Article
Sensor-OR: Towards Data-Driven Workflow-Recognition in the Connected Operating Room
Search Medline for
Authors
Published: | April 21, 2016 |
---|
Outline
Text
Background: Until recently the operating room (OR) was considered as a black box. Only retrospectively, manually written documentation reported what happened during the surgery. No possibilities to measure quantitative data existed, due to this lack of of diditalisation no automatic documentation was possible. Also, it was impossible to predict the estimated time of surgery without interrupting the surgeon. Previous approaches to address this problem used self-developed sensors in the OR which is difficult to standardize and to employ in a non-research-setting. We used a new approach to record medical device data during surgical interventions directly from medical devices in an integrated OR.
Materials and methods: The foundation for data recording was a clinically used sensor-OR. This was an OR equipped with the OR1 FUSION™ (KARL STORZ GmbH & Co. KG, Tuttlingen, Germany) enhanced by a data logger to collect the medical device data. Different devices were connected in the OR, as devices specifically needed for laparoscopy e.g. laparoscopic camera, light-source, thermoflator, electric devices, as well as devices always present in an OR (table and light). Device parameter recorded by the data logger were updated up to a 50ms interval and stored in a database.
Results: In the OR n=14 laparoscopic surgeries of different types (e.g. pancreatic, reflux, colorectal-surgery) were performed and data from up to seven devices was recorded. In total 2.359.133 datapoints of 220 parameter were recorded for 30 hours of surgery.
Conclusion: Within this approach, available data from existing devices in an integrated OR was recorded without additional sensors. The setup needed only one additional device (data logger) in the OR. This data could be used for quantitative analysis of a surgery, leading to automatic documentation or updates regarding the progress of a surgery. This detailed information about an operation could help to predict postoperative complications if intraoperative events can be detected properly. Also, given the possibility to equip not only a single OR, but several ORs with data loggers the idea of an integrated operating floor arises. By data-driven process analysis the logistics of the OR could be optimized. At the same time, it is important to be aware of the sensitivity and the amount of data being possibly recorded. To analyze this amount of data, powerful analysis methods (Big Data) will be necessary to provide conclusions concerning the process or the patient outcome. Additionally surgeons will have to develop a model of data protection, due to the sensitivity of data for both, patient and surgeon.
Acknowledgements This work was carried out with the support of the Federal Ministry of Education and Research (BMBF) as part of the OR.NET-project, by the German Research Foundation (DFG) as part of project A01 in the SFB/TRR 125 Cognition-Guided Surgery and by the Federal Ministry for Economic Affairs and Energy (BMWi) as part of the InnOPlan-project. We thank the company Karl Storz GmbH, Tuttlingen, Germany for providing the data logger.