 This study demonstrated the feasibility and potential of using deep learning models based on pre-chemotherapy-contrast-enhanced ultrasound, CEUS, videos to predict the efficacy of neo-adjuvant chemotherapy for pancreatic cancer. Two CNN models were developed, one which incorporated both US and CEUS videos, and another which focused on specific ROIs within CEUS videos. Both models achieved high accuracy, recall, precision, and F1 scores, with the latter achieving higher accuracy than the former. Additionally, 38.2% and 40.3% of the original videos could be accurately classified by the two models when the naked eye was unable to make an accurate distinction. These results suggest that deep learning models can provide valuable insights into the efficacy of neo-adjuvant chemotherapy for pancreatic cancer. This article was authored by Yeming Xiao, Ying Nandang, Wuan Cheng, and others. We are article.tv, links in the description below.