What will the future look like when machines are used in medicine?

Using Artificial Intelligence to help doctors diagnose cancers: Predicting the response of immunotherapy to checkpoint inhibitors and other treatments

The medical world believes that Artificial Intelligence has a huge potential in helping doctors make diagnoses. It’s already being used as a radiological tool, helping with CT scans and X-rays. Another program called OpenEvidence, led by scientists at Harvard University, the Massachusetts Institute of Technology and Cornell University, is using AI to read through the latest medical research studies and synthesize the information for users.

One example of where AI tools could surpass human abilities, according to Poon, is the use of digital pathology to predict tumoral responses to immunotherapy. It is thought that the tumour microenvironment — the milieu of cancerous, non-cancerous and immune cells that can be sampled using a biopsy — influences whether an individual will respond well to various anti-cancer drugs. If you can see millions of patients with the same type of cancer who have already taken a checkpoint inhibitors or other immunotherapy and you look at the exceptional responders and the non-responseers, you could start to discern a lot of these patterns that an expert might not be able to see.

“If you scale that to the level of a health system, you can start to see how there’s a plethora of choices to make about the devices themselves and a plethora of decisions on how to integrate them, purchase them, monitor them, deploy them,” he says. “It can quickly become a kind of IT soup.”

Foundation Models: A Medical AI Platform for Detecting Patterns in High-Dimensional Images and Genomic Data, with a Specialization on Generative Artificial Intelligence

“It’s becoming now better known in the field that you have to do an external validation,” Liu says. But, she adds, “there’s only a handful of institutions in the world that are very aware of this”. If there’s no testing of the model in a particular setting, there’s no chance of knowing if these tools are actually helping.

The following month, Google researchers described in a preprint6 That approach together with the Med-PaLM model can answer some open-ended medical queries almost as well as a physician. The Med-PaLM Multimodal system was able to not only read chest X-ray images but also draft a medical report in natural language.

These are examples of what some scientists call a foundation model. The term self-supervised learning describes models that are trained on broad data sets and use a method called self-supervised learning. They are called base models or pre-trained models as they form a basis for performing different tasks.

Foundation models are “a very flexible framework”, says Karthikesalingam, adding that their characteristics seem to be well suited to addressing some of the limitations of first-generation medical AI tools.

Big tech companies are investing in medical-imaging foundation models that use multiple image types, and incorporate electronic health records and genomics data.

As these models are trained on ever more data, some scientists are optimistic that they might be able to identify patterns that humans cannot. A study was done by researchers at Google to show how an artificial intelligence model can identify someone’s age and gender from a picture of their eyes. That is not something that experienced ophthalmologists can do. “So, there’s a real hope that there’s a lot of scientific information embedded within these high-dimensional images.”

With artificial intelligence seemingly working its way into every technology out there, one area where it’s considered particularly promising is in helping doctors make medical diagnoses.

Mansour has been helping test an experimental version of UpToDate that uses generative AI to help doctors access more targeted information from its database.

“AI won’t replace doctors, but doctors who use AI will replace doctors who do not,” Succi says. It’s the same as writing an article on a typewriter or on a computer. It’s that level of leap.”

“It needs improvement,” says Dr. Marc Succi of Mass General Brigham, who was one of the paper’s authors. “We’ve drilled down on specific parts of the clinical visit where it needs to improve before it’s ready for prime time.”

“It’s a time-consuming and very haphazard process,” says Dr. June-Ho Kim, who directs a program on primary care innovation at Ariadne Labs, which is a partnership of Brigham and Women’s Hospital and the Harvard T.H. Chan School of Public Health. Natural language summaries of it are very useful and a large language model that can digest that is able to do that.

He once saw it cite a journal article in his area of expertise that he wasn’t familiar with. “And I then looked to see if I could find the study in that journal. It didn’t exist,” Bonis says. My next question was “Did you make this up?” It said yes.”

Gen AI can help doctors make better diagnoses better — Peter Bonis, MD, EMEA, EPR, npr, QED, etc

At this point, Wolters Kluwer Health is just sharing the AI-enhanced program in a beta form for testing. Bonis says the company needs to make sure it’s entirely reliable before it can be released.

Peter Bonis says that if you have a question, it’s possible to keep the context of it. “And saying, ‘Oh, I meant this,’ or ‘What about that?’ And it knows what you’re talking about and can guide you through, in much the same way that you might ask a master clinician to do that.”

He scrolls down a list of responses to things he has received, such as a mosquito-borne disease. Mansour says he wishes this list could be more specific: “I think gen AI gives you the opportunity to really refine that.”

“Here’s an example,” he says as he looks at his computer. “If I meet a patient who is visiting from Hawaii.” When a patient’s symptoms make him worry about an infection that he got back home, he types “Hawaii” and “infection” into UpToDate.

It is a huge database of articles written by experts in the field, who are all pulling from the latest research, which is what it is called.

Source: AI could help doctors make better diagnoses

UpToDate, a computer program to help patients with infectious diseases diagnose and manage invasive infections in transplant patients: A laugh with Mansour

The doctor uses a computer program called UpToDate when a patient comes in with a mystery disease. It’s an incredibly common tool, with more than 2 million users at 44,000 health care organizations in over 190 countries.

Invasive infections in transplant patients are cared for by Mansour. “Got a nice picture of mushrooms in my office,” Mansour says with a laugh. I like helping patients with destructive mold and yeast infections.