Online time and resource management based on surgical workflow time series analysis

Area of Science:

• Medical Informatics
• Surgical Workflow Analysis
• Time Series Forecasting

Background:

• Operating room (OR) management is critical for hospital efficiency, particularly resource and time allocation.
• Current OR management relies on manual processes, leading to inefficiencies in this cost-intensive unit.

Purpose of the Study:

• To enhance hospital effectiveness and efficiency through automated resource and time management in the OR.
• To develop an exact online forecast of surgeon resource usage and remaining intervention time.

Main Methods:

• Utilized time series analysis and a linear time-variant system for novel online approach.
• Calculated power spectral density and spectrogram of surgical perspectives (e.g., instrument usage) to compare workflows.

Main Results:

• Predicted remaining intervention time for lumbar discectomy with an error of 21 min 45 s ±9 min 59 s, focusing on right-hand movements.
• Achieved 74% sensitivity and 75% specificity in forecasting technical resource usage within 20 minutes, analyzing instrument use.

Conclusions:

• Automated surgical workflow recording has minimal intraoperative impact, as not all perspectives require recording.
• Predictions aid OR staff and technicians, potentially reducing resource conflicts and improving patient care.