Lab's weekly topic - week 49, 2023 - What's limitations and possible solutions for a new AI tools to detect blood poisoning?


A New, Transparent AI Tool May Help Detect Blood Poisoning

It is a promising application using AI for medical services. It is a hot topic using AI to help people’s health issues. This novel algorithm can scan electric records and may reduce sepsis death, but the wide application still needs more time.

  1. A solution based on machine learning was developed to help physicians detect sepsis earlier, which could save lives. Johns Hopkins Hospital developed the solution and has been deployed at three different hospitals for a try. Electronic records are used directly for an automation alert. 
  1. User side: Physicians could not have a high willingness to try a new tool they are not familiar with.
  2. Solution side: The solution depends on the electronic records. It has missing fields and values in the records, which could block the solution.
  3. Model side: It is better that we can explain the results predicted by the model detecting sepsis. Deep learning usually has a better performance compared with the shallow, traditional machine learning methods. While it is a black-box method and needs further effort to understand the prediction.
  4. Market side: It could have an interest conflict when the solution is deployed widely.
Possible fix and future development
  1. For the missing fields in electronic records, we could have more training and fill the fields as much as possible. Also, we can develop a model that can handle some missing values for the solution.
  2. We could exploit the high performance of deep learning for the alteration of sepsis, which could improve the accuracy significantly. For the explanation, it could be helpful that some research can be used to explain deep learning models.


Popular posts from this blog

Lab's weekly topic - week 50 2023 - Detailed explanation for encoder of U-Net

Lab's recommendation: Awesome books to learn machine learning and AI (continue updating)