AI and the MD: report

According to the author of  “AI and the MD,” published in the latest edition of the U of T Magazine, Patchen Barss, the role of artificial intelligence (AI) in health care will continue to grow.

So-called “smart algorithms” will soon be found not only on our personal devices, but in doctor’s offices. Most notably, Barss states, AI will change how we treat disease.

Barss interviewed Anna Goldenberg, who is a senior scientist at the Hospital for Sick Children and professor of computer science at the University of Toronto;  Marzyeh Ghassemi, who was recently appointed to the Vector Institute for Artificial Intelligence in Toronto; and Quais Morris, who teaches molecular genetics at the University of Toronto.

Goldenberg is working on an algorithm that could predict whether a baby with Li-Fraumeni Syndrome will develop cancer before turning six years old.

La-Fraumeni Syndrome is a rare genetic mutation that stops the effective functioning of TP53, a tumor-suppressing gene.

Some children with Li-Fraumeni Syndrome will get cancer in the first few years of their lives; others can spend decades of life tumour-free.

Ghassemi hopes to harness the power of wearable technology, such as smartphones and smart watches, to help people manage depression.

Ghassemi is working on apps that will turn these wearable devices into background monitors.

The app would track biological information like sleep cycle and heart rate, as well as the user’s social behaviors, such as frequency of going out and socializing. The app could also track changes to voice intonation. Changes in any of these things could indicate a descent into depression.

Morris hopes to use AI to improve our understanding of cancerous tumors.

Morris is using machine-learning algorithms as a means to combine clinical information with molecular data. By doing so, Morris hopes to learn more about how tumors evolve and answer questions such as:

Why does a particular chemotherapy work for one patient, but not for another? Who’s at risk for which types of cancers, and why? Why do some people get cancers and others do not?

In Conclusion

While AI can reasonably be expected to transform health care, Goldenberg, Ghassemi, and Morris agree that, for the foreseeable future, the role of AI will be to “augment the capacity of human medical professionals, not to replace them,” Barss writes.

As Morris states, we need humans “for the subjective things.” A machine is not (yet) capable of making decisions intuitive decisions or complex judgement calls.

For more on AI and health care research conducted at the University of Toronto, read the full article.

Leave a Reply

Your email address will not be published. Required fields are marked *