6.871/HST.956 Week of 2/10 Reflection Questions
For this week's reflection questions, we request that you watch the following two videos:
1. ML for Cardiology: https://youtu.be/UVUq0-32iHw
2. ML for Mammography: https://youtu.be/ZXwv4dyCv5I

And answer the following questions. Please answer each question in 100-200 words. These are open-ended questions, and you will be graded not on agreement with the course staff's opinions as to the answers, but rather based on the thoughtfulness, clarity, and cohesion of your answer. Let us know on Piazza if you have any questions!
Sign in to Google to save your progress. Learn more
Email *
MIT Kerberos ID *
ML for Cardiology 1: Rahul says at the beginning of the talk that "these ideas really need to make it into practice," and, throughout the talk, highlights several of the key barriers to doing so. Based on the video and our lectures so far, choose one of the barriers Rahul mentions, explain it briefly, then provide additional commentary on your perspective on this barrier (e.g., thoughts on solutions, analogies to other fields, reasons why this barrier might sensibly exist in our ecosystem in the first place, etc.). *
ML for Cardiology 2: How much farther did Rahul have to go in “validation” of a model than you were expecting / might be required in a non-medical setting? *
Ml for Mammography 1: Compare this talk to Rahul’s ML for Cardiology talk, specifically on the axes of impact and barriers to impact. Why (or why not) is the application described here a good case for the deployment of AI for medicine according to Rahul’s taxonomy? *
ML for Mammography 2: In this talk, they identify a number of specific, technical, modelling challenges that (while only identified here in the context of mammography), are emblematic of medical data on the whole. Comment on these briefly. Which were or were you not expecting? In what ways are these technical challenges also found in other fields? Are there any broad commonalities between these challenges? Etc. *
A copy of your responses will be emailed to the address you provided.
Submit
Clear form
Never submit passwords through Google Forms.
reCAPTCHA
This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy