Meta-Diagnostic For Identification And Preprocessing of medical image data for in-silico trials (DIVE-MED)
Problem: Clinical studies are essential for the successful launch of new medical products, but are often time consuming and associated with high costs due to, among others, manual identification of suitable patients and data processing.
Goal: The project aims at accelerating and simplifying the selection of patients for clinical studies in the field of neuroradiology while increasing data protection.
Method: With the help of deep learning (DL) techniques, medical images from the Picture Archiving and Communication System (PACS) of the clinic will be automatically preprocessed, depersonalized and evaluated for inclusion into clinical studies.
By reducing required human resources while speeding up clinical trials, DIVE-MED would essentially promote earlier access for patients to treatments with newer technologies.
Related Publications:
- Nielsen et al. (2021):
Deep Learning-Based Automated Thrombolysis in Cerebral Infarction Scoring: A Timely Proof-of-Principle Study
STROKE. 2021;52(11):3497-3504. - Nielsen et al. (2020):
Time Matters: Handling Spatio-Temporal Perfusion Information for Automated TICI Scoring
MICCAI 2020. 1. Aufl. Springer Nature Switzerland, 86-96.