Date/time
Tuesday 30th September 2025 at 0800 CDT (US Central) / 0900 EDT (US Eastern) / 1400 BST (UK) / 1500 CEST (Central Europe) / 1600 EEST (Eastern Europe) / 1830 IST (India) / 2100 CST/AWST (China / Western Australia) / 2300 AEST (Eastern Australia).
Chaired by Matt J Dunn (Cardiff University).
Agenda
- Welcome
- Membership
- ISCEV collaboration
- Scope
- Context
- ISCET’s purpose
- Election of subcommittee chair
Meeting recording
Attendees
15 people attended the meeting.
Notes
Chairperson
- Only 15 of the 27 subcommittee members were present, so it was agreed the election will take place offline via short bios and an asynchronous vote.
- It was also suggested that a vice-chair role be created to ensure continuity, especially given the clinical workload of potential chairs.
- The wider ISCET group had previously specified that the chair should be an active clinician, but some members noted that non-clinicians with deep eye-tracking expertise could be equally valuable. The requirement may be revisited.
Collaboration with ISCEV
- ISCET is modelling its process on ISCEV, which has published clinical electrophysiology standards since 1989.
- ISCEV and ISCET plan a joint guideline.
- The initial publication will be called guidelines rather than standards, with the option to elevate later.
Purpose and scope of guidelines
- Guidelines are intended as defaults, not binding rules.
- They define how to collect and process data, not when to test or how to interpret results (to avoid liability and maintain clinical flexibility).
- The primary audience includes new clinicians, reducing the steep learning curve in clinical eye tracking.
- Guidelines are clinical-only. Research use is outside the remit, though guidelines may indirectly become defaults for research, as they have with ISCEV.
Protocols and structure
- Protocols should lead to datasets sufficient for clinical decision-making, while remaining flexible for difficult cases (e.g., infants, low vision, structural anomalies).
- Debate on organisation:
- Disorder-based (‘nystagmus’ guideline).
- System-based (fixation, saccades, pursuits, etc.), since many tests are the same across diseases.
- For now, the subcommittee will work on the whole set relevant to nystagmus, as a set. Organisation of the ‘library’ of ISCET guidelines can happen at a later date.
- Emerging consensus: define branches or modules (tests) that clinicians can select depending on the clinical question, without prescribing order.
- Example questions: Is this nystagmus? Where is the null zone? What is foveation quality?
- These questions will serve as the starting point for protocol design.
Minimum vs ideal protocols
- Discussion of whether guidelines should set minimum requirements, ideal standards, or both.
- Minimum hardware standards may be lower than expected if valid clinical information can still be derived.
- Ideal versions may later be defined and then simplified back for practical use.
Validity and quality metrics
- It was suggested to define validity measures (robustness/error metrics, physiological plausibility checks) rather than rigid hardware criteria.
- This would allow for technological flexibility and future developments, while ensuring data reliability. The minimum requirements would need to be clearly understood by clinicians/manufacturers.
- Calibration in patients with abnormal fixation was highlighted as a major challenge.
Next steps
- The subcommittee will work on a draft document comprising the questions that a clinician may need to answer in relation to nystagmus, with detail on the tests that would need to be performed for each question. The final guideline may not necessarily include these questions, but this provides a starting point to discussing which protocols need to be defined.
Next meeting
The next meeting will be arranged by the subcommittee chair, once elected.
Chat
16:06:14 From Matt J Dunn : https://iscev.wildapricot.org/resources/Documents/guidelines%20and%20policies/ISCEV%20Process%20for%20Standards,%20Guidelines%20_%20Extended%20Protocols.pdf
16:42:53 From Thom Wilcockson : I’m afraid I need to leave. Please share the meeting recording. Thanks.
16:43:27 From Larry Abel : Some systems have variable sample rates (e.g., Tobii Glasses); those probably should be avoided
16:46:21 From Helena Lee : I think we have to remember that we are trying to make EMR usable and accessible to new (non expert) clinicians in busy clinics. Clinicans like trees and branches, and guidelines mean that there is a choice and its not prescriptive.
16:49:32 From selfj : Reacted to “I think we have to r…” with 👍
16:57:50 From Mervyn Thomas : Thanks for putting this together. Apologies I have to leave to another meeting
17:05:27 From Helena Lee : Agree with Larry. Alot of qualitative interpretation is very useful. With OCT scans we often just look at qualitative features/abnormalities even though there are ways of segmenting every single layer. When scans arent centred perfectly – automated segmentation is not great but we still get really useful qualitative info.
17:05:44 From Rebecca McLean : Reacted to “Agree with Larry. …” with 👍
17:08:05 From Larry Abel : Reacted to “Agree with Larry. Al…” with 👍
17:08:15 From Larry Abel : I don’t think our recordings need to be quantified to answer some of the relevant clinical questions. I wasn’t talking about phone videos but recordings of eye movement position versus time with an eye tracker.
17:08:41 From Sian Handley : Agree with everyone about the value of uncalibrated data in kids. I suggest this is something than is learnt with experience and can be explained / caveated in a guideline that would be beneficial for people new to the area (as Helena brought us back to)
17:09:32 From Amanda Douglass : Pebble pad?
17:10:08 From Andreas Sprenger : I totally agree in case you want to know whether or not a Nystagmus is present. If you want to document the Progress of a disease or an Intervention (e.g., surgery) you Need to have calibrated data.
17:11:35 From Rasha Moustafa : Reacted to “I totally agree in c…” with 👍
17:12:35 From Larry Abel : Replying to “I totally agree in c…”
Agreed; it’s just that no one who has sent me patients to record has asked for that. But the data usually could be calibrated offline if that was wanted. But nystagmus type, null location, periodicity all can be analysed with relatively uncalibrated data.
17:12:54 From Amanda Douglass : Replying to “I totally agree in c…”
Looking at it over time is asking a different question
17:14:53 From Amanda Douglass : Perhaps having a clinician as one of the chair or deputy?
17:14:59 From Fiona Bríd Mulvey : I think that so long as the person is looking relatively straight ahead, the small difference in spatial precision between calibrated and uncalibrated data wouldn’t be hugely significant for detecting nystagmus. If the person’s gaze is far from central position, the difference would grow – with increasing error with increased gaze (eye in head) angle
17:15:03 From selfj : Reacted to “Perhaps having a cli…” with 👍
17:15:38 From audrey.bonnan : Reacted to “I totally agree in c…” with 👍
17:16:15 From selfj : Sorry. Have to leave. Thanks Matt and all.