Surgical process modelling strategies: which method to choose for determining workflow?

Figures & data

Figure 1. Literature selection procedure and results.

Table 1. Modelling strategies concepts, definitions and dependencies.

Concept	Definition	Criteria and dependencies
	Aspects
Granularity level Data acquisition Model representation Modelling approach Generalization	Description of the procedure at different levels of detail/abstraction.	Purpose of study Data acquisition
	Low to high
Acquiring data of surgical procedure for modelling and analysis.	Purpose of study Granularity level Modelling approach Benefits and drawbacks Available equipment and recourses
Manual/Computer-based
Representation method of modelling surgical procedure	Purpose of study Data acquisition Benefits and drawbacks
Descriptive/Numeric
Direction of modelling (from low to high granularity levels or vice-versa)	Data acquisition Granularity levels Benefits and drawbacks
Top-down/Bottom-Up
Surgical procedure analysis and generalizing the model.	Purpose of study Similarity metrics Statistical analysis Data mining and data warehousing

Figure 2. Different methods of manual data acquisition and the corresponding benefits (indicated with ‘+’) and drawbacks (indicated with ‘−’).

Table 2. Benefits and drawbacks of manual and computer-based data acquisition technologies.

	Benefits	Drawbacks
Manual	Less initial effort for data acquisition (No need to set-up computer-based acquisition systems). Easier data interpretation. Acquisition from different sources other than OR (interview, literature study, etc.). Interactions with clinical team in the OR.	Time consuming data acquisition. Not possible to acquire very low-level data (e.g. at the motion level). Possibility of human error.
Computer-based	Possibility of acquiring very low-level data. Precise data acquisition. Automate data acquisition.	Time consuming and complex data interpretation. Possible error in data recording. Possible error in data interpretation. Time consuming setting up computer-based acquisition systems. Usually physical object attached to the tools, clinical team or patient.

Figure 3. Chains of modelling strategies and their compatible aspects.

Figure 4. Granularity levels used in the study, from lowest level to highest level, from left to right, respectively.

Figure 5. Workflow diagram of the laparoscopic surgery (a) at phase level and (b) at activity level for the “inserting trocars”. The data was obtained from online and offline observations in OUH and Erasmus MC.