
ML-based Dental Photography Web Solution
Solution: Web system running on Machine Learning and Computer Vision algorithms
Industry: Healthcare
Technologies: React.js, Next.js, TypeScript, Canvas, Node.js, GraphQL
Methodologies: Agile
Project Duration: 12 weeks
Project Team: 1 Designer, 1 PM, 2 Backend developers, 1 Frontend developer, 1 QA
Challenge
DeepSmile Technology, a French dental start-up, wanted to build a web service for dentists and orthodontists, allowing them to obtain high-quality dental images.
Oral healthcare specialists take dental images prior to and after the treatment, to evaluate the patient’s dental condition and the treatment quality. The OralPix system would cut the time and effort spent on taking high-quality images of the patient’s teeth and jaws. Empowered by Computer Vision (CV) and Machine Learning (ML) technologies, the system would align images for a fee.
The company engaged OZiTAG into a ground-breaking new project, which became one-of-a-kind in France.
Solution: Description and Key Capabilities
OZiTAG developed the solution on top of advanced ML algorithms that transform photos into high-quality dental images and help dentists and orthodontists effectively evaluate them.
A client creates a profile that allows him/her to submit images and receive their edited version in a flash. Once a customer logs in and uploads dental images, the system converts and manipulates them to achieve image alignment and decrease the noise in the image and the impact of camera sensitivity. The multi-step process includes image centering, framing, photo straightening, facial trimming for background standardization, and quality improvement. As a result, the client receives a card of 8 photos, edited in accordance with the industry standards, and arranged in a preferable order.
The system supports a hassle-free card payment function, ensured by the Stripe integration. The web app doesn’t store any sensitive card data but sends it to the PCI-compliant payment gateway for a 100%-secure transaction.
We configured a combination of Node.js, TypeScript, and GraphQL for high processing speed and reliable performance, and added rigorous stress-testing to ensure the system would meet the client’s load requirements.
For the frontend part of the solution, our engineers selected React.js and Next.js as an effective and elegant tech stack for a dynamic website, battle-tested by them in complex web projects.
Development Approach
Our team put in place the Agile methodology: we set up a transparent process with regular communication and a clear vision of responsibilities.
Results
With OZiTAG’s help, the client managed to deliver a custom dental photography web solution suited for the needs of dentistry and oral health. The solution enables users to evaluate patients’ treatment with the photos they need, without wasting time and effort on their manual editing.