A survey of context recognition in surgery

Area of Science:

• Computer Science
• Medical Informatics
• Surgical Technology

Background:

• Context awareness is increasingly important in modern surgical environments, especially with advancements towards 'operating rooms of the future'.
• Understanding the context within surgery is vital for developing intelligent support systems that can enhance surgical performance and patient safety.

Purpose of the Study:

• To systematically organize and review existing approaches for context recognition in surgical settings.
• To identify and categorize different types of context relevant to surgical procedures.
• To provide recommendations for future research directions in surgical context awareness.

Main Methods:

• Conducted a comprehensive literature search of major electronic research databases for publications between 2010 and 2015.
• Identified and categorized studies based on three main types of context: surgical workflow, surgeon's cognitive state, and surgeon's technical state.
• Grouped 52 relevant studies by the type of context detected and the sensors employed.

Main Results:

• Identified three primary contexts: surgical workflow, surgeon's cognitive state, and technical state.
• Summarized various approaches for context recognition, highlighting the sensors and methods used.
• Found that current methods and evaluations have room for improvement, particularly in machine learning applications for cognitive state recognition.

Conclusions:

• Recommends increased use of machine learning for recognizing complex relationships in surgeon's cognitive states.
• Advocates for improved validation protocols, including more in-situ evaluations and a higher number of participants.
• Provides a structured overview of recent context recognition methods to aid the development of next-generation context-aware surgical support systems.