Although there is limited research on whether digital interventions can effectively prevent acute or subacute pain from developing into chronic pain, Hinge Health is here to prove the opposite! Hinge Health is the complete clinical care solution for personalized physical and behavioral support and pain reduction. To enable its advanced motion technology to better address its patient needs, the company turned to SuperAnnotate to start building more accurate training datasets for its AI.
TLDR: Here is why Hinge Health chose to work with SuperAnnotate:
- Above 96% annotation accuracy: With SuperAnnotate, Hinge Health's annotation quality is consistently above 96% compared to the previous 80%.
- 50% cost savings due to fewer revisions: By spending significantly less on revisions, Hinge Health saves 50% of its budget due to the annotation quality it receives from SuperAnnotate.
- World-class annotation service and DataOps: SuperAnnotate provides annotators with deep human pose and healthcare experience along with a dedicated AI Project Management team, guaranteeing successful delivery even for the most complicated projects.
- Intuitive interface for faster project completion: SuperAnnotate makes your project onboarding a breeze, so you can start labeling industry-specific data in a matter of seconds.
Pioneering the new standard in advanced motion technology
Hinge Health is on a mission to democratize pain reduction. Their app includes everything a patient may require, including a complete clinical care team, a digital MSK (musculoskeletal) clinic, and HingeConnect, the medical data exchange platform powered by Hinge Health. The solution combines wearable detectors and computer vision technology to track a user's movement during exercise therapy programs. With Hinge Health, patients receive feedback on their body/joint movements and activities, which are then monitored by respective care teams to later provide individual virtual physical therapy.
Problem
Consistent, super-high annotation quality is at the core of Hinge Health's computer vision technology. The data team at Hinge Health was challenged to find a solution that would not only ensure high-quality annotations of body parts and joints but would also speed up the quality assurance process without having to give up the cost-efficiency aspect. Hinge Health's objectives were soon met once they got SuperAnnotate on board.
SuperAnnotate as a solution
"The things we were looking for were improving annotation quality and consistency, finding an all-in-one platform that included software and workforce management, and an incredibly knowledgeable team for our difficult use case. SuperAnnotate had it all." Joseph Martorana, data annotation lead at Hinge Health, recalls.
SuperAnnotate’s Data Curation and QA systems were key parts of the SuperAnnotate platform for Hinge Health. “The curation tool/Explore has improved a lot in just the last year, and it is awesome! The efficiency of the QA workflow and the ability to review data and take actions in bulk saves us a tremendous amount of time, speeding up the QA process from 3 days to just an hour,” Martorana mentions.
It was the combination of the SuperAnnotate platform with integrated services that ultimately drew Hinge Health's data team to SuperAnnotate. By providing a vetted workforce ready to take on projects of any volume and complexity, SuperAnnotate was able to help significantly reduce the amount of time Hinge Health's team spent managing the tedious annotation processes while still delivering premium quality annotations on time.
Today, by leveraging the power of SuperAnnotate, Hinge Health is able to locate and track 27 joints (including palms, elbows, and shoulders), effectively measure patients' range of motion, and provide exclusive virtual physical therapy.
Results
SuperAnnotate's marketplace of annotation services helped Hinge Health achieve outstanding results in annotating healthcare motion training data. "Definitely the best annotators I've seen, and I've worked with a lot of companies," Martorana mentioned. The platform’s key points tools, semantic segmentation capabilities, and analytics technologies were incredibly helpful in estimating poses and movement, excluding crowds from target objects, and tracking annotator activity/progress. The results were then used to create 3D models of people to analyze and provide better therapeutic recommendations in virtual physical therapies. Here is what the team at Hinge Health observed throughout the collaboration:
- Cost-efficient: Hinge Health managed to save 50% of its annotation budget while dramatically increasing its labeling accuracy.
- Higher annotation quality/accuracy: The annotation accuracy that SuperAnnotate delivers to Hinge Health is consistently above 95%-96%, compared to the previous 80% with other vendors.
- Exclusive DataOps program and project management: SuperAnnotate's AI Solutions team is always available and adaptable to accommodate changes in terms of timelines, guidelines, and new project priorities.
- Quick to integrate customer feedback: SuperAnnotate takes customer feedback seriously. The team keeps actively implementing customer feature requests and introducing novel functionalities for a smoother platform experience.
Key takeaways
Hinge Health’s achievements were only the beginning of the thriving and long-lasting collaboration with SuperAnnotate. Yet the list of industries trusting SuperAnnotate with their training datasets is growing rapidly.
All of our customers observe positive changes in at least two of the following categories: labeling accuracy/quality, annotation or quality assurance speed, cost reduction while maintaining the quality, and ease of use/overall platform experience.
Follow in Hinge Health's footsteps to become SuperAnnotate's next customer and drive tomorrow's innovation with top-quality training data—SuperData, for your AI.
