Hyunkwang Lee

Ph.D. in EE at Harvard SEAS (@hseas)

[Hyunkwang Lee @hyunkwang_lee]

@IAmSamFin Congrats!

[Hyunkwang Lee @hyunkwang_lee]

@mmdmansouri @MontefioreNYC So proud of you! But please be safe!

[Isaac Kohane @zakkohane]

Just heard from colleague that Deep Learning in Medicine course @harvardmed @HarvardDBMI https://t.co/nwqnMa41Qe was oversubscribed this year even more than last & going well by distance learning during the pandemic. Kudos! CC @AndrewLBeam @kunhsingyu

[Hyunkwang Lee @hyunkwang_lee]

@IAmSamFin Good luck! Sam. let me know if i can join you defense via Zoom

[Andy Biotech @AndyBiotech]

#COVID19 South Korea daily new cases declines again to 76, lowest since 2/22

- Total cases: 8,086➡️8,162

- Deaths: 72➡️75

- Total tested: 261,335➡️268,212

Very impressive how huge amount of work on broad testing + tracing + isolating is paying off against the #coronavirus!

[Alexandre Cadrin-Chênevert @alexandrecadrin]

When AI > AI+human, we face important ethical questions.

"In the reader study, the performance level of AI was 0.940, significantly higher than that of the radiologists without AI assistance (0.810). With the assistance of AI, radiologists' performance was improved to 0.881."

[Jeremy Howard @jeremyphoward]

TIL Google invented universal language model pre-training with BERT.

[Google DeepMind @GoogleDeepMind]

In collaboration with @GoogleHealth, we've developed an AI model that can exceed human performance at detecting breast cancer in screening mammograms.

[Alan Karthikesalingam @alan_karthi]

Delighted to kick off 2020 w/our research published @nature. A massive achievement for our team, working over the last few years with @CRUK & many collaborators including patients. A great step to better breast cancer care: now on to clinical validation. https://t.co/7AXUKFSsXW

[Nature Portfolio @NaturePortfolio]

Editorial: Many PhD students and postdoctoral researchers are overworked and overstressed — and their mental health is suffering because of it. https://t.co/XMh95WptNI

[Hyunkwang Lee @hyunkwang_lee]

@IAmSamFin @AndrewLBeam Thank you so much and I am sorry to hear that. Hope you feel better and we will catch up soon (and talk about our ongoing project 😀)

[Hyunkwang Lee @hyunkwang_lee]

Successfully defended my PhD dissertation today. Wouldn't have been possible without the support of my wife, family, friends, my dissertation committee (David Parkes, Demba Ba, @AndrewLBeam)and all the collaborators. Thank you all!

[Hyunkwang Lee @hyunkwang_lee]

Thank you for being my committee member! You made my day so special! Gio was happy to meet you!

[Andrew Beam @AndrewLBeam]

Congratulations Dr. Lee! @hyunkwang_lee gave an amazing thesis defense this morning. So happy to have been a part of his committee and hear about his impressive body of work.

[Hyunkwang Lee @hyunkwang_lee]

@AndrewLBeam @maithra_raghu I also really enjoyed this paper when it was uploaded in arxiv! I wanted to do the same analysis for the GrayNet pretrained model if i could get access...

[RadNet @RadNetImaging]

Another #RSNA meeting is officially behind us. We wanted to take a moment to recognize all of our colleagues, partners, vendors, prospects, and especially the team at @RSNA  for contributing to another successful week. From our team to all of you, thank you! #RSNA2019

[Maithra Raghu @maithra_raghu]

How does transfer learning for medical imaging affect performance, representations and convergence? Check out the blogpost below and our #NeurIPS2019 paper https://t.co/jNesB0ODBD for some of the surprising conclusions, new approaches and open questions!

[Hyunkwang Lee @hyunkwang_lee]

This is really useful to think about the liability for physicians, discussed by JDs.

[Andrew Beam @AndrewLBeam]

A question that always comes up when discussing AI in medicine:

"Who is responsible when a medical AI system gets a diagnosis wrong?"

Finally, a useful summary just published in
@JAMA_current: https://t.co/AxP0HvICUh

[Scientific Reports @SciReports]

Using #ArtificialIntelligence to read chest radiographs for #tuberculosis detection: A multi-site evaluation of the diagnostic accuracy of three deep learning systems #Radiography https://t.co/z8Fzky03Eh

[Eric Topol @EricTopol]

No matter how experienced the radiologist, accurate interpretation of malignant lung nodules on chest X-rays was ameliorated by #AI #openaccess https://t.co/FAVsqU6HnG
@radiology_rsna @RSNA

[Bálint Botz @BalintBotz]

Intriguing: "most #deeplearning algorithms have been developed using reconstructed, human-interpretable medical images. ... this study showed SinoNet performed better for pathology detection directly from sparse sinograms than reconstructed images"
https://t.co/F4EnVcLIUj