Adapting the procedure of multifocal visual evoked potentials (MFVEP) for use in children; maximising cooperation and diagnostic information

Visual field (VF) testing in the paediatric population is a clinical problem that previously has not been addressed. Gold standard testing in adults; Humphreys visual field testing is a subjective test that relies on cooperation from the patients, making them impractical in younger children. Objective measures such as the visual evoked potential (VEP) are not spatially sensitive. Researchers have attempted to address this with multifocal visual evoked potentials (mfVEP), but there was a limitation of fixation stability and concluded the need for an attractive fixation.

This project developed the adapted mfVEP, which replaced the central 10° of the visual stimulus with a DVD input and utilised eye movement recording to filter out data, where fixation fell outside this window. This thesis validated this technique by comparing it to standard mfVEP using a test-retest design with Wilcoxon sign rank, inter-class correlation coefficient and Bland-Altman plots to compare amplitude, latency, reproducibility, ability to detect a pseudo defect and for the unaffected clusters to retain reproducibility. The results showed high agreeability and similar reproducibility on the clinically significant value of amplitude. The latency of the adapted mfVEP was up to 7.5ms earlier than standard mfVEP but produced similar reproducibility measures.

In the pseudo defect groups, which masked hemifield and quadrant field, the amplitude measure was able to detect all defects and retain reproducibility of the unaffected clusters. The latency measure was highly variable and did not show good detection or retainment of reproducibility.

Preference was assessed using qualitative interview technique and showed preference for the adapted mfVEP as participants felt more relaxed and the stimulus was easier to look at and fixate on compared to standard testing that requires effort to maintain fixation; not to blink or look away.

Finally, a case series was presented which showed that the adapted mfVEP could detect a quandrantopia in a 6-year-old child, but also could detect other hemifield and functional visual loss. It showed the adapted mfVEP was consistent with the findings of VF and VEP testing, but also was able to detect additional information if the VF could not be performed or if the VEP was unable to detect a defect. Good fixation was achieved with children as young as 1-year-old, but the limiting factor of poor signal to noise (SNR) ratio made these un-interpretable. Although in all these cases some form of VEP was successful, in addition the VEP could predict the successfulness of the mfVEP. If poor VEPs either SNR or degraded signals, then the mfVEP would not be recordable.

In conclusion a technique was developed that improves fixation of the mfVEP and show its application in the paediatric population. There are several limitations, and future work should focus on repeating this study in healthy and non-healthy paediatric patients, considering behaviour aspects. Furthermore, to determine normative data, critical clinical values, and sensitivity/specificity. Also to improving SNR by optimising montages, stimulus sensitivity and recording times to make this more applicable to younger age ranges.

Outputs

• Poster Presentation at ECCN
• Oral Presentation at ANS 75th Anniversary