Fig. 4

Model Interface. When an image is submitted to the server as shown in (A), the model segments the flap within the photograph as illustrated in (B), and displays the probability of vascular compromise along with the most appropriate classification grade accordingly.