Daniel Raff

@raffrica

Family Medicine Resident. Data Scientist. Co-Founder

Vancouver, Canada
Vrijeme pridruživanja: kolovoz 2011.

Tweetovi

Blokirali ste korisnika/cu @raffrica

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

  1. Prikvačeni tweet
    5. srp 2018.

    Wrote about what I learned this past year during UBCs Masters of Data Science:

    Poništi
  2. 5. sij

    A well written and accessible summary of the ML4H session at NeurIPS

    Poništi
  3. 4. sij

    Just learned this is called Eroom's Law (Moore's backwards):

    Prikaži ovu nit
    Poništi
  4. 4. sij

    What a depressing quote: "The number of drugs receiving regulatory approval per billion dollars of investment has been consistently halving every nine years, while the number of medical publications has been doubling with the same periodicity (Fig. 1)."

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    25. stu 2019.

    Thanks , , for great conversation. We need to focus more on data than on fancy models. I urge everyone to think about the following: 1/3. Learning health system needs data. We can’t benefit from each other's data unless we share.

    Prikaži ovu nit
    Poništi
  6. 23. stu 2019.

    Caveats: 1) Smart Compose is a double-edged sword and will need to be studied. Could introduce inaccuracies but on the other hand could improve data quality. 2) Search will be highly variable based on local implementations because of the logistics of where the data is hosted,etc

    Prikaži ovu nit
    Poništi
  7. 23. stu 2019.

    - auto-populating relevant lab results based on what I type.. this could save me so much time. My current workflow is going back and forth with up to 1 sec load times and 2-3 clicks between each back-and-forth. Maddening.

    Prikaži ovu nit
    Poništi
  8. 23. stu 2019.

    - the level of search we are accustomed to from our non-medical lives but in the chart

    Prikaži ovu nit
    Poništi
  9. 23. stu 2019.

    What it does well: - visual representation of lab results and changes - version control to highlight the diffs between copy-forwarded notes while ensuring they’re still readable (working on the information ‘starvation in the midst of plenty’ problem).

    Prikaži ovu nit
    Poništi
  10. 23. stu 2019.

    This short demo works on so many problems that frustrate me on a daily basis. Here's hoping it will be available in Canada and in a primary care setting.

    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet

    Much of what’s being sold as "AI" today is snake oil. It does not and cannot work. In a talk at MIT yesterday, I described why this happening, how we can recognize flawed AI claims, and push back. Here are my annotated slides:

    Prikaži ovu nit
    Poništi
  12. 1. stu 2019.
    Poništi
  13. 25. lis 2019.

    It’s still early days for algorithmic healthcare, but we need to be better. We need to anticipate biases in datasets and features, and have built in post-market surveillance for biases. How long has this been on the market before the bias was detected?

    Poništi
  14. 23. lis 2019.

    This thread is gold and should be required reading for anyone with an upcoming ID rotation

    Poništi
  15. proslijedio/la je Tweet
    21. lis 2019.

    This is how you do AI research, without the hype... Independent evaluators: ✅ Comparison of different models: ✅ Multi-site data at different disease prevalence: ✅ STARD criteria: ✅ Blinding: ✅ Next step: Prospective testing

    Poništi
  16. 15. lis 2019.

    Seems obvious in retrospect to put pediatric eye screening tools in the hands of parents (who already have the basic tool in their hands aka smartphones)

    Poništi
  17. 8. svi 2019.

    I love this article on how your food choices affect climate change because it weaves in evidence while being actionable with links for recipes.

    Poništi
  18. 4. velj 2019.

    “A disorder like depression is many, many illnesses. It’s like fever. We don’t go beyond just taking someone’s temperature and saying ‘this person has a fever, therefore we need something to bring down the fever.’ So everybody goes on an antidepressant.”

    Poništi
  19. proslijedio/la je Tweet
    27. sij 2019.
    Poništi
  20. 18. pro 2018.

    My old teacher wrote this up and it is fire

    Poništi
  21. proslijedio/la je Tweet

    A bit of an exaggeration to suggest that it saved a life with the information presented, but I bet we'll see a lot of these.

    Poništi

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:

    ·