Software analyzed brain tumors within minutes – Ingenieur.de – The news portal for engineers

While a doctor takes up to 60 minutes to analyze a brain tumor on a magnetic recording, the new fully automatic BraTumIA computer program can determine the malignant tissue structures within only ten minutes to the last detail. The software can even perform calculations dimensional – making them far unique in the world

software to optimize therapies

Behind the software put doctors and engineers from Bern.. They belong to the local University Hospital and the Institute for Surgical Technology and Biomechanics (ISTB) of the University of Bern and want to improve the treatment of tumors. “The precise segmentation of the tumor tissue allows us to use the image more and more accurate information to optimize therapy,” explains Prof. Roland Wiest, neuro-radiologist at the University Hospital. “This is hugely important because new treatment strategies for gliomas – malignant tumors – on accurate growth data of the tumors are dependent.”

software constantly learning

BrainTumIA joins the trend of medical Data Mining a. The software continuously improved by collecting new statistical data – similar to the usual data mining, in which data are collected to buying habits of customers on the Internet. This technology has made it the software to ensure accuracy with respect to international attention. The US daily Washington Post she asked as showpiece regarding data mining in medical and analytical range before

For brain tumor patients, the medical data mining is crucial. Analyze doctors images of the tumor tissue manually, fallacies are theoretically in different directions. The software, however, makes most mistakes that always go in the same direction. These doctors can quickly and precisely check and reduce to a minimum.

Currently, the research group is working on two new versions of the software for patients with stroke and multiple sclerosis (MS). In MS patients, the software inflamed brain tissue in the cerebral white matter (plaque) to recognize. In stroke patients, they should predict which brain areas remain damaged expected after treatment.