Diseases of the pancreas can have serious consequences for patients and are difficult to diagnose. Early detection is crucial for patients' chance of survival. In Europe today, more than 90 per cent of patients die from pancreatic cancer within the first five years after diagnosis. The Medi-Globe Group, a leading international medical technology company based in Achenmühle, Bavaria, will collaborate with the French Institut Hospitalo-Universitaire (IHU) Strasbourg to develop the world's first AI software to detect pancreatic disease during endoscopic ultrasound (EUS) examinations. The key technology here is artificial intelligence. As part of the large-scale clinical trials on the way to market maturity, the AI software could now go into testing on patients, the so-called first-in-human application.
"The new software has the potential to revolutionise the diagnosis of pancreatic diseases and especially the diagnosis of pancreatic cancer. This would be a dream come true for the whole society and an important milestone in medicine and medical technology," enthuses Prof. Dr. Leonardo Sosa Valencia, one of the world's leading gastroenterologists and a top expert in the field of pancreatic disease detection.
"With the AI software, we will be able to provide gastroenterological specialists with a virtual expert in the future and give them the opportunity to significantly improve the diagnostics that determine therapy. Very good results have already been achieved in the initial clinical applications. Small lesions have been detected, even in complex cases. This gives great hope to the treating physicians and, above all, affected patients and will sustainably improve patient care in the medium term," says Dr Markus Schönberger, Director Business Development at the Medi-Globe Group.
Deep Learning: Clinical experience of world-leading gastroenterologists
The AI software brings together the clinical experience of the world's most experienced gastroenterologists. It bundles the knowledge from countless EUS examinations of the pancreas into an algorithm and transforms it into an application for endoscopic ultrasound examinations. In the process, the expert knowledge is used to train a neural network according to the Deep Learning method. The innovative diagnostic software created according to this method is intended to find details during the examination of the pancreas that often remain hidden from the doctor's view. For diagnosis, it is crucial to detect pathologically altered areas in the pancreas at an early stage so that tissue samples can then be taken with pinpoint accuracy, e.g. by means of a fine-needle biopsy, the evaluation of which is decisive for the decision on therapy.
"The interpretation of EUS images is demanding, strongly dependent on the user and requires several years of training and specialisation. The partnership with Medi-Globe Group aims to demonstrate the utility of modern AI approaches in the analysis of these complex images and to support gastroenterologists to detect pancreatic cancer more effectively," explains Prof. Dr. Nicolas Padoy, Director of AI Research at IHU Strasbourg.
"We cooperate very closely with leading international university hospitals. With the help of artificial intelligence, together with the researchers at IHU, we will be able to make the clinical experience knowledge of world-leading gastroenterologists available to all doctors in the future. We are thus consistently expanding our role as an innovation leader in minimally invasive diagnostics and therapy," says Marc Jablonowski, Chief Technology & Innovation Officer (CTIO) at the Medi-Globe Group.
Organisator: international calendar