact will be population health. the normal/abnormal detection on imaging may be accessible to those that are not specialized to interpret imaging. so, they will know when something is wrong but may not know what it is #JACR |
| Rich Heller @reh3md RT @radhelper: T4 we need to embrace these tools starting in medical school to encourage the next generation of radioloigsts to lead these efforts. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @radhelper There are entire curricula of AI and data science courses available as MOOCs https://t.co/NoMjwUYfFz #JACR |
| Radhelper PLLC @radhelper we are scaring away too many talented students away from radiology and other specialties. it's a bit myopic in my opinion. #JACR |
| Bernadette Keefe MD @nxtstop1 T4 Via JACR T4: How does someone learn more about #datascience, #AI and informatics? Is there additional training or certification? #JACR |
| Vasanth @vasanthdrv T2 - existing deep learning networks are black boxes. nobody knows how they do what they do or what they learn.there is no reasoning that is an essential part of diagnoses made by humans. they will be better at the quantitative tasks that don't apply logical reasoning #JACR |
| Becky Haines @grebeccahaines RT @pfanderson: @radhelper There are entire curricula of AI and data science courses available as MOOCs https://t.co/NoMjwUYfFz #JACR |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: T4 lots of knowledge about #AI and #ML that is available. one example: T4 lots of knowledge about #AI and #ML that is available. one example: https://t.co/Y7L9hApidY #JACR |
| Bernadette Keefe MD @nxtstop1 RT @Jenny_Writer: T4 Lots of information out there, including on the #ACRDSI blog https://t.co/V8LDK0kxs3 #JACR |
| Charles Kahn, MD @cekahn @JACRJournal Nice intro on #AI and #DeepLearning in recent AJR issue by Bernie King. @RSNA soon to publish new journal on Radiology AI, and lest we forget... #JACR special issue on AI in #Radiology! #JACR |
| Radiology Partners @Rad_Partners T4 continued: @RadiologyACR has formed a Data Science Institute (DSI) @datascienceins with a mission that includes creating standards and use cases for AI tools, and educating radiologists about the space. #JACR |
 | blah blah blah @FalgunHChokshi @radhelper @grebeccahaines @sethmhardy @US_FDA They are slow. But will catch up. Usually need either an amazing use case or cases or a big tragedy to speed up action. #JACR |
| Radhelper PLLC @radhelper RT @Rad_Partners: T4 continued: T4 continued: @RadiologyACR has formed a Data Science Institute (DSI) @datascienceins with a mission that includes creating standards and use cases for AI tools, and educating radiologists about the space. #JACR |
| Radhelper PLLC @radhelper RT @cekahn: @JACRJournal Nice intro on #AI and #DeepLearning in recent AJR issue by Bernie King. @RSNA soon to publish new journal on Radiology AI, and lest we forget... #JACR special issue on AI in #Radiology! #JACR |
| Charles Kahn, MD @cekahn @radhelper Very myopic! We've had one well-known (nonrad) faculty member tell our med students that there's no future in radiology or pathology. #JACR |
| Bernadette Keefe MD @nxtstop1 Thread: The role of #ArtificialIntelligence in the practice of radiology — some thoughts from #jacr tweet chat attendees https://t.co/6cwxi5kdaK |
| blah blah blah @FalgunHChokshi @nxtstop1 .@coursera good resources and courses. @AndrewYNg ‘s https://t.co/gf2HF7Iggx #JACR |
| Radhelper PLLC @radhelper T4 another nice summary from @JACRJournal https://t.co/iFFGY2QB4H #JACR |
| Bernadette Keefe MD @nxtstop1 RT @FalgunChokshiMD: @nxtstop1 .@coursera good resources and courses. @AndrewYNg ‘s https://t.co/gf2HF7Iggx #JACR |
| blah blah blah @FalgunHChokshi @drsxr @radhelper @grebeccahaines @sethmhardy @US_FDA They NEED to happen. And soon. #JACR |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: T4 another nice summary from @JACRJournal https://t.co/iFFGY2QB4H #JACR |
| Jenny Jones @Jenny_Writer T4 A couple of good podcasts on #AI from the #RLI, including one from a pathologist's perspective https://t.co/xIprVRU3qv #JACR |
| JACR @JACRJournal Only a few minutes left, what are your closing thoughts on data science? #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @Rad_Partners: T4 continued: T4 continued: @RadiologyACR has formed a Data Science Institute (DSI) @datascienceins with a mission that includes creating standards and use cases for AI tools, and educating radiologists about the space. #JACR |
| Radhelper PLLC @radhelper @Vilavaite @pfanderson @cekahn #AI tools will have the ability to interpret from the raw data including sinograms and K-space #JACR |
| Bernadette Keefe MD @nxtstop1 RT @Jenny_Writer: T4 A couple of good podcasts on #AI from the #RLI, including one from a pathologist's perspective https://t.co/xIprVRU3qv #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @Vilavaite @cekahn This is just one example, in a narrow domain. It's happening in lots of fields. Sometimes they shut it down, sometimes the humans guide-on-the-side. #JACR |
| blah blah blah @FalgunHChokshi @drsxr @radhelper Doubt residency programs are going to teach anything close to #AI and #Machinelearning. Most people barely understand what this technology is really. 90% Hype right now. Searching for a platinum needle in a mountain of steel needles! #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: T4 another nice summary from @JACRJournal https://t.co/iFFGY2QB4H #JACR |
| Becky Haines @grebeccahaines Good model for education w @asset25 https://t.co/qE4vjk2yqn #jacr https://t.co/WyxNCXDf7o |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @FalgunChokshiMD: @nxtstop1 .@coursera good resources and courses. @AndrewYNg ‘s https://t.co/gf2HF7Iggx #JACR |
| JACR @JACRJournal RT @radhelper: T4 another nice summary from @JACRJournal https://t.co/iFFGY2QB4H #JACR |
 | Sergey Minaev @sminaev2015 RT @JACRJournal: Be sure to include T1, T2, etc in your tweets so we know which questions you're responding to. Don't forget to include #JACR https://t.co/Z71OJMq250 |
| PF Anderson @pfanderson@disabled.social @pfanderson Final thoughts: what can radiology learn from how AI has been implemented in other domains to avoid the pitfalls, and maximise the benefits for both clinics and patients? #JACR |
| Radhelper PLLC @radhelper @JACRJournal radiology must continue to embrace technology as it has historically. the data science and #AI generation is no different. we must evolve as a group and empower our community with tools that make sense and improve care #JACR |
| Jenny Jones @Jenny_Writer #AI will be another tool in the box for radiologists to strengthen their role as consultants in care, a la #Imaging3 #JACR |
 | Mark Punyanitya @markpun RT @FalgunChokshiMD: @drsxr @radhelper Doubt residency programs are going to teach anything close to #AI and #Machinelearning. Most people barely understand what this technology is really. 90% Hype right now. Searching for a platinum needle in a mountain of steel needles! #JACR |
| Radhelper PLLC @radhelper RT @grebeccahaines: Good model for education w @asset25 https://t.co/qE4vjk2yqn #jacr https://t.co/WyxNCXDf7o |
| Sergey Minaev @sminaev2015 RT @FalgunChokshiMD: @drsxr @radhelper Doubt residency programs are going to teach anything close to #AI and #Machinelearning. Most people barely understand what this technology is really. 90% Hype right now. Searching for a platinum needle in a mountain of steel needles! #JACR |
| Sergey Minaev @sminaev2015 RT @radhelper: T4 another nice summary from @JACRJournal https://t.co/iFFGY2QB4H #JACR |
| blah blah blah @FalgunHChokshi @radhelper @Vilavaite @pfanderson @cekahn Would be nice. But, honestly, all this #AI #innovation will be driven mainly by the private sector, with some collaboration by #academics, and government regulation. But these is a commercializeable #technology. Like it or not, if the product made doesn’t sell, no cigar. #JACR. |
| Bernadette Keefe MD @nxtstop1 @FalgunChokshiMD @drsxr @radhelper Don’t mind the hype if it means that here, on the ground, we are having serious conversations about preparing for the future practice and how implementation / validation should take place ~ #jacr. |
 | UtarSystems @UtarSystems RT @FalgunChokshiMD: @radhelper @Vilavaite @pfanderson @cekahn Would be nice. But, honestly, all this #AI #innovation will be driven mainly by the private sector, with some collaboration by #academics, and government regulation. But these is a commercializeable #technology. Like it or not, if the product made doesn’t sell, no cigar. #JACR. |
 | WioAI @WIOMAX_PA RT @nxtstop1: Thread: Thread: The role of #ArtificialIntelligence in the practice of radiology — some thoughts from #jacr tweet chat attendees https://t.co/6cwxi5kdaK |
 | Kingchi Engineering @kingchieng RT @FalgunChokshiMD: @radhelper @Vilavaite @pfanderson @cekahn Would be nice. But, honestly, all this #AI #innovation will be driven mainly by the private sector, with some collaboration by #academics, and government regulation. But these is a commercializeable #technology. Like it or not, if the product made doesn’t sell, no cigar. #JACR. |
| Radhelper PLLC @radhelper Thank you all for an extremely invigorating Tweet Chat! The next decade in medicine and radiology is going to be fascinating to say the least #JACR |
| Charles Kahn, MD @cekahn And that's a key lesson. #AI is just that: a tool that can be applied (for good or ill), and (wise) radiologists will use it to improve their ability to deliver care! #JACR https://t.co/rVdRRNon51 |
| PF Anderson @pfanderson@disabled.social @pfanderson @FalgunChokshiMD @radhelper Didn't you hear? BitCoin crashed. https://t.co/zRZ6EH7rmj #JACR |
| blah blah blah @FalgunHChokshi @radhelper @Vilavaite @pfanderson @cekahn As much as we radiologists talk about owning this, we don’t have the capital to own it. There is A LOT of money floating around both for #investments and #development. It’s THE thing for #venturecapital investments. Most will fail. But need just 1-2 big wins. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @radhelper: @Vilavaite @pfanderson @cekahn #AI tools will have the ability to interpret from the raw data including sinograms and K-space #JACR |
| Jenny Jones @Jenny_Writer RT @nxtstop1: @FalgunChokshiMD @drsxr @radhelper Don’t mind the hype if it means that here, on the ground, we are having serious conversations about preparing for the future practice and how implementation / validation should take place ~ #jacr. |
| Radhelper PLLC @radhelper @Vilavaite Probably won't change much. We still have to teach children mathematics in the era of calculator and computers. Same rules apply #JACR |
| Jenny Jones @Jenny_Writer RT @cekahn: And that's a key lesson. #AI is just that: And that's a key lesson. #AI is just that: a tool that can be applied (for good or ill), and (wise) radiologists will use it to improve their ability to deliver care! #JACR https://t.co/rVdRRNon51 |
| Charles Kahn, MD @cekahn Amazing discussion! Thank you all for joining the tweetchat. It was a ton o' fun! #JACR |
| Radhelper PLLC @radhelper RT @FalgunChokshiMD: @radhelper @Vilavaite @pfanderson @cekahn Would be nice. But, honestly, all this #AI #innovation will be driven mainly by the private sector, with some collaboration by #academics, and government regulation. But these is a commercializeable #technology. Like it or not, if the product made doesn’t sell, no cigar. #JACR. |
| Bernadette Keefe MD @nxtstop1 RT @cekahn: And that's a key lesson. #AI is just that: And that's a key lesson. #AI is just that: a tool that can be applied (for good or ill), and (wise) radiologists will use it to improve their ability to deliver care! #JACR https://t.co/rVdRRNon51 |
| Bernadette Keefe MD @nxtstop1 RT @radhelper: Thank you all for an extremely invigorating Tweet Chat! The next decade in medicine and radiology is going to be fascinating to say the least #JACR |
| Seth M Hardy, MD MBA exFACR @sethmhardy Truth. https://t.co/uT0OSicVDh #jacr https://t.co/rhS2TNIASS |
| PF Anderson @pfanderson@disabled.social @pfanderson @FalgunChokshiMD @radhelper @Vilavaite @cekahn If we don't insert some guidelines & requirements up front, we'll end up with an Elsevier of medical imaging, who own 95% of both images & analytic tools #JACR |
| Jenny Jones @Jenny_Writer Thanks for leading a great discussion @cekahn @radhelper! I learned a lot, as always. #jacr |
| blah blah blah @FalgunHChokshi @nxtstop1 @drsxr @radhelper Definitely, but we, who have limited #capital, will be adjunct factors in #AI’s tech dissemination and adoption curve. We can help validate and even implement. But it’s going to be products we buy to do this and that comes from industry. Not what many want to hear. #JACR |
| Becky Haines @grebeccahaines @vasanthdrv @ACRRFS DK @judywawira @jeremyphoward can u advise? #jacr |
| blah blah blah @FalgunHChokshi @pfanderson @radhelper #Bitcoin crashes many times and then bubbles up. This is not news. #JACR |
| blah blah blah @FalgunHChokshi RT @sethmhardy: Truth. https://t.co/uT0OSicVDh #jacr https://t.co/rhS2TNIASS |
| Charles Kahn, MD @cekahn There are many kinds of "capital", and knowledge is one. #AI systems are increasingly available free (github), computing grows cheaper daily in the cloud. For systems based on data (images), a large image collection is capital, too! #JACR https://t.co/KiK6HbHnWA |
| blah blah blah @FalgunHChokshi @nxtstop1 @drsxr @radhelper Frankly, what I love about all this hype is that we are engaging and educating our #radiology colleagues about this technology to some degree and that people are talking about it and not just living in denial. But we must temper the talk with reality. Hard truths. #JACR |
| Becky Haines @grebeccahaines @FalgunChokshiMD @nxtstop1 @drsxr @radhelper Educated radiologists will be key in developing tools as well as implementing in clinical practice. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @FalgunChokshiMD @radhelper @Vilavaite @cekahn Can we at least ensure SOME open genuinely useful tools for analysis (like mothur in microbiome work) and some mandated diversity & competition in corporate controls? I think #JACR shd sponsor a software carpentry training session https://t.co/vhI9rCPJ4T |
| blah blah blah @FalgunHChokshi @cekahn Agreed, but the big drivers of #AI adoption will be enterprise level platforms, which, for better or worse, will be what percolate the technology dissemination of this technology. Unfortunately, we saw this with the #EHR, which sucks. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson @FalgunChokshiMD @radhelper LOL! Read the title of the article. It was about how that crash was different from the normal Bitcoin crash. Was getting a lot of buzz last week. #JACR |
| Becky Haines @grebeccahaines This sounds like a job for @ruthcarlosmd #jacr https://t.co/B9hJKUlHMP |
| Mark Punyanitya @markpun RT @grebeccahaines: @FalgunChokshiMD @nxtstop1 @drsxr @radhelper Educated radiologists will be key in developing tools as well as implementing in clinical practice. #JACR |
| blah blah blah @FalgunHChokshi @drsxr @radhelper @Vilavaite @pfanderson @cekahn Comelletly agree. But it’s the lowest common demolition that can get monetized and ROI realizes faster. We don’t like it, but I believe that’s how it’ll happen. #JACR |
| PF Anderson @pfanderson@disabled.social @pfanderson RT @cekahn: There are many kinds of "capital", and knowledge is one. #AI systems are increasingly available free (github), computing grows cheaper daily in the cloud. For systems based on data (images), a large image collection is capital, too! #JACR https://t.co/KiK6HbHnWA |
| Y, like yellow, E @jackyeclayton RT @cekahn: There are many kinds of "capital", and knowledge is one. #AI systems are increasingly available free (github), computing grows cheaper daily in the cloud. For systems based on data (images), a large image collection is capital, too! #JACR https://t.co/KiK6HbHnWA |
| blah blah blah @FalgunHChokshi @grebeccahaines @nxtstop1 @drsxr @radhelper Hope so. #JACR |
| blah blah blah @FalgunHChokshi @drsxr @nxtstop1 @radhelper #diversity of though is an untapped asset! Agreed. #JACR |
| Bernadette Keefe MD @nxtstop1 @FalgunChokshiMD @drsxr @radhelper Engagement & education is key; it IS the future, for us, and for many fields. That the monetary benefits may accrue elsewhere –––common. Media hype can be tempered thru better communication w/#healthcare media (establish ties w/ #journalists, answer Qs, inform)They r open! #JACR |
| blah blah blah @FalgunHChokshi @drsxr @nxtstop1 @radhelper **thought. Love autocorrect #JACr |
| PF Anderson @pfanderson@disabled.social @pfanderson @Jenny_Writer @radhelper #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| PF Anderson @pfanderson@disabled.social @pfanderson @pfanderson @drsxr @reh3md #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| PF Anderson @pfanderson@disabled.social @pfanderson @cekahn @JACRJournal #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| PF Anderson @pfanderson@disabled.social @pfanderson @grebeccahaines #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| PF Anderson @pfanderson@disabled.social @pfanderson @nxtstop1 @Rad_Partners #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| JACR @JACRJournal The #JACR Data Science collection is free online now https://t.co/tvHKEMIFNd https://t.co/1wIKn7Jggx |
| Bernadette Keefe MD @nxtstop1 RT @pfanderson: @nxtstop1 @Rad_Partners #Storify story "#JACR Chat on #AI, #DataScience, and #DeepLearning in Medical Imaging " https://t.co/Ak8hAos0CW |
| blah blah blah @FalgunHChokshi @nxtstop1 @drsxr @radhelper For sure, we rads must be engaged, or we will again go the way of hospitals and healthcare organizations in he 1980’s with HMO administrators controlling things. Wait, they still do!! #JACR |
| Radiology Partners @Rad_Partners It’s an exciting time in #radiology. #AI is another step in the evolution of our profession – not a threat, but an opportunity. Thanks for hosting this tweet chat, @radhelper @cekahn @JACRJournal. It's always great to hear thoughts of others in the field. #JACR |
| blah blah blah @FalgunHChokshi @drsxr @cekahn Yes. Any threat to their market position will@elicit a reaction. #AI is so hyped up that no matter what, they have to react. But reaction is very different than #vision. We can help add #vision and practical #utility. That will sell! #JACr. |
| blah blah blah @FalgunHChokshi @drsxr @radhelper @Vilavaite @pfanderson @cekahn I’m already in the ring!! Come on, need more folks. #JACR |
#JACR content from Twitter.