In 3rd world developing countries, the majority of the people do not have access to good medical consultation. The idea was to have an easily accessible, low cost kiosk located in these small cities. A patient will be able to use facial recognition to login and contact specialized doctors world wide in order to discuss medical concerns for free.
What it does
Firstly a user would login into the kiosk using facial recognition. Then he or she would ask for help with their area of concern from various on-call doctors worldwide (from Cardio to ENT). The patient would wait in a queue depending on how many on-call doctors are online, until a doctor starts reviewing the patients medical folder. The doctor can request the patients vitals (heart rate, temperature) via the kiosk which has these sensors connected. The patient will then take these tests as instructed. Once the doctor has the vitals he can continue diagnosing the patient by asking questions. We have overcome the language barriers by using Google translate to allow for smooth integration between various countries. This way doctors world wide can help diagnose patients during their on call "dead hours".
How we built it
For the hardware, Arduino is used to take measurements from a heart rate sensor as well as a thermometer, both of these readings need some time to calibrate. For facial recognition hard cascade was used to detect the face, then the image is cropped around the face. Dlib was used to locate facial points. 100 photos are taken and sent to a training API for a simple secure facial recognition. For the patients and doctor UI pyhton, vue.js, and tkinter for a simple login and user data input to a firebase database. Whereas Google's translation API is used to translate from the doctors language to the patients language and back for simple diagnostic communication.
Challenges we ran into
A complete redesign of the system set us back in time as we had a problem with setting up server. Integrating node with python was a difficult challenge to overcome as well.
Accomplishments that we're proud of
Learning a lot of different skills from front end to back end. The idea of this product can help reduce the mortality rate in 3rd world countries, as it allows for early professional consultation by professional doctors worldwide.
What we learned
Learned how to run front end development using python, used Google Cloud services for web development APIs.
What's next for KioDoc
KioDoc can include a medicine translator that finds the active ingredient in the medicine prescribed by the doctor and finds the equivalent medicine in the patient's country.