Now this blog post really is at the cutting edge of research, because we have barely scratched the surface of this topic at UEA, and we are still waiting for funding to carry out the main body of the work (the team expects to hear in May whether we have got through the first round of the European Research Council funding competition).
There is currently great deal of scientific interest surrounding the relationship between the speech patterns of individuals, and whether this can be used as a diagnostic tool for developmental problems in children, as well as progressive medical conditions such Alzheimer’s Disease in later life. The jury is out on how useful this might be in practice. There is also increased pressure for earlier, more accurate diagnosis of progressive neurological conditions, as a result of drug companies coming up with new products, as well as a recognition that there are increasing levels of this kind of disease within many Western countries, which has a lot of consequences for welfare funding and so on.
Most of the research into conditions such as Alzheimer’s focuses on physical changes to the brain, and genetics, as well as the role of environmental factors such as diet and stress in predisposing individuals to develop the disease. Some attention is given to aspects of memory loss and declining neurological function. However it seems that the least well researched aspect is the decline in linguistic ability, which is a normal function of ageing, but which seems to accelerate in the case of certain diseases. This particular decline is well documented, and we have found research papers referring to it that go back a couple of decades, which got us quite excited about the possibilities of what might be found using new ways of tracking speech using powerful computers, and researching the problem systematically. This theory has even been tested by another researcher on the novels of Iris Murdoch, a famous Alzheimer’s sufferer, and the writer after whom our software, “IRIS”, will be named.
At the moment, we know so little about the role of speech in Alzheimer’s, that it makes early detection considerably more difficult, particular in situations where there is pressure on local primary healthcare services. If a better understanding of the precise nature of linguistic decline in such circumstances could be achieved, it would give a simpler, more solid and more readily understood indication of speech and language impairment in the pre-clinical phase, before people appear on the psychiatrist’s radar, so to speak. To this end, we are planning to develop computer software that should be able to test and track linguistic decline in relation to these conditions, and which would complement existing diagnostic tests such as the well established Mini Mental State Examination, a 30-year-old questionnaire used to test thinking ability. There is also the potential for such software to be used in family doctor surgeries as a cheap and simple screening mechanism for referrals to specialists. During the study, we are also planning to collect speech data from a sample of already diagnosed patients as well as patients who form part of a control group, and to analyse that data using new technological tools. This will allow us to start developing a more sophisticated and reliable system of objective speech measurement, and it will also enable us to build a small network of technologists, social scientists, speech therapists and psychiatrists who will be able to work together in developing and testing it further, with the ultimate aim of improving the accuracy of early stage diagnosis amongst all social groups, and building a new UK linguistic analysis hub.