Leandro L Di Stasi, PhD1, Carolina Diaz-Piedra, PhD1, Juan F Ruiz-Rabelo, MD, PhD2, Joaquin Roca-Gonzalez, PhD3, Samuel Romero, PhD4, Andres Catena, PhD1. 1Mind, Brain, and Behavior Research Center – University of Granada, Granada, Spain, 2Department of General Surgery, Reina Sofia University Hospital, Cordoba, Spain, 3Electronics Technology Department, Technical University
Objective of the technology or device
Surgeons’ cognitive overload is associated to medical errors and sentinel events. Electroencephalography (EEG) is a reliable tool to monitor surgeons’ cognitive load (CL). However, the ability to objectively measure CL online in healthcare scenarios remains a major challenge. Due to technical and methodological difficulties, very few studies have investigated the effects of surgeons' CL on EEG activity. In recent years, user-friendly and portable EEG devices have overcome many of these barriers, offering powerful instruments to assess CL unobtrusively. Here, we aimed to monitor surgeons' CL by means of their EEG activity. We used a low-cost, wireless, and dry EEG system (MindWave Mobile, NeuroSky Inc., USA) during simulated surgical tasks.
Description of the technology and method
We conducted the study in conformity with the Code of Ethics of the WMA. Fifteen surgeons from the Andalusian healthcare system (7 surgical trainees vs. 8 attending surgeons) participated in the study. EEG signals were recorded while surgeons performed two exercises (Pattern Cut and Peg Transfer: low vs. high complexity), using two surgical procedures (laparo-endoscopic single-site surgery [LESS] vs. multiport laparoscopy surgery [MPS]), with LESS being more cognitively challenging than MPS. The MindWave Mobile offers high recording quality, with a good trade-off between cost (~120€) and device performance. The device consists of a single dry electrode placed on Fp1 referenced to the left earlobe mounted on a light helmet-like system (512 Hz) (Figure 1). The headset uses a Bluetooth connection to send raw data to a recorder unit (smartphone). Artifacts were detected using the first derivate algorithm (we removed portions of the raw data exceeding ±50μV). Then, artifact-free data were segmented in four periods (one for each task), and segmented again in 2-second long epochs. We calculated the power spectra (Hz) for the δ (0.5–4), θ (4.0–8), α (8.0–13), and β (13.0–30) frequency bands. We computed the average power for each task and, finally, the (δ+θ)/(α+β) ratio (DTABr). To analyze the effect of the experience, surgical procedure, and exercise complexity on the DTABr, we performed a single-factor mixed ANCOVA (using sleepiness [Epworth Sleepiness Scale scores] as a covariate).
The interaction between surgical procedure and exercise was significant (p=0.03). DTABr was similar for the Pattern Cut exercise in both surgical procedures (p>0.05), but significantly lower for the Peg Transfer during the LESS procedure (p=0.04).That is, when surgeons –independently of their expertise– performed the most complex exercise with the most demanding procedure, they exhibited greater high-frequency and lower slow-frequencies amount of brain activity. Because task complexity modulates arousal, the most complex situation may have increased surgeons' arousal level, which in turn could have influenced DTABr. Surgical performance showed the same tendency.
Conclusions / future directions
Quantifying surgeon's CL during his/her duties have numerous implications to prevent errors and adverse events. Our preliminary results have potential impacts on the development of neuroergonomic tools to monitor performance in healthcare professionals and in the specifications of future guidelines for residency programs.