064 #amld2020 #health
Tweets related to Artificial Intelligence and Public Health at the EPFL Applied Machine Learning Days 2020.
There was a lot of inspiring content this year related to the intersection of Artificial Intelligence and Public Health at the Applied Machine Learning Days conference in Lausanne, Switzerland, which I covered to some extent in my previous blog post. See also the interviews on this topic with Max Tegmark (MIT), Marinka Zitnik (Harvard), Gavin Brown (U.Manchester), Aleksandra Kovachev (Delivery Hero), and Wessel Valkenburg & Aziza Merzouki (U.Geneva) at Heidi.news
In roughly chronological order, my focus here is on live coverage of the A.I. & Health track organised by Elaine Nsoesie. The photo above from the startup pitch of @Retin_AI.
Packed room for the AI & Health track at Applied Machine Learning Days in Lausanne. Thanks @ensoesie @ @marcelsalathe! Happy that ISI Foundation is represented here by @mtizzoni and @danielapaolotti pic.twitter.com/9cXo1okrlQ
— Ciro Cattuto (@ciro) January 27, 2020
One of the big topics of the moment is, of course, the coronavirus outbreak.
Results from @Crowdbreaks_org tracking of tweets on #Wuhan virus #ncov: 10 million tweets/day at the moment! Presented by @marmuel_ from @epfl_en Digital Epidemiology Lab at #AMLD2020 pic.twitter.com/olG8IAD8aL
— Applied Machine Learning Days (@appliedmldays) January 27, 2020
Over 10K retweets… The data’s relatively clear that @cnni is a major source of misinformation around the very unlikely “snake source” of coronavirus.
— Marcel Salathé (@marcelsalathe) January 26, 2020
(Btw @WHO as source is wrong)
For a good update: https://t.co/x1DKAYc8GR pic.twitter.com/u9XyE7MQ53
At least 48M people in 15 Chinese cities are effectively quarantined in China's struggle to stop spread of 2019-nCoV.
— Helen Branswell (@HelenBranswell) January 25, 2020
The mind reels.https://t.co/cJLENhp8e0
There's definitely fuel in the fire. Let's get back into that "packed room":
Co-Founder of @evidation, @calimagna talked about how to use Person-Generated Health Data to develop digital measures. They allow for example to capture outcomes relevant to the patient and they can be use to predict flu outbreaks! pic.twitter.com/lQz56aBH5i
— EPFL IC (@ICepfl) January 27, 2020
@calimagna presentation on monitoring flu and other health conditions #AMLD2020 Health & AI track pic.twitter.com/Sh25c6snzD
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 27, 2020
@calimagna speaking on Developing Digital Measures from Person-Generated Health Data at the AI & Health track #AMLD2020 pic.twitter.com/fuqhffUq0B
— Elaine Nsoesie, PhD (@ensoesie) January 27, 2020
Dr. Gonzalez Hernandez @gracielagon from @penn uses twitter data from pregnant women to study birth defects #DigitalEpidemiology #AMLD2020 #AIandHealth pic.twitter.com/dbgV2Eh897
— Applied Machine Learning Days (@appliedmldays) January 27, 2020
My @UPennIBI Health Language Processing Center webpage https://t.co/xPmixAyVPN has inks to our publications and freely available annotated datasets, and to the upcoming #SMM4H workshop shared tasks. We are happy to help!
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 30, 2020
If you want to find out more about our work on monitoring #infectiousdiseases in online news articles using #NLProc, make sure to visit the (AI & Health) poster session @appliedmldays #AMLD2020 pic.twitter.com/wE40MttQ1F
— auss (@Auss_Abbood) January 27, 2020
A novel technique for mapping MedDRA terms to ICD terms (medical database standard) with the generation of word embeddings and context analysis. Here, the corpus used was a record of diagnoses provided by a public hospital in the US. @Roche pic.twitter.com/XQBJM9KLIB
— EPFL IC (@ICepfl) January 27, 2020
Stance analysis for text classification on social media to determine health trends (vaccines, CRISPR...) need continuous annotation of data and model retraining to avoid concept drift. The collaborative platform @crowdbreaks_org offers such infrastructure
— EPFL IC (@ICepfl) January 27, 2020
@danielapaolotti presents how @Influenzanet and other national flu platforms across the world collect data from the population by letting them enter their symptoms and environmental and geographical details to predict and follow influenza trends.
— EPFL IC (@ICepfl) January 27, 2020
@danielapaolotti at #AMLD2020 AI & Health detailed report on the analysis of self-reported ILI (influenza like illness) on the influezanet platform from multiple countries @MatthewScotch pic.twitter.com/3hhmP72n6p
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 27, 2020
@marmuel_ at #AMLD2020 AI & Health track - very interesting work on stance (sentiment) analysis for tracking health trends. pic.twitter.com/sWMd9d9zdU
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 27, 2020
For those attending the #AMLD2020 AI & Health track, this is the blog I referred to. The idea is: even if the cookies are on the counter, it doesn't mean you can eat them! Responsible use of data and putting yourself in your subject's shoes goes a long way @ensoesie @calimagna https://t.co/W6FUTk8Ly2
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 27, 2020
Facilitating clinical and pharmaceutical research on #AMD through a #MachineLearning software that analyses disease #biomarkers in #OCT scans. Agata Mosinska, @Retin_AI, just concluded her presentation at the #AMLD2020. @EPFL #AI #ophthalmology #ARTORG #spinoff @unibern pic.twitter.com/8zHCTPmoPj
— Monika Kugemann (@MKugemann) January 28, 2020
The beauty and complexity of data on menstrual cycle! @LauraSymul presenting at @appliedmldays and I think inspiring other with her energy to research menstrual health. As she points out.. no menstruation no human race 😊 #AIandHealth #AMLD2020 pic.twitter.com/m28lCyriOE
— Talia Salzmann (@TaliaSalzmann) January 27, 2020
@mtizzoni at #AMLD2020 AI& Health Track - interesting work on using health forum data to predict cesarean section status @anna_wexler pic.twitter.com/RjtZ9ulycD
— Graciela Gonzalez-Hernandez, PhD (@gracielagon) January 27, 2020
Maroussia Roelens from @unige sharing challenges and potential of using ML to predict epidemics using digitized information from Integrated Management of Childhood Illness in Burkina Faso #IMCI #globalhealth #AIHealth pic.twitter.com/LKCdLnkbrO
— Applied Machine Learning Days (@appliedmldays) January 27, 2020
#ML for cities in #Africa is the topic of @nunuska's (@GoogleAI) talk at #AIandCities. #AMLD2020 pic.twitter.com/FySCYkmAyX
— Konstantin Klemmer (@kklmmr) January 28, 2020
If you still think human brains are best suited for all medical diagnosis tasks, think again (via @KyleTheDavid and @kklmmr).
Closing thoughts:
One thing I would have told my younger self was you are not alone.
— Elaine Nsoesie, PhD (@ensoesie) January 27, 2020
You don’t have to solve this problem on your own. All you have to do is pick one brick and lay it down. Somewhere in the world someone is also laying a brick.
Don’t stay safe. Stay free. #AMLD2020 https://t.co/kG55iKfulg
With best wishes to all involved. Keep me posted on any other tweets or material you'd like to share.
