Different image number for different cases

[hiteshnitetc]

Hi. In nerve segmentation data sets there are data of 47 cases and 120 images of each case with mask and there is no case wise numbering in test data. But in my data for different cases i have different number images (with masks) like for case 1 i have 10 images and for case 2 i have 2 images. Will this effect my segmentation? And one more question is how many images required for a good training, in nerve segmentation data there are there are 5635 images, but i donot have that much images.