How things change! It used to be the case that if you were walking along and passed a lone character having a somewhat argumentative conversation with themselves, you’d deftly cross to the other side of the street! Now, of course, it just means that they have a mobile phone with a Bluetooth earpiece and are conversing with a friend by phone. In the old days my wife and I told our children off for using their device at the breakfast table, and now, with the kids long gone, we find ourselves doing just the same thing and we don’t bat an eyelid!
Once upon a time, as a vet, you could be confident in having all the information you needed at your fingertips, just by remembering what you had been taught in vet school. Or at least if you couldn’t remember it, you were in any case streets ahead of the client who was bringing their ill pet to you. At least they thought you knew it all! Now, of course, the client has Googled the condition they reckon their animal has and has found out what the rest of the world thinks you should be doing about it.
Now, of course, the client has Googled the condition they reckon their animal has and has found out what the rest of the world thinks you should be doing about it
How do we cope with that? In a way we need to move with the times just as we have at the breakfast table, making best use of the technology available. These days at breakfast time I’m most likely to be inputting some text to go with a picture of an interesting case I saw the day before to entertain and hopefully educate through my Instagram. And that means in the consulting room if I have an interesting case I’ll Google it myself in front of the client, showing them how to use the internet to the best effect – using Google Scholar to access peer-reviewed papers rather than just going on Facebook or Twitter.
So, yesterday’s case was an elderly Greyhound with an irritated eye and a white corneal opacity that looked just like the pictures of stromal calcification in a paper by Jane Sansom on the condition in eyes of dogs with Cushing’s disease – calcinosis corneae not calcinosis cutis! It took a few seconds to find that article online to show to the owner and Jeremy, the referring vet who was in the consult room with me. It was Jeremy who pointed out that nothing else about the slim, fully haired dog suggested a diagnosis of Cushing’s – so back to square one! Dr Google is by no means always right!
But today’s case was a blind Shih Tzu and Google Scholar immediately revealed the perfect paper detailing the gene mutation causing progressive retinal atrophy in this breed: a mutation in the JPH2 junctophillin gene. Don’t worry if you haven’t heard of that – I hadn’t until this morning. Every day is a school day, hey?! Back to our Greyhound, I wondered if artificial intelligence could work out whether a dog was Cushingoid through its symptoms without a need for complex blood tests? What I love about a quick Google search is the other stuff you come up with – rather like seeking a particular volume in a second-hand bookshop. It’s the other tomes you come across that can quite divert you from your first aim.
I wondered if artificial intelligence could work out whether a dog was Cushingoid through its symptoms without a need for complex blood tests?
And so, on with a quick Google Scholar scan on Cushing’s and artificial intelligence. Straight away up comes a fascinating paper on using a neural network to recognise patterns in tetrahydrocortisol and pregnanetriol blood levels in patients with Cushing’s – sadly human ones, not canine, but you can’t have everything! With Cushing’s in humans, there are two tests to try to differentiate between adenoma, hyperplasia and carcinoma. The paper starts off nice and simple, explaining pattern recognition and neural networks. I can cope with a publication when it’s all prose. But the moment equations start rearing their ugly heads I’ve got a problem! So a Bayes formula telling me that the neural network needs to minimise SCjkp(jIx) sends me completely cold! Apparently, the posterior probabilities p(kIx) are central to the theory, but that just leaves me happy that veterinary medicine doesn’t need Bayesian statistics very often!
Gorgeous Google thankfully led me away from the maths and happily to a biographical sketch of the Reverend Thomas Bayes. Bayes was a non-conformist minister whose family came from Sheffield and made their livelihood in cutlery. The Act of Conformity in 1662, which said that all church ministers had to use the Book of Common Prayer, led to Thomas’s grandfather being ejected from his parish church along with over 2,000 other clergy. This also meant that Thomas couldn’t go to Oxford or Cambridge, so he studied in Edinburgh under the mathematician James Gregory, even though his family thought he was studying theology!
Thomas must have done that as well because his first publication was Divine Benevolence – how what God wanted was the happiness of all his creation. He managed to juggle these two subjects, for as a minister in Tunbridge Wells, he did enough “geometry and mathematical learning” to be elected as a Fellow of the Royal Society. This was the time of Isaac Newton and his differential calculus which Bayes strongly supported, but his main interest seems to have been in probability. Bayes’ theorem gives the probability of one event occurring given prior knowledge of another. There’s not enough room (or brain power on my part!) to discuss this further, but for me the interesting part is seeing how Bayes’s understanding of random probability of events links in with his concept of a God wanting good things for his creation. What does Bayes do with regard to miracles, I wonder?
Well, I’ve spent an hour reading through various papers on Bayes and miracles without much light being shone on the question I just asked, but for the indisputable fact that for all the statistics and probabilities based on chance, which Bayes worked on throughout his life, he remained a Christian believing in God’s love. Maybe the key problem is that you can’t put that into a mathematical equation however hard you try. And that will never change!
