top of page
Search

1 | Covid & AI - An Unlikely Match?

  • Writer: Romil Shah
    Romil Shah
  • May 11, 2020
  • 2 min read

Updated: Apr 5, 2022

Research Spotlights

AI & Covid

For the last decade, there has been so much talk of the AI revolution in healthcare. Touted by VCs as the doctor replacement, $2 billion has been invested in AI healthcare companies in 2019 (projected to increase in 2020). As someone on the front lines, any actual implementation of ML has been scarce even though plenty of studies have demonstrated its value. COVID-19 sped up the adoption curve. Across the country, AI has been used to diagnose COVID, predict a decline in patients hit hardest, and develop new drugs that can help COVID patients. For the first time, it’s not just research; machine learning is being used daily to actually help clinical decision making. We are seeing AI help clinicians at a time when eyes across the world are watching medicine closely.


A Computer is your best Chart Checker

In healthcare, data is power. Medical data is roughly a $60 billion dollar business each year. Individual medical records are worth as much as $1000 per record. In surgical patients, most of that data is hiding behind long free-text notes where doctors describe their indications, the intricacies of what they found in the surgery, and the complex decision making. I just published a paper on using an NLP (natural language processing) algorithm to structure and automatically collect important information such as what a patient’s anatomy looks like intra-operatively, the surgical approach, implant type and size, and complications from surgeon notes. My research results suggest that algorithms like this will ultimately lead to better quality / quantity of surgical data, thus improving patient care.


Monthly Links


One Big Question

One of the leading pediatric surgeons talked about how it is easier to talk to kids about surgery and even examine them over the webcam because in the comfort of their homes.


The widespread adoption of telehealth will lead to several unknown pros and cons – what will stay and what will go is yet to be determined. A lot of great things have came out of this space recently - we need to make sure to not let it go to waste.



 
 
 

Comments


bottom of page