Post by Willow on Mar 5, 2013 20:45:09 GMT 9.5
CSIRO developing sensors to 'taste' disease by phone
I didn't know whether to put this in health or IT!! But its a nifty idea!
DIAGNOSING diseases such as malaria and tuberculosis in the developing world could be radically improved by a sensor attached to a mobile phone being developed by Australian researchers.
Leader of CSIRO's Medical Devices Stream, Scott Martin, said 20 researchers were building a phone-attached miniature sensor capable of detecting an array of bacterial diseases using a breath sample or urine sample. It would be used in countries without adequate pathology services.
Dr Martin said the sensor contained inkjet-printed material with an electrical resistance that changed depending on the sample.
Given the lack of even 3G data services in developing countries, the phone will not be used primarily for transmitting data but for displaying test results and advising health professionals of treatments and suitable pharmaceuticals based on the findings.
"Because we're targeting developing nations, we're not focusing specifically initially on smartphones. The prevalence of mobile telephones in developing nations is very high, but they can't afford the sort of smartphones that we currently have," Dr Martin said.
The phone software therefore would be a feature phone program rather than a modern-day smartphone app.
Dr Martin said CSIRO was backing up the sensor system with an artificial intelligence program that built a profile of the range of diseases that could be sensed based on trials under way at Royal Melbourne Hospital and St Vincent's Hospital in Melbourne.
"It's a different sort of sensor to the ones commonly used in health. Normally when you go to a doctor and have a test done, they take a sample from you and send it out for very specific testing," he said.
"Ours doesn't have a single sensor looking for a single property of the sample. It has an array of sensors. It's not looking for a specific target. It's looking for a fingerprint that's indicative of the disorder."
Dr Martin said the concept of the sensor was a bit like wine tasting.
"Wine itself is a very complex liquid with many components to it, but we can taste wine and even small differences between regions and so our sensor is doing the same sort of thing. We're loading it up with a range of different tastebuds, if you like, that use a similar approach to animals and humans."
He said the project used lots of supercomputer time and machine-learning to find out "the taste profile" of what disease a person had.
The mobile sensor project is being partnered with the Nossal Institute for Global Health, which is developing a smart cable that adds some processing power outside of the phone.
The cable also has components to enhance co-ordination and control and a range of different sensor modules that could be plugged in.
"Any specific levels of computations would happen on the smart cable, which would be a very cheap device," Dr Martin said. "There would be some software that sits on the telephone that can interpret the results from the smart cable."
A health professional could read the test results and then the software would offer guidance as to appropriate medicines and treatment.
Dr Martin said the project began in 2011 and would run for five years. It was being funded by the Science and Industry Endowment Fund.
"At the moment we're taking a lot of patient samples from two hospitals in Melbourne from people who come into the emergency department. We're interested in distinguishing between two types of people - people with clear infections and people who do not have infections."
Of the 20 staff involved, eight were full-time and 12 part-time staff with multidisciplinary backgrounds and representing more than 16 nationalities.
The project's aim was to produce a prototype. Decisions were yet to be made on specific countries for rollout, but the project was looking at partnerships with India, where tuberculosis was a major problem, and the team had on-the-ground experience in Mozambique and Tanzania, Dr Martin said.
I didn't know whether to put this in health or IT!! But its a nifty idea!
DIAGNOSING diseases such as malaria and tuberculosis in the developing world could be radically improved by a sensor attached to a mobile phone being developed by Australian researchers.
Leader of CSIRO's Medical Devices Stream, Scott Martin, said 20 researchers were building a phone-attached miniature sensor capable of detecting an array of bacterial diseases using a breath sample or urine sample. It would be used in countries without adequate pathology services.
Dr Martin said the sensor contained inkjet-printed material with an electrical resistance that changed depending on the sample.
Given the lack of even 3G data services in developing countries, the phone will not be used primarily for transmitting data but for displaying test results and advising health professionals of treatments and suitable pharmaceuticals based on the findings.
"Because we're targeting developing nations, we're not focusing specifically initially on smartphones. The prevalence of mobile telephones in developing nations is very high, but they can't afford the sort of smartphones that we currently have," Dr Martin said.
The phone software therefore would be a feature phone program rather than a modern-day smartphone app.
Dr Martin said CSIRO was backing up the sensor system with an artificial intelligence program that built a profile of the range of diseases that could be sensed based on trials under way at Royal Melbourne Hospital and St Vincent's Hospital in Melbourne.
"It's a different sort of sensor to the ones commonly used in health. Normally when you go to a doctor and have a test done, they take a sample from you and send it out for very specific testing," he said.
"Ours doesn't have a single sensor looking for a single property of the sample. It has an array of sensors. It's not looking for a specific target. It's looking for a fingerprint that's indicative of the disorder."
Dr Martin said the concept of the sensor was a bit like wine tasting.
"Wine itself is a very complex liquid with many components to it, but we can taste wine and even small differences between regions and so our sensor is doing the same sort of thing. We're loading it up with a range of different tastebuds, if you like, that use a similar approach to animals and humans."
He said the project used lots of supercomputer time and machine-learning to find out "the taste profile" of what disease a person had.
The mobile sensor project is being partnered with the Nossal Institute for Global Health, which is developing a smart cable that adds some processing power outside of the phone.
The cable also has components to enhance co-ordination and control and a range of different sensor modules that could be plugged in.
"Any specific levels of computations would happen on the smart cable, which would be a very cheap device," Dr Martin said. "There would be some software that sits on the telephone that can interpret the results from the smart cable."
A health professional could read the test results and then the software would offer guidance as to appropriate medicines and treatment.
Dr Martin said the project began in 2011 and would run for five years. It was being funded by the Science and Industry Endowment Fund.
"At the moment we're taking a lot of patient samples from two hospitals in Melbourne from people who come into the emergency department. We're interested in distinguishing between two types of people - people with clear infections and people who do not have infections."
Of the 20 staff involved, eight were full-time and 12 part-time staff with multidisciplinary backgrounds and representing more than 16 nationalities.
The project's aim was to produce a prototype. Decisions were yet to be made on specific countries for rollout, but the project was looking at partnerships with India, where tuberculosis was a major problem, and the team had on-the-ground experience in Mozambique and Tanzania, Dr Martin said.