Machine learning projects

We developed a web service for a Belgian retail company. This web service enables our client to organise and maintain their database of 3 million products based on product image analysis

To automate document flow to Peppol and Doccle, IxorThink designed a natural language model for IxorDocs. Our system detects client and payment information in invoices and other documents.

In collaboration with J.Delabie PhD, Toronto, we are training a model to recognise and count blast cells in bone marrow smears.

In colaboration with Dr. De Paepe, AZ Sint Jan we are developing a ML model that classifies dysplastic colon polyps. A model that lies at the base of a whole slide screening tool.

Together with Prof. Raedt, UGent, we aim to develop a ML model that will automatically detect and annotate epileptic seizures in EEG signals of rat and mice.

IxorThink developed a web based platform. The platform is used for collecting data and showing results of our machine learning models. The platform uses a vendor neutral whole slide scan viewer as well as a viewer for time series.