Doctor Bot

Total healthcare solution

Comment

Inspiration

Recently, my mother was being admitted to the hospital. Before she was admitted, no doctor was ready to see her from close due to the Covid-19 virus outbreak. At this point, we were not able to diagnosis the disease. Further, after her discharge, there were a lot of hardback prescription to keep account. We had to store images of every prescription which was very inconvenient. Further, when we wanted to visit the doctor for follow ups, we had to wait for 2 hours sometimes to meet the doctor. After going through all the hassle, during brainstorming for a project idea to develop for HTN I thought this project would be a perfect one to help all the people out there.

What it does

  1. NLP Chatbot - DoctorBot engages with users like a doctor and provides a detailed analysis based on inputs provided by the patient on follow up questions. With the help of NLP the patient can enter it's symptoms in normal language and DoctorBot is able to identify the feelings and symptoms of users. DoctorBot asks various types of questions including multiple choice single choice as well as group single type which accounts for more precise results. For example on the left slide the user entered the text I have mild headache and cough". A human is prone to making some spelling mistakes as our patient may have done in headache spelling but our bot is smart enough and usually captures spelling and grammatical errors. Next up the bot asks various questions to the patient and predicts the disease with the help of answers answered. The bot is smart enough to even dodge the symptoms which the users mentions as no. which in above case was cough. The predictions contains the symptoms observed over the course of interview and the diseases predicted along with it's probability which can be downloaded for future purpose and are also stored in the database.
  2. Book Appointment - The patient can book an appointment with a doctor near him through the application. The patient can also book appointment without using the DoctorBot according to his choice. The application takes the start and end time of a doctor as input and provides slots accordingly.
  3. Prescribe - Once the appointment has started the doctor can prescribe medicines and symptoms observed and form a E-Prescription which are stored in database and can be used by the patient for further visits to doctors.
  4. View Previous Prescriptions and Predictions - Once a patient has booked an appointment to a doctor, the doctor has the patients information along with his previous prescriptions and predictions which can help the doctor for better diagnosis and prescribe more accurate healthcare. The patient is also able to view his own previous prescriptions and predictions.

How I built it

The following steps were followed during the development of project:

  1. Found and wrote down the scope of the project
  2. Create a work breakdown structure (Chatbot, appointment booking, etc)
  3. Break each work package into tasks.
  4. Determine project dependencies through creating a flowchart
  5. Determine total time needed for each task defined previously.
  6. Identify important milestones of the project to focus on them more
  7. Create a project management timeline for the duration of 2 days

Challenges I ran into

Various challenges were faced during the course of development of project. Some of them are:

  1. The time available to build the project was very less, so to prioritise things of the project was a challenge initially.
  2. Use of Infermedica API as there was no proper documentation available.
  3. Making sure once a patient has booked an appointment on specific slot, that slot should not be provided to anyone else.
  4. Making sure that the patient is not able to book an appointment on the day the doctor has a holiday.

Accomplishments that I'm proud of

  1. The fact that I was able to give my best and build something useful given the time duration.
  2. Was able to build something that can bring a difference to people lives.

What I learned

  1. Make use of the Infermedica API and build a successful chatbot.
  2. Learn how to use firebase database along with Angular
  3. Use of socket programming

    What's next for Doctor Bot

    1. Video streaming between patients and doctor in case of unavailability of any one.
    2. Reminders(notifications) given to patients based on the prescription prescribed by the doctor.
    3. Patients can buy medicines through application prescribed by doctor.
    4. Patients can upload various reports which can be used for analytics of his/her well being like their fitbit data, blood test reports.

Share this project:

Updates