The collaboration started on February 8, the two partners said.
The massive medical imaging data produced during the management of AP-HP patients can be made available to clinician-researchers and engineers. They authorize the development of algorithms for medical decision-making, in terms of diagnosis, therapy and predictive clinical development.
“To be usable, these data need to be standardized and structured, whether they come from examinations by scanner (computed tomography), MRI (magnetic resonance imaging), PET (positron emission tomography), ultrasound (ultrasound) or conventional radiology (ionizing rays). This is the work that data managers and engineers of the AP-HP carry out daily in order to control and improve their quality “, underlined the AP-HP and Imageens in a press release joint press release.
Their partnership, signed “in the fall of 2020”, involves the co-development of a software solution supported by AI methods. This software called “Label” automatically labels and classifies imaging data in large quantities to “facilitate” their reuse for studies and research.
The AP-HP and its researchers benefit from the free use of the software and the possibility of exploiting the imaging data thus annotated and classified and the AP-HP has an operating license of ‘Label’ for its uses in research.
In accordance with the protection of patients’ personal data, the start-up Imageens is only authorized to work “on de-identified imaging data”, in order to improve the current version of “Label”, to validate its effectiveness and its reliability in a large-scale application, and to commercialize it – “in the long term” – to other health and research establishments.
Founded in 2017, Imageens is a Parisian medical AI start-up, a spin-off from AP-HP and Sorbonne University. It develops software solutions making it possible to add value to medical imaging data, via two products: “Label”, the Ia tool for processing medical imaging data mentioned above and “ArtFun +”, software for calculating biomarkers. imaging predictive of patient risk of mortality and serious cardiovascular events.