Algorithm 1: Bone diagnosis technique using hybrid SFNet

1: Create a fractured and healthy image dataset. 2: Apply augmentation technique rotation, flip horizontal, flip vertical and scaling to increase the size of the dataset. 3: Find the edge in an image using the improved canny edge algorithm discussed in Section 4.3. 4:  For I = 1 to 20 train the model       (a): Input grey and canny images to hybrid SFNet       (b): Apply Equations (9) and (10) to convert logits into probability values       (c): Calculate training and validation loss for each epoch using equation 155: Find overall training accuracy using the equation discussed in Table 3. 6: Find overall validation accuracy using the equation discussed in Table 3. 7: Find the loss of the hybrid SFNet. 8: Plot a training and validation loss graph for 20 epochs.