First up this month was Eddie who is researching pre-NHS healthcare from 1900 to 1948 and especially how it was funded. There were a number of different institutions, foremost among these were the Voluntary Hospitals. These were centres for acute care and were essentially private, although not in the same way that hospitals are private these days as they were not run for profit. They were funded by a combination of philanthropy, charity and mutuality schemes. These were like insurance where people would make payments out of their wages in order to get a certain level of cover.
Then there were the workhouse infirmaries. These would often be full of the long term sick and the elderly and people would often be sent here from the voluntary hospitals in order to convalesce. These were actually an accidental growth following the implementation of the poor law.
On a smaller scale were the cottage hospitals. Based in villages and towns these would usually have fewer than 25 beds and relied on local doctors and GPs, who were sometimes underqualified or whose methods were outdated. Ripley cottage hospital was founded after a miner was crushed and had to be taken ten miles on a horse and cart to the nearest hospital.
Below that were district nursing associations. These were for people who couldn't make it to a surgery. They would be paid for by a combination of subscription schemes and charity events such as fetes and tea parties.
None of the network really had any connections with the other parts, they were all very independent. The voluntary hospitals were run by local aristocrats or businessmen and they wanted to keep control of their own domains. In 1911, the National Insurance Act was passed in parliament. This provided a level of health insurance and saw the start of standardised GP cover. In 1948, the government essentially compulsorily purchased all of the voluntary and cottage hospitals. The infirmaries transitioned into regular hospitals and the NHS was born.
Key learning: Unlike in modern private hospitals, you wouldn't be turned away from a voluntary hospital even if you weren't insured.
The second speaker was Sunni, who is stimulating neurons using light as part of a team that is trying to create a simple, living, artificial brain. When you have cells on a petri dish and add a chemical, it is distributed throughout the dish and it stimulates all of the cells. Sunni is trying to stimulate a single cell.
Instead we use a drug whereby we can change its confirmation by shining a light at it. Then, when we remove the light, the drug is no longer active. The closed form of the drug is colourless, UV light hits it and it becomes colourful. If we add whit light, it loses its colour again. The UV light causes a bond to break open in the drug.
Cells have hidden receptors. We need to get the drug very close to the cell itself before we activate the drug. Then the idea is that the cell that we stimulate is a neuron. Once we can stimulate a single neuron, we can start to understand how learning works. This is because as we learn, a pathway is created in the brain. We are trying to replicate the creation of this pathway.
Sunni is using dopamine as the drug that he is trying to activate. The long term goal is to develop this technique so that the drug could go into the brain and then you would shine a UV light through the skull to release the dopamine.
Key learning: Sunni's team have managed to stimulate a motor-neurone to move a ball on a screen.
Last up was Yves, talking about using mathematics in image processing. Specifically, he's looking at how do you detect objects that you care about in an image and discard the rest? Is each individual pixel part of what we're interested in or not?
As we know, oil and water don't mix; this is called phase separation and the models used to describe it are interesting from a mathematical point of view. This is very similar to the segmentation that Yves is doing, separating the pixels that we're interested in from those that we aren't.
Since, in the words of Hungarian mathematician George Polya, "mathematics isn't a spectator sport", it's time for some audience participation. First of all, we're given pieces of paper that represent oil or water. Following a very simple algorithm to swap our pieces of paper, the water and the oil separates out into a couple of clumps of water and a couple of clumps of oil. Yves then hands out 25 more pieces of paper that make up a 5x5 picture. By comparing each image to a dictionary picture and then using the same algorithm used before, we put the picture back together.
Instead of using humans, Yves is using computers which are capable of doing similar calculations millions of times over and far faster than a group of people in a pub. The computer builds a network of pixels based on how similar they are to the dictionary picture that is initially specified and it weights the pixels based on how similar they are. To tell whether the pixels are similar, the computer looks at the local neighbourhoods of each pixel and compares them. This whole technique works better with better contrast.
Currently Yves is using this technique to help out some biology researchers who are looking at the size of certain markings on the heads of birds. The computer can tell which feathers make up the marking vs those that do not and can then measure the size of the marking on each bird.
Key learning: "Counting wolves is the same as counting apples".