Deap Learning AI for Rare Diseases

I have a kidney disease. An extraordinarily rare one at that. For my entire life, I've spent most of my life in a hospital bed or at a doctor's office. No one could help me. So I started working on a Deep Learning AI to remove bias and prejudice from the equation. 

Doctors tend to have biases they are human. The issue is that those biases can lead to misdiagnosis and mistreatment sometimes harming the patient. This is much more common than people realize. I've collected data from hundreds of volunteers with rare diseases that have all told the same story. "No one will help me, I feel so alone". The solution isn't removing the doctor but allowing them to better view what could be the issue with a patient without guessing. 

The AI will consist of a database with patient data. The data will be secure in confidential but their symptoms are needed. The algorithm will match symptoms with the corresponding correct disease and confidence rating. 

People with rare diseases like me need help. Anyone with a rare disease can benefit from this. Not only the patients but doctors will spend less time trying to guess and modify their prognosis to fit the disease and more time focusing on the patient. 

My solution directly aligns with this challenge. I have planned this for years and finally have a location to share my insight with this challenge. The population I want to serve and the solution I designed directly address misdiagnosis and correctly allocates resources where they are needed most,  the patient. 

I am going to school to finish my Cyber Security & Computer Science Masters and need the resources and time to complete my project. I don't have money to hire or collect data. I need financial stability and growth to be able to complete my dream. 

There is currently not any redily available way for doctors to review symptoms reliably without vast research and hundreds of NIH (National Institute of Health) documents. This product centralizes the data allowing for simple and intuitive use. 

The product is simplistic in design using readily available data the AI uses a deep learning algorithm to scour data and predict disease based of their corresponding symptoms.

https://github.com/leanderme/s...

Similar symptom checkers exist but are not used in medical environments. 

The system uses preexisting and new data uploaded regularly to the NIH. All personal information is scrubbed and verified by the NIH. Collecting more data could introduce security vulnerabilities but this can be mitigated during the collection step by not asking personal information to be given.

Currently none.

I hope that if I receive the resources needed to develop my solution the product could be in trial phases in the next 5 years. Without financial stability the product cannot come to market or testing. 

Sadly, since I'm unable to make more than just ideas and minor programming progress my measured output is low.