Scanning and processing doctor's handwriting for CM

Processing handwritten medical documents can be a slow and error-prone task, especially when large volumes of information need to be transferred into digital systems. CM wanted to simplify this process and make handwritten doctors' notes easier to process at scale.

Industry

Life sciences & Healthcare

Domains

Customer Service

The challenge of handwritten data

Nowadays everything seems digitalised, but many doctor's certificates are still handwritten. For CM, Belgium's largest health insurance fund, this meant manually processing a mountain of paperwork every single day. These certificates contain essential information needed to process refunds for their 4.5 million members – but deciphering handwritten text is time-consuming, expensive, and prone to errors.

Key takeaways

Increased efficiency and speed in certificate processing.

Improved customer satisfaction due to faster refunds

An intelligent solution

Vectr.Consulting developed an AI-powered solution to help CM dig through this mountain of paperwork. This system can automatically read and transcribe doctor’s certificates by processing images in two phases:

Advanced image processing techniques detect text boundaries and standardise handwriting styles.

A custom-trained AI, based on neural networks and optimised for handwriting, reads and transcribes the text.

Our experts designed this system specifically for CM's needs, combining their data and business knowledge to vastly outperform generic AI models.

Identifying dangerous intersections

To make this analysis even more powerful, Vectr.Consulting developed a system that transforms pixel coordinates from the images into real-world geographical coordinates, allowing them to map the trajectories of different road users.

This detailed analysis of traffic patterns helps MOW identify high-risk intersections and take steps to prevent accidents. By detecting near misses, for example, they can pinpoint locations where infrastructure improvements or targeted awareness campaigns might be needed.

Happier customers, faster refunds

The results speak for themselves, but don't take our word for it! Here's what Roel De Spiegeleer, ICT director at CM had to say about their new system.

"Thanks to the handwriting AI, we are now capable of automatically processing nearly 20% of all A-type doctor’s certificates, which greatly enhances and speeds up our patient reimbursement process.”