Case Study: Detecting brain hemorrhage in Digital Imaging and Communications in Medicine by using Artificial Intelligence.
Hi! Recently I was working with Neurocare.AI is American medical startup. Dr. Junaid Kalia is a leading person in the start-up and is a neurosurgeon. He is a very talented and knowledgeable person felt good working with him. Neurocare.AI is working on digital health Clinic services I mention the link to the Neurocare.AI website you can see it.
The idea behind the project is that we will create a service for a brain hemorrhage detection system by using an artificial intelligence module that detects brain hemorrhage in DICOM images. The first problem is that to label the data DICOM images that are used to train over a module for detecting brain hemorrhage so In our first phase we need 30,000 brain X-rays and CT scans. we need a web application that will be able to show and label those 30000 scans and after that, we have implemented the AI algorithm. In the end, we have one API that is used for a brain hemorrhage detection system. For release web applications a lot of open source libraries are present one of the most used is OHIF viewer that we use for creating a web application. Simply we cloned OHIF open-source library and with some modifications and adding a labeling module at the end we will create a web service to label the data you can see the link below
One of the biggest challenging things for us was that there is no open source library present for mobile applications because there is a wonderful idea behind mobile applications. So I have done a deep study of Cornerstone Library DICOM images DICOM tags how DICOM images are working a lot of research and hard work are involved to understand everything. In the end, we created our own mobile application that shows Dicom images and controls window width window leveling, and this kind of thing manually and views automatically with control modality from mobile we control the system and sites from mobile filtering data filtering by patients so this is a lot of functionality we create for mobile application. Actually, the idea behind the mobile application is that doctors have this mobile application when patients have CT scans from a medical laboratory so this CT scan is automatically uploaded to Cloud which is a pipeline created specially for this application on Google Cloud. before showing this image on the doctor’s mobile application AI algorithm Apply to this image and detecting the brain hemorrhage or anything like that and get a result before doctor observation so that after doctor observation the doctor can observe the high alert patient and high thread patients. And treat them as first priorities. you can see this whole process on the link below
For online demo
Best Regards
Muhammad Umar Hayat
Director V3 Solution
www.v3solution.com
www.developerumarhayat.com
