The problem Aashirbaad - Living healthier together solves

Many people die every year in Nepal due to a lack of proper health facilities and late diagnosis of disease. Aashirbaad aims to solve:

1. The rural areas of Nepal do not have proper health facilities and health personnel. So, people living in those areas are deprived of primary care and diagnosis facilities. Aashirbaad helps patients to get an early diagnosis of diseases. Also our website recommends nearby hospitals and doctors according to the diseases.

2)Due to increased health problems, patients have to wait for a long time to see their doctor. Due to this, a lot of time is wasted and also is not much efficient approach. To solve this Aashirbaad allows users to find out about their health conditions by simply uploading the scans within few seconds.

3)Doctors need more time to go through each reports manually. Automating this task saves the valuable time of doctor. As the time is saved, doctors can focus on more important tasks.

4. It is costly to visit doctor frequently to show the medical reports time and again. People can get benefitted if they can predict disease using phone. They can save the cost of travelling and also save their time.

Challenges we ran into

We ran into various problems during the project and tried to solve those issues as follows:

1. It was difficult and time consuming to train the neural network. We used Google Collaboratory to train the model faster.

2. It was difficult to get the better accuracy and it needed lots of training and fine tuning of the model to get a good accuracy.

3. The amount of data available was not sufficient to train the deep neural network. As a solution to this, we used theTransfer Learning approach to load the weights from pretrained model. Also, we used Image Augmentation to generate more data.