Deep learning for improved breast cancer monitoring using a portable ultrasound scanner
Insight
SEPTEMBER 20, 2019
Segmentation Since a few patients had multiple images in the dataset, the data were separated, by patient, into three parts: training (80%), validation (10%), and testing (10%). The model was a modified U-Net and trained on GPU hosted by Amazon Web Services (AWS) EC2 instances. on test data. and the recall is 0.85
Let's personalize your content