VertebrAI is a cloud service to assist radiologist for the detection of DDD using Deep learning with X-ray and MRI images.
Our mission is to create an automated diagnostic platform to identify the severity of disk degeneration disease (DDD). We aim to make degeneration grading more accurate and data driven. DDD is a very prominent issue in America. Unfortunately only 57% of those those with evidence of DDD are diagnosed. This indicates the need for innovation within industry to increase the diagnosis performance. Despite the changing landscape of health insurance. The prices for MRI and X-ray scans are incredible high. Thus deterring many working class Americans from getting diagnoses. Further, the inefficiency of the entire process in both the physicians and patients’s POV drives the low diagnosis rate. We aim to decrease radiologists interpret time which will consequently decrease appointment waiting time for patients. Due to minimal blood supply to intervertebral disks, once they are damaged, minimal healing happens. The best treatment is to prevent degeneration occurrence through early diagnosis, which our product will enable.
What it does
We built a cloud service to detect DDD using pattern recognition and deep learning with X-ray and MRI images as input. We estimate that we could reach above a 95% accuracies trained using publicly available datasets.
How I built it
We collect and train 2000 images dataset with pattern recognition and deep learning models as a cloud service. We also build a website front end to utilize the service for easy use.
Challenges I ran into
Collect dataset, recruit radiologists to label data, etc.
Accomplishments that I'm proud of
The idea, our teamwork and spirits.
What I learned
Collaboration, time management.
What's next for VertebrAI
We will collaborate with hospitals, and collect dataset of X-ray and MRI images simultaneously with radiologists’ grading to train our algorithms.