A stack of hundreds of two-minute long ECG signals should be analyzed with a minimum of energy using a machine learning (ML) algorithm:
Based on the concept of the new Federal Agency for Jump Innovations (PSRIN-D), the BMBF initiated three pilot innovation competitions. One of them presented the participants with the task of developing the most energy-efficient AI system possible as a hardware implementation on an ASIC or FPGA. With this, a stack of hundreds of two-minute long ECG signals should be analyzed with a minimum of energy using a machine learning (ML) algorithm: Is the patient healthy or does the recording show atrial fibrillation? It is known that atrial fibrillation is a common cause of strokes. An energy-efficient and inexpensive analysis could therefore prevent many strokes.
Click here for more information.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Leave a Reply