Using MID and high level visual features for surgical workflow detection in cholecystectomy procedures /
Sherif Mohamed Hany Shehata
Using MID and high level visual features for surgical workflow detection in cholecystectomy procedures / استخدام صفات بصرية متوسطة و عالية المستوى لتمييز المراحل الجراحية فى عمليات استئصال المرارة بالمنظار Sherif Mohamed Hany Shehata ; Supervised Fathi Hassan Saleh , Nicolas Padoy - Cairo : Sherif Mohamed Hany Shehata , 2016 - 53 P. : photographs ; 30cm
Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Computer Engineering
We present a method that uses visual information in a cholecystectomy procedures video to detect the surgical workflow. While most related work relies on rich external information, we rely only on the endoscopic video used in the surgery. We fine tune a convolutional neural network and use it to get mid-level features representing the surgical phases. Additionally, we train DPM object detectors to detect the used surgical tools, and utilize this information to provide discriminative high-level features. We present a pipeline that employs the mid and high level features by using one vs all SVMs followed by an HHMM to infer the surgical workflow. We present detailed experiments on a relatively large dataset containing 80 cholecystectomy videos. Our best approach achieves 90% detection accuracy in offline mode using only visual information
Cholecystectomy Deformable part models Surgical workflow
Using MID and high level visual features for surgical workflow detection in cholecystectomy procedures / استخدام صفات بصرية متوسطة و عالية المستوى لتمييز المراحل الجراحية فى عمليات استئصال المرارة بالمنظار Sherif Mohamed Hany Shehata ; Supervised Fathi Hassan Saleh , Nicolas Padoy - Cairo : Sherif Mohamed Hany Shehata , 2016 - 53 P. : photographs ; 30cm
Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Computer Engineering
We present a method that uses visual information in a cholecystectomy procedures video to detect the surgical workflow. While most related work relies on rich external information, we rely only on the endoscopic video used in the surgery. We fine tune a convolutional neural network and use it to get mid-level features representing the surgical phases. Additionally, we train DPM object detectors to detect the used surgical tools, and utilize this information to provide discriminative high-level features. We present a pipeline that employs the mid and high level features by using one vs all SVMs followed by an HHMM to infer the surgical workflow. We present detailed experiments on a relatively large dataset containing 80 cholecystectomy videos. Our best approach achieves 90% detection accuracy in offline mode using only visual information
Cholecystectomy Deformable part models Surgical workflow