Skip to content

Latest commit

 

History

History
155 lines (122 loc) · 6.77 KB

README.md

File metadata and controls

155 lines (122 loc) · 6.77 KB
logo

ML Based Mental Health Diagnosis & Consulting App


Idea in brief

MindCrafter is a mental health diagnosis application that uses data from android smartphones, smartwatches that automatically measure patterns of behavior. Real-time smart devices data measurement provides a complete picture of behavioral functioning. This helps MindCrafter to provide Dynamic AI tailored diagnosis and consulting to each person.

assistant-guide

Plan to bring it to Life

assistant-guide
Working Explanation

I am primarily focusing on the five most important mental issues. 2 out of 5 adults have these issues. I'll be using Smart devices with Android OS to collect behavioral data like sleep times & heart rate of the user.

Anxiety (Excessive Fear)
Physical Symtoms Data Collection Methods
Pelpetations (Sudden Heart Rate Fluctuations) Heart Rate Monitoring
using SmartWatch or Facial Recognition
Sudden Fluctuations in Body Tempreture Body Temperature Monitoring
using SmartWatch or device's built-in temperature sensor
Involuntary Tremors in body Symptom Analysis Questioning
using Application Notifications
Voluntary Tremors Voluntary Leg or Hand Shaking
using Motion sensors
Issues with Stomach (Digestion) Symptom Analysis Questioning
using Application Notifications
Android Based treatments: Meditation, Entertainment, Therapy Session, etc.
Depression (Feeling Upset)
Physical Symtoms Data Collection Methods
Loose of Interest in Social & Entertainment Activities Relaxation Time
Time Spend on Social & Entertainment Media
Excessive Guilt Feelings Therapy Session Analysis
using Therapy Questioning in Application
Sudden Low Physical work & Excersize Meditation, Yoga & Physical Activity Tracker
using Application & SmartWatch
Sleep Abnormalities (Early Wakeups & Disturbed Sleep Patterns) Sleep Tracking
using SmartWatch & Device usage
Poor Concentration Symptom Analysis Questioning
using Application Notifications
Android Based treatments: Therapy Session, ANT clearing Therapy, etc.
Bipolar Disorder (Severe Mental Disorder)

This type of disorder can be analyzed with mania & Depression together. Because of this mental issue. The user always switch between Mania Phase & Depression Phase for 7 t0 15 days for each Phase.

Mania (Depression is discussed above)
Physical Symtoms Data Collection Methods
Sudden Excessive Interest in Social & Entertainment Activities Activity Time
Time Spend on Social & Entertainment Media
Excessive Happiness Feelings Therapy Session Analysis
using Therapy Questioning in Application
Sudden increase in Energy Levels, Physical work & Excersize Meditation, Yoga & Physical Activity Tracker
using Application & SmartWatch
Very fewer Sleep Patterns with high energy levels Sleep Tracking
using SmartWatch & Device usage
Grandeur Feeling Symptom Analysis Questioning
using Application Notifications
Android Based treatments: Long Therapy Sessions, Specialised Consultancy, etc.
187 More Issues with Symptoms and Data Collection Method can be analyzed for ML Model training.
Banner

Product & Features

Each user generates more than 100+ behavioral data points each day. Which can help MindCrafter to analyze the situation of the user. Today users are digitizing more personal data than ever before.

Product Features
Android Application Consulting ChatBot
Medidation Guide (Full)
Daily Yoga Scheduling
Emotional State Monitoring
Google Assistant (Service) Consulting ChatBot
Guided Meditation Session(Quick)
Daily Yoga Scheduling

Project Technology Profile

Frameworks Purpose
Flutter Android Application Development
Firebase Performance & Monitoring
Tensorflow Lite On Device tailored ML Model
Dialogflow Consulting Bot
Google API's Youtube Videos & Music

Google’s help

  • I've got a very limited understanding of feature extraction from behavioral data for ML model training that is different for each user. Google could help me with a better understanding of the Android OS & Data Collection methods that I can use.
  • Google could help with Google API's integration like Youtube API & Music API.
  • This application could impact thousands of people but with Google, it can impact millions of people who don't even know that they need help.
open-header

Plan on using On-Device ML technology

On-device APIs can process your data quickly and work even when there’s no network connection. Cloud-based APIs, on the other hand, leverage the power of Google Cloud Platform's machine learning technology to give you an even higher level of accuracy.

On Device ML Kit API's Use Cases (examples)
Text recognition Language Processing & Conversation Engine
Face detection Facial Emotion & Heart rate Tracking
Object detection & tracking AR Based Meditation Guide
Smart Reply ChatBot Reply Helper
AutoML model inference On-device ML inference
Custom model inference On-device ML inference

Project Timeline

Start Date End date Milestones
2 April 29 April Study about digital diagnosis
1 May 18 May Planning about ML integration
19 May 2 June Analysis Modelling
3 June 23 June Development (Templates & Operations)
24 June 3 July Tensorflow lite Integration
4 July 20 July Machine Learning Training & Debugging
23 July 3 August Functional Testing
4 August 20 August User Interface Designing
21 August 31 August User Experience Testing
Final Deployment 1 September Deployment & Testing

Why Mental Problems ?

National Alliance on Mental Illness Report - Sept. 2019
1 in 5 U.S. adults experience mental illness each year
1 in 25 U.S. adults experience serious mental illness each year
1 in 6 U.S. youth aged 6-17 experience a mental health disorder each year
Suicide is the 2nd leading cause of death among people aged 10-34 

There is only one psychiatrist per 100 000 people

Worried 😟 ?

Smart Devices can help us here.
  • More Than 100 Behavioural Datapoints per user/day
  • Diagnosis Test Results
  • Consulting
  • Tailored Machine Learning Model

Statistics References

https://www.nami.org/learn-more/mental-health-by-the-numbers
https://www.who.int/whr/2001/media_centre/press_release/en/