title: Eye Tracking in Surgery
status: established
last_updated: 2026-05-31
sources: Bapna 2023 Eye Tracking Surgery Review, Tolvanen 2022 Eye Tracking Workload Surgery
tags: [eye-tracking, surgery, workload, expertise, systematic-review, fixation, pupil-dilation, blink-rate, applied]

Eye Tracking in Surgery¶

Status: established
Last updated: 2026-05-31
Sources: Bapna 2023 Eye Tracking Surgery Review, Tolvanen 2022 Eye Tracking Workload Surgery
Tags: [eye-tracking, surgery, workload, expertise, systematic-review, fixation, pupil-dilation, blink-rate, applied]

Summary¶

Eye tracking in surgical contexts is applied to two main purposes: assessing workload and stress during procedures, and differentiating levels of surgical expertise. Systematic reviews confirm that pupil responses, blink rate, and gaze patterns (fixation count, fixation duration, dwell time) are valid indicators of both workload and skill level. Most studies use simulated environments; evidence from real operating theatres remains limited, and the field lacks standardised metric parameters and analysis techniques.

Body¶

Context¶

This article draws on two PRISMA-compliant systematic reviews of eye tracking in surgery that address different questions. Tolvanen et al. (2022) reviewed 17 studies (2010–2020) on eye tracking as a workload indicator; Bapna et al. (2023) reviewed 80 studies (January 2000 – December 2022, from 12,054 screened records) on eye tracking across the surgical disciplines, with an emphasis on differentiating expertise and on gaze training. Within this knowledge base the article is the applied surgical case of two foundational strands: the workload arm draws directly on the cognitive-load mechanisms set out in Pupil Dilation Cognitive Load, and the expertise arm rests on the spatial-gaze metrics whose extraction is treated in Fixation Saccade Detection. It sits alongside the other applied-domain articles, Gaze Based Hci And Usability and Visual Occlusion Attentional Demand.

Key Points¶

On workload, Tolvanen et al. (2022) concluded that pupil responses, blink rate, and gaze indices are valid indicators, with surgeons' pupil responses, gaze patterns, and blinks consistently associated with perceived workload (PDF p. 2, orig. p. 1341). Twelve of their studies reported pupil responses and four reported blink data (PDF p. 3, orig. p. 1342); blink rate was found to be controlled by task difficulty and is described as a potential indicator of focus and effort in surgery (PDF p. 6, orig. p. 1345). Their reading of pupil dilation as a workload index is consistent with the broader TEPR literature (see Pupil Dilation Cognitive Load).

On expertise, Bapna et al. (2023) found that number of fixations, fixation duration, dwell time, and cognitive workload differentiated novice from expert performance (PDF p. 1, orig. p. 2600). Across their corpus 17 studies addressed training, 48 assessment, and 20 the technology itself, with 72 of the 80 conducted in simulated environments and 26 different eye-tracking devices used (PDF pp. 1–2, orig. pp. 2600–2601). Experts showed longer fixation periods on key anatomical areas and directed more of their gaze to the surgical scene rather than to their instruments — a strategy termed target locking — while measures such as gaze entropy changed as performance improved over repeated robotic-simulator sessions (PDF p. 5, orig. p. 2604).

On gaze training, Bapna et al. (2023) report that eight studies demonstrated its effectiveness, with novices taught more expert-like behaviours including target locking, linked to improved performance (PDF p. 1, orig. p. 2600; PDF pp. 5–6, orig. pp. 2604–2605).

On limitations, Tolvanen et al. (2022) note that 15 of their 17 studies measured workload in simulated settings, with reported differences between operating-room and simulated measurements, so findings may not fully generalise (PDF p. 2, orig. p. 1341). They also identify distractions and other non-technical factors as recognised underlying factors in successful surgery that the literature underrepresents (PDF p. 2, orig. p. 1341). Bapna et al. (2023) reach a parallel methodological conclusion: utility is broad but the literature lacks standardisation of metric parameters, hardware, and gaze-analysis techniques (PDF pp. 1–2, orig. pp. 2600–2601).

Conclusion¶

The two reviews are complementary, covering adjacent purposes — workload assessment (Tolvanen et al., 2022) and expertise differentiation plus training (Bapna et al., 2023) — and they agree on the central limitations. Both find eye-tracking metrics valid for their respective purpose, and both converge on the same unresolved problems: most evidence comes from simulation rather than real theatres, and the field has no standardised metric battery or analysis method. The combined picture is a validated but not yet standardised toolset whose generalisation to the social environment of real surgery remains open.

References¶

Bapna, T., Valles, J., Leng, S., Pacilli, M. & Nataraja, R. M. (2023) 'Eye-tracking in surgery: A systematic review', ANZ Journal of Surgery, 93(11), pp. 2600–2608. doi: 10.1111/ans.18686. bapna2023surgery

Tolvanen, O., Elomaa, A-P., Itkonen, M., Vrzakova, H., Bednarik, R. & Huotarinen, A. (2022) 'Eye-tracking indicators of workload in surgery: A systematic review', Journal of Investigative Surgery, 35(6), pp. 1340–1349. doi: 10.1080/08941939.2021.2025282. tolvanen2022surgery

Open Questions¶

What standardised metric battery would allow comparison of gaze-based competency assessment a