Zoi Capital | Digital Health – AI in Healthcare – Venture Capital

AI Doctor Book

How Do You Get the Data?

  • Ronald M. Razmi, MD
  • October 26, 2021

Data to AI is what oil has been to the world economy for the last century. Unless you have plenty of it, you will not be able to get far! It’s not enough to get a hold of sufficient data to train a model, you also need to have other data sets of the same type to validate your model on. Then, you need to have the right data flowing into your model once it is launched in the real-world. What does that mean? It means that if your model requires data that is not readily available in the setting that the model is used for, it will not be able to perform as intended. If it does not have access to the input data it needs to come up with predictions or do its pattern recognition, it will not be of much use. This is especially salient in the world of healthcare, where data is siloed and often not inter-operable.

Of course, being able to get a hold of large datasets means that companies will be making deals with medical centers. That may come with its own set of issues and challenges. A healthcare artificial intelligence start-up founded by three insiders at Memorial Sloan Kettering (MSK) Cancer Center, Paige.AI,  was able to secure large datasets of pathology slides for its pathology AI products. The company has an exclusive deal to use the cancer center’s vast archive of 25 million patient tissue slides, along with decades of work by its world-renowned pathologists.

 

The arrangement has sparked considerable turmoil among doctors and scientists at Memorial Sloan Kettering, which has intensified in the wake of an investigation by ProPublica and The New York Times into the failures of its chief medical officer, Dr. José Baselga, to disclose some of his financial ties to the health and drug industries in dozens of research articles. He resigned after the controversy broke out, Hospital pathologists have strongly objected to the Paige.AI deal, saying it is unfair that the founders received equity stakes in a company that relies on the pathologists’ expertise and work amassed over 60 years. They also questioned the use of patients’ data — even if it is anonymous — without their knowledge in a profit-driven venture. This resulted resulted in many resignations and people giving up their equity stake and pathologists protesting the deal.
This represents the type of problems that can emerge as private companies and medical centers try to use existing data to form partnerships and launch new ventures. Objections are raised about using existing patient data (without their consent) to launch for-profit ventures and many times,  medical center insiders get equity stakes in these companies.
This is before you even worry about the availability of the data in most settings once the model becomes operationalized in the real-world!!
kindersco

Share:

More Explore:

AI Medical Coding V

October 14, 2024

In the final post in this series, I’ll discuss the approaches that can make a very necessary and tedious part of administering healthcare more automated

Read More »

We are interested in hearing from companies creating the next wave of innovations in healthcare

Contact Us