Matthew Lungren MD

@mattlungrenMD

Faculty Co-Director Interventional Radiologist Machine Learning Committee &

Palo Alto, CA
aimi.stanford.edu
Vrijeme pridruživanja: ožujak 2011.
110 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @mattlungrenMD

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @mattlungrenMD

  1. Prikvačeni tweet
    5. srp 2018.

    Will Radiologists lose jobs to AI?

    1 reply 10 proslijeđenih tweetova 32 korisnika označavaju da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    29. sij

    My 1st -guided echocardiography today (fun & by democratizing acquisition it will change the field) What does that mean? You put the transducer on the patient's chest ->the AI tells you how to properly position it and then automatically captures video images when spot on.

    21 reply 295 proslijeđenih tweetova 894 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. 25. sij

    Terrific analysis! Really love how open source medical data and code led to new insights and understanding for others - and bonus that paid it forward with extensive breakdown notes and code! Thank you!

    After a dismal performance in the RSNA Intracranial Haemorrhage competition I pulled apart the 2nd place solution to see what they had done so right. Post 3 of 4 now up with jupyter notebooks :D Blog series: Notebooks:
    Prikaži ovu nit
    1 reply 4 proslijeđena tweeta 18 korisnika označava da im se sviđa
    Poništi
  4. 23. sij

    Announcing new medical imaging dataset for an important clinical use case - and our first for ultrasound/ cardiac echo. Hope this release will help ML researchers engage in medical applications and improve diagnostics for patients worldwide. More to come in 2020 🤖

    We're super excited to announce the release of our latest shared dataset with >10k expert-labeled echocardiogram videos! Check it out at Special thanks to
    81 proslijeđeni tweet 248 korisnika označava da im se sviđa
    Poništi
  5. proslijedio/la je Tweet
    20. sij

    Interesting paper showing how dozens of studies have accidentally leaked large amounts of data from train->test dataset, by duplicating data items prior to doing a random split.

    Our paper "Overly Optimistic Prediction Results on Imbalanced Data: Flaws and Benefits of Applying Over-sampling" has been published on arXiv: What did we do? A thread... (1/6)
    Prikaži ovu nit
    21 reply 98 proslijeđenih tweetova 528 korisnika označava da im se sviđa
    Poništi
  6. proslijedio/la je Tweet
    19. sij

    The rise of robot radiologists? No. A classic quote by : " won't replace radiologists, but radiologists who use AI will replace radiologists who don't." In new : Also : by

    10 replies 130 proslijeđenih tweetova 270 korisnika označava da im se sviđa
    Poništi
  7. proslijedio/la je Tweet
    16. sij

    Cows make milk. They milk themselves. Other cows check the milk (for free). Cows - get this - PAY THE FARMER to take the milk away. Then the farmer (you won't believe this, honestly) sells the milk *back to the cows.*

    Explain academic publishing to me like I am Five
    2.880 proslijeđenih tweetova 7.619 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    12. sij

    Just wrapped up the breakout session at with this rockstar group! you were dearly missed! – mjesto: UCSF Smith Cardiovascular Research Building

    1 proslijeđeni tweet 29 korisnika označava da im se sviđa
    Poništi
  9. proslijedio/la je Tweet
    11. sij

    Oh and BTW, if you want to be one of the first to see this material, then you'll want to join our Deep Learning course in SF, running for 8 weeks from March 17, 2020. Apply here:

    In our forthcoming book & course, you'll learn how to build a real deep learning web app from scratch, including downloading images using Bing's API. You'll also learn what can go wrong! (h/t )
    12 replies 38 proslijeđenih tweetova 150 korisnika označava da im se sviđa
    Poništi
  10. proslijedio/la je Tweet
    7. sij

    I completed my 1st data science project ~30 years ago. Since then I've been continuously developing a questionnaire I use for all new data projects, to ensure the right info is available from the start. I'm sharing it publicly today for the first time.

    55 replies 1.072 proslijeđena tweeta 3.714 korisnika označava da im se sviđa
    Poništi
  11. proslijedio/la je Tweet
    5. sij

    I wrote an opinion piece on the recent Nature article on AI for breast cancer screening. I hope it puts the results in perspective. Some of the media coverage has been a bit overhyped.

    Artificial intelligence shows promise for breast cancer screening. Learn more in a terrific N&V by
    7 replies 44 proslijeđena tweeta 136 korisnika označava da im se sviđa
    Poništi
  12. proslijedio/la je Tweet
    4. sij

    Comparison of the breast AI paper against the Editorial Board recommendations for Assessing Radiology Research on Artificial Intelligence

    14 replies 130 proslijeđenih tweetova 267 korisnika označava da im se sviđa
    Poništi
  13. proslijedio/la je Tweet
    2. sij
    Odgovor korisniku/ci

    "[A]uthors are required to make materials, data, code, and associated protocols promptly available to readers without undue qualifications." Unless it's a super glamorous paper that will get tons of press coverage for Nature, in which case don't worry.

    3 proslijeđena tweeta 44 korisnika označavaju da im se sviđa
    Poništi
  14. proslijedio/la je Tweet
    3. sij

    RADIOLOGY Editorial Board: Methods to assess AI research.

    56 proslijeđenih tweetova 119 korisnika označava da im se sviđa
    Poništi
  15. proslijedio/la je Tweet
    2. sij

    Our code and the weights of trained networks are publicly available for that reason. I hope that sharing the code will become a standard in the future. I believe that journals, especially the big ones, should require it. 7/N

    38 proslijeđenih tweetova 227 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    2. sij

    Congrats to Google, but let’s not forgot the team from NYU who last year published better results, validated on more cases, tested on more readers, and made their code and data available. They just don’t have the PR machine to raise awareness.

    Great news! Our paper "Deep Neural Networks Improve Radiologists' Performance in Breast Cancer Screening" has been published at ! This marks the end of a 3-year long endeavor, at and , to demonstrate the potential of CNNs on this task. (1/8)
    Prikaži ovu nit
    15 replies 348 proslijeđenih tweetova 1.111 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    2. sij

    Path to a Free Self-Taught Education in !

    36 proslijeđenih tweetova 26 korisnika označava da im se sviđa
    Poništi
  18. proslijedio/la je Tweet
    2. sij
    Odgovor korisnicima i sljedećem broju korisnika:

    Too bad that others in research community cannot validate/benefit from results! Does licensing justify why can't release de-identified data for resarch? If so why publish? UK data was public - data is the critical need to advance field!

    1 reply 2 proslijeđena tweeta 14 korisnika označava da im se sviđa
    Poništi
  19. proslijedio/la je Tweet
    1. sij
    Odgovor korisnicima i sljedećem broju korisnika:

    The enriched set might contribute to poor performance. In the real clinical setting, experts read with certain targets in mind, like recall rate of 10% and cancer detection rate 5 in 1000. The study had 25% cancer prevalence, and did not tell the readers. Way different!

    1 reply 4 proslijeđena tweeta 8 korisnika označava da im se sviđa
    Poništi
  20. 1. sij

    ‼️Important to be clear about difference between FINDINGS and DISEASES on chest X-rays... Yes of course there are thousands of diseases - but it is possible to identify nearly all of them recognizing only 14 findings. Analogy: 26 simple letters can compose millions of words!

    "Even if deep-learning 'AI' systems have begun to identify 14 diseases on chest radiographs, that's 14 disorders down, 2,944 to go."
    1 reply 15 proslijeđenih tweetova 31 korisnik označava da mu se sviđa
    Poništi
  21. 31. pro 2019.

    Credit (and kudos) for creating this great resource: Sjoerd Kerkstra

    3 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi

@mattlungrenMD još nije objavio nijedan Tweet.

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·