DeepKnee autonomous MRI reporting system
We have developed a software framework for autonomous medical image interpretation based on deep learning technology. Our focus is on the assessment of MRI exams of the extremities without human intervention.
Currently our easily integrable software enables the fully automatic, prompt & robust annotation of the relevant anatomic details and detection of the most frequent knee lesions: ACL & meniscal tears, Baker cysts.
Recognition of cartilage abnormalities will be available by the end of 2022, and medial collateral ligament tear by 2023Q2.
We are also working on the most frequently occurring ankle lesions: cartilage problems of the talar dome & tears of the lateral ligaments
You don’t have to change your favorite dicom viewer to look at the suggestions of deepKnee. Our software can be seamlessly integrated in your PACS environment without any interruption in your workflow.
Our framework is currently integrated and under test in the systems of the following companies and institutions:
- Semmelweis University
- Pécs Diagnostic Center
- Budai Health Center
- MedicalPoint Ltd.