Computer Assisted Instruction in Laparoscopic Surgery using Deep Learning

Minimally invasive surgery (MIS) is famous to cause less harm the skin compared with regular operation, due tiny surgical instruments and small incisions used. It provides many advantages patients like a shorter hospital stays, reduced pain faster recovery. In addition, MIS offers possibility of video record surgery. These videos are used for teaching purposes, evaluating surgeons also they treated as evidence in case lawsuits from patients. On other hand, these types surgeries difficult learn teach. That’s why tend check possible technical error. Since medias commonly very long, this manual quality assessment (SQA) process, without any support search, take so much time effort. To surmount issue, we present neural network based solution, identify index videos, using three fine-tuned Convolutional Neural Network VGG19, Inception v-4 NASNet-A. Finally, benefits proposed approach on Cholec80 dataset.