TB Monitor Framework Luke Geneslaw, Karthic Aragam, and Paul Chang
- Using the framework
This project was meant to suppliment the research and work being
done at the University of Lima, led by PhD candidate German Comina. The TB Monitor Framework is meant to help TB patients monitor their health by tracking the number of times they cough throughout the treatment process and showing them their trends in an app. As their health improves, they will notice a decrease in their cough rates. The framework also includes a module called the sanitizer that is used to help record quality audio. The sanitizer is an open source python project.
The TB Monitor app is currently only on Android, so to install it
simply put the tbmonitor.apk (found in App/ folder) on your android device and install. As for the sanitizer, the only system requirement is to have python installed. Download Sanitizer/sanitizer.py.
3. Using the framework
To begin tracking your health with TB Monitor, open the app and
hit "Inscribirse," then enter your registration info. You now have an account, so when you login you can view graphs of your cough counts over various periods of time. Initially, you won't have any data for the graph to display. Once you upload some cough data, you can use the tabs dscroll through the buttons below the graph to view past data. If you would like to see data displayed, you can use our test account to login.
email : firstname.lastname@example.org password : 12345 Note: recording cough data is out of the scope of this project,
because ultimately this project will be handed off to the team in Peru and paired with the recording technology that they have been developing.
In order to use the Sanitizer on a wav file, the name of the wav file
must be hard-coded into "sanitizer.py". The Sanitizer then outputs a wav file with a name that can also be hard-coded. To run the sanitizer: $ python sanitizer.py
There are several improvements that will be made to this project by
the Peru team in the future, as well as some potential features that would further improve the benefit of the TB Monitor Framework. Below is a list of future improvements in relative order of urgency.
- Adding audio recording to the front end is crucial. Currently, the backend provides a Web API that is capable of receiving audio files via HTTP requests, but there is no native way to record and send files to the server at this point.
- Adding the cough detection algorithm to the server is also necessary for the usability of the framework. The Peru team has a Matlab cough detection algorithm that they will merge with our backend so it can not only receive audio files but also process them. Currently, if the Web API receives and audio file for a patient, it does update their data and new information can be viewed in the app, but the data does not reflect the actual number of coughs found in the patient's audio file.
- Adding security to the interactions to the interaction between the app and the backend is also necessary. Currently, the login and registration process has some security built in, but this can be improved, and security needs to be added to Web API requests and responses.