Publicado em Aug. 4, 2020

AI in Medicine: is there a reason for so much attraction?

Health and AI share structured knowledge based on formal learning or observation that allows a clear and organized description of what has been learned.

In the book "Far from the tree", by Andrew Salomon, the author comments that "... one learns to read until the 4th grade, and then one reads to learn". The chronological reference of this sentence is less important than the conclusion about learning: it only happens through elements of language, written or spoken. Of course, those who cannot read can learn things by observation or experience. More empirical learning. You can observe the fire and notice that it is hot and then learn that what is hot burns. Whoever learned to read will be able to [d] write what fire is, that it is hot and that it burns. And anyone who read about fire from someone who described it, acquired a knowledge without having to get burned. This is the process that builds and transmits knowledge.

Artificial Intelligence, which appears around the four winds, is experiencing its third wave of proliferation on the planet. The first wave, where the expression "Artificial Intelligence" was created, dates from 1956 and was contained in the academic environment. The second wave, in the 1980s, had the strong participation of Edward Feigenbaum, who in 1994 received the Turing award. His pupil Edward Shortliffe, is the creator of the most successful AI experiment in this second wave: a program called Mycin [https://pt.wikipedia.org/wiki/Mycin]. What did Mycin do? He simply supported the process of diagnosing infectious diseases at Stanford University School of Medicine. Detail: Mycin had a success rate of 69%, considered higher than the internists of that time. Remember that Mycin was classified as an "expert system" and that it used a knowledge base with around 600 production rules. All this paragraph to say that from an early age, Artificial Intelligence has been around Health. And within it, it creates experiments like Mycin. I wonder why ?

Was this proximity a casual attraction? Or a deliberate intention? In fact, neither. The reason for this proximity between AI and Health lies in the essence of the two areas. AI, in any of its development modalities, needs a knowledge built as described in the first paragraph of this text: structured from a formal learning or observation that allows to create a description of what was learned, in a clear and organized. And the area of ​​Health in general, Medicine in particular, is a productive sector that presents knowledge always very well structured, stored in specialties with clear limits. The most important: the knowledge of the Health area has the quality ensured by strict curatorship. It is for all these reasons that AI approaches the Health area.

In addition to all this essential reasoning, another more mundane reason must also be considered: the growth of knowledge stored in Health - the quality of which is not always acceptable - continues to grow at a dizzying rate. At this time, the support of the machine - quality not always guaranteed - brings a certain encouragement to the health professional, desperate for not having been able to absorb what was published the day before!

A historical retrospective that considers how much AI has already used the Health area to make progress and how much it has already paid back to the sector, will show that this equation has an unbalanced result: the development of AI has already taken more from Health than it has brought for her. But the hope hangs in the air that certain solutions of this new intelligence could even help fight the coronavirus ...

All of this will be reviewed in more detail in the event announced in the illustration for this article. See you there and, who knows, we managed to balance this equation a little more ...

474 reads 211 Likes

About the author

Fabio Gandour
Tecnologia só faz sentido se estiver a serviço das pessoas.

Fabio Gandour

Scientist & Innovation Designer

Gandour foi cientista-chefe funcionário da IBM por 20 anos. Sua responsabilidade inicial na empresa foi dedicada a Informática em Saúde. Neste segmento, atuou no desenvolvimento de soluções e estratégias de marketing. Mais recentemente, foi Gerente de Novas Tecnologias, estabelecendo um efetivo canal de colaboração entre os laboratórios da IBM Research Division e o mercado local.

O cientista é graduado em Medicina pela Universidade de Brasília e PhD em Ciências da Computação.

Read too

World Cup 2022 – Day 12

World Cup 2022 – Day 12

12/04/2022 - Analysis of yesterday's games according to Artificial Intelligence

474 reads 211 Likes
4° Encontro de Networking Associados I2AI
Próximo Evento
Dec. 15, 2022

4° Encontro de Networking Associados I2AI

Reserve sua agenda para participar do 4° Encontro de Networking Associados I2AI em 15/12/2022 das 19h00 às 21h30. Neste dia promoveremos muito networking!!!