Why artificial intelligence has not revolutionised healthcare… yet

[AussieNeil]

'Artificial intelligence and machine learning are predicted to be part of the next industrial revolution and could help business and industry save billions of dollars by the next decade.

The tech giants Google, Facebook, Apple, IBM and others are applying artificial intelligence to all sorts of data.

Machine learning methods are being used in areas such as translating language almost in real time, and even to identify images of cats on the internet.

So why haven’t we seen artificial intelligence used to the same extent in healthcare?'

Olivier Salvado, Group Leader Biomedical Informatics, CSIRO explains the difficulties: theconversation.com/why-art...

Here's a fun and possibly apocryphal account, illustrating the difficulty with one form of artificial intelligence - neural networks and which explains how the US military became the proud owner of a multi-million dollar mainframe computer that could tell you if it was sunny or not... neil.fraser.name/writing/tank/

Neil

Photo: Cape du Couedic lighthouse on the SW tip of Kangaroo Island

[Foggymind]

Hi Neil,

Thanks for those links, I found them very interesting. As someone who has built many robots at home and dabbled with artificial intelligence I personally believe the use of AI in medicine is a very long way off. Even if the technology was available today I think the level of trust for these systems by both the health authorities and the public would be low. Many years ago I was involved as an IT business analyst in the devolvement of a system for the NHS to transcribe most of their paper records to a computer based storage solution. After months of analysis of the work involved in this venture it became obvious the man hours necessary to input all the manual record data including the cross validation of that data would be prohibitively expensive and prone to mistakes. The project was eventually dropped by the NHS and to this day they still do not have a central electronic repository of medical record data.

What was learnt at the time was that there was very little trust in computerised systems and many preferred hanging on to what they knew worked and what they trusted. I think this frame of mind is still prominent today.

Referring back to the neural networks, these too I believe have a good few years development ahead before they truly become usable and trustworthy. Currently, neural networks consisting of millions of nodes and connections and running on superfast computers just about match the computing power of a worm.

This is of course my own personal view based on my own experiences and knowledge. I’m sure that computers will one day touch every aspect of our lives and hopefully make a good job of it, I just don’t think we have got there just yet.

Kevin – Essex, UK

[Cllcanada Top Poster CURE Hero in reply to Foggymind]

Computers touch every aspect of my life... from my coffee maker to my thermometer to my camera to my refrigerator, Uber car routing and beyond.

SIRI what time is my next med?

Chris...your next med is Rosuvastatin at 4pm... would you like me to set a reminder sound?

Here are a few more ...

statnews.com/sponsored/?ske...

[Foggymind in reply to Cllcanada]

You raise a very good point there Chris, I should have thought that last sentence through a bit more. What I was trying to say (I think) is that AI and neural networks don’t really impact us in a big way yet. Most of the computer power in our homes, cars, and workplace etc. are microprocessors running straightforward programs written to hopefully cope with any eventuality it comes across. Most of us know of the MS Windows blue screen of death, that’s where the program running didn’t cope, didn’t have an escape clause for something it wasn’t expecting and so it just crashed as it were. AI and neural networks can learn things without it being hard programed by a human. They can modify their behavior based on the things they learn and can handle the same events differently next time.

Kevin – Essex, UK

[Cllcanada Top Poster CURE Hero in reply to Foggymind]

Blue screen of death... how appealing... MS Windows... whatz dat? 🖥

I only know about the Macintosh toolbox and Turi... 😆

turi.com/products/

Face recognition on my iPhone camera uses neutral nets...one day they will recognize tumours...

backchannel.com/an-exclusiv...

[AussieNeil Administrator in reply to Foggymind]

Interesting to hear of your professional experiences in this field, Kevin. I find it sobering that it was only a few short years ago that our global Internet reached the connectivity level of just one human brain...