Sign in Subscribe
Machine learning

Artificial Intelligence Aids Minimally Invasive Surgery with Advanced Camera Detection

Strategic Insights

— 1 min read

Researchers at the University of Alberta have successfully developed an artificial intelligence system to detect camera rotation in minimally invasive surgery (MIS) procedures, mitigating the difficulty and cognitive load on surgeons. This innovative approach leverages a convolutional neural network (CNN) to automatically correct camera rotation issues during MIS procedures. The system achieved an accuracy of 96% on a dataset of 2116 video frames from trans-nasal MIS procedures.

Key Takeaways:

  • The proposed system uses a CNN to detect camera rotation, addressing a significant challenge in MIS procedures.
  • The dataset consisted of 2116 video frames, with 497 frames labeled as 'tilted' and 1619 frames as 'non-tilted.'
  • The ResNet50 model was trained on the dataset for 10 epochs, achieving an accuracy of 96.9% at epoch 6.
  • The final F1 score was 0.94, and the Matthews Correlation Coefficient was 0.9168, with no significant bias toward either class.
  • The trained ResNet50 model demonstrated a high success rate in predicting significant camera rotation without favoring the more frequent class in the dataset.
  • This research establishes a foundation for developing an automatic correction system for camera rotation in MIS procedures.
  • Zhong Shi Zhang and his team at the Surgical Simulation Research Lab, University of Alberta, led the research.
  • Yun Wu and Bin Zheng were also contributors to this research.

Statistics:

  • 96.9% accuracy at epoch 6
  • Validation loss of 0.0242 before validation accuracy began to decrease
  • 96% accuracy on the test set
  • Average loss of 0.0256 on the test set
  • F1 score of 0.94
  • Matthews Correlation Coefficient of 0.9168
  • Dataset consisted of 2116 video frames
  • 497 frames labeled as 'tilted'
  • 1619 frames labeled as 'non-tilted'
  • 10 epochs of training
  • ResNet50 model achieved 96.9% accuracy

Sources:

  • Information (Journal) - http://www.mdpi.com/journal/information/
  • "Automatic Detection of Camera Rotation Moments in Trans-Nasal Minimally Invasive Surgery Using Machine Learning Algorithm." (Information, 2025,16(4):303)
  • Medical Devices & Surgical Technology Week (News Report) - "Research on Machine Learning Described by Researchers at University of Alberta (Automatic Detection of Camera Rotation Moments in Trans-Nasal Minimally Invasive Surgery Using Machine Learning Algorithm)" (May 18, 2025, p 978)

Read more