A digital stethoscope, which uses a smartphone and the Cloud, could save thousands of children’s lives in developing countries.
The stethoscope has been created by four university students from Victoria for the Imagine Cup, which challenges students around the world to develop unique products using technology.
The creators of the digital stethoscope, Team StethoCloud, include Hon Weng Chong, Andrew Lin, Kim Ramchen and Masha Salehi, and they are hoping the product can be readily used in developing countries.
“The idea behind this is that due to the rapid proliferation of mobile phone technology in developing countries, such as India and Africa, and the mobile phone being a readily accessible, we can then harness the availability of the technology [to help reduce childhood mortality],” Weng Chong tells Computerworld Australia.
“They already have this technology and hopefully every village chief would have a mobile phone or every town centre where someone could bring the child in and have the application diagnose them.”
The product also has potential applications in countries like Australia where rural residents could utilise the product.
How it works
The team has taken a traditional stethoscope and embedded a microphone in the head. A wire is threaded through tubing and a 4.5mm stereo jack has been attached to the end, which is plugged into a smartphone’s microphone inline connector.
“What you do is you take the stethoscope, you place it on the back and there are six locations [which you record], because in medicine that’s how a doctor would listen to you. We take a five-second recording in each of these spots,” Weng Chong says.
“We take all of these recordings, we store them temporarily in a cache on the phone and once it’s all done, we can upload it to the Cloud. This is actually running our infrastructure on Azure right now.”
Users answer a series of questions from the World Health Organisation guidelines for pneumonia which are asked before taking the recording. The application includes videos which show symptoms that healthcare workers may not be familiar with and gives an explanation of the symptom and what it looks like.
The audio recordings are then run against algorithms to detect respiratory rate and three diagnoses are returned – no pneumonia, pneumonia and severe pneumonia. A treatment plan is then provided.
While the prototype application currently only operates on Wi-Fi or 3G connectivity, the team is developing the application to operate offline.
“Because this plugs into the phone and it acts like an inline microphone, the idea is that we can use this to call up our server that’s running a Skype gateway. It then transmits the audio signals that we have coming in as a phone call,” Weng Chong says.
“That then transmits it as an audio file across to our server and all the textural data will be sent as a SMS. We then do all the computation as we would, but instead of pushing the result back by 3G, we push it back as a SMS.”
It also only currently runs on Windows phones, but Weng Chong says it may also be developed for Android phones. However, there are no plans to integrate the product with iPhones. “You won’t find any iPhones in developing countries," Weng Chong says.
The digital stethoscope costs about $30 to make — about $25 for the stethoscope head, $3 for the stereo jack and $2 for the microphone —and can be made as cheaply as $15 to $20. This compares to other stethoscopes on the market that are about $400 to $800 for digital stethoscopes and $200 for traditional stethoscopes, Weng Chong points out.
One key to making the product accessible financially is the digital sound processor which is used to digitise the recordings, as opposed to digital stethoscopes which embed a chip inside the head of the stethoscope and costs several hundred dollars, according to Weng Chong.
Weng Chong says he would like to improve the stethoscope by embedding better microphones which are smaller, more sensitive, cancel noise and more uni-directional.
“We’re also writing up a research protocol to start doing some live trials on patients at the children’s hospital in Melbourne, so we believe the key is to have lots of data, and we want to do it in a controlled setting where we’re just collecting samples,” Weng Chong says.
Ramchen, who has helped with developing the algorithm for the stethoscope, says it currently has about 80 per cent accuracy in a quiet room. About 17 per cent of the time the results will be inconclusive and 3 per cent of the time it will slightly over estimate results.
He says carrying out multiple samples can help achieve more accurate results and the team is working on improving the hardware and algorithms.
Eventually they would like to include Big Data in the application and are looking for “techy” doctors, nurses and healthcare practitioners to help collect data — the more data in the algorithm, the better the accuracy of results.
“If we can harness the Cloud storage platform to get lots of data, you can run a Big Data approach on it and using even really generic data can be very powerful,” Ramchen says.
Weng Chong estimates it will be about 18 to 24 months until the product progresses beyond the first prototype to a commercially available product.
Follow Stephanie McDonald on Twitter: @stephmcdonald0
Follow Computerworld Australia on Twitter: @ComputerworldAU