The article reports on a head-to-head comparison of two optical coherence tomography (OCT) systems in a pediatric craniosynostosis population. The goal? To see if a handheld device can reliably reproduce measurements from a table-top instrument, supporting non-invasive monitoring for suspected intracranial hypertension.
By looking at both quantitative optic nerve metrics and qualitative signs of raised intracranial pressure, the study investigates intermachine agreement and reproducibility. It also explores the potential for pooling data across devices to improve clinical care and research.
Study design and methods
This prospective cross-sectional study evaluated reproducibility and intermachine agreement between the widely used Spectralis table-top OCT and the handheld Envisu C2300 system. The study included 20 school-aged children with craniosynostosis, examining 40 eyes in total.
Each participant underwent sequential imaging with both devices. The order was randomized, and all scans were evaluated by masked clinicians to reduce bias.
The protocol combined quantitative optic nerve and peripapillary measurements with qualitative grading for signs of intracranial hypertension (IH). Scans went through semi-automated segmentation, followed by a scale-factor correction to align measurements reported in visual angles across devices.
All analyses accounted for within-subject correlation. The workflow aimed to reflect practical clinical routines, not just idealized research conditions.
Participants and imaging protocol
All eligible children completed imaging on both platforms with high success. There was 100% recruitment and bilateral imaging success.
Clinicians evaluating the scans were masked to device identity, and imaging order was randomized. The study captured a broad set of metrics, including cup depth, disc width, nasal and temporal peripapillary RNFL thickness, and other axial nerve head parameters.
Quantitative analysis and metrics
Quantitative analyses relied on semi-automated segmentation with a scale-factor correction to harmonize units between Spectralis and Envisu. Measures were presented in visual angles for standardization.
The statistical approach included intraclass correlation coefficients (ICCs), limits of agreement (LOA), paired t-tests, and coefficients of variation (CV). These metrics helped gauge intermachine reproducibility and any systematic biases.
Key quantitative findings and agreement
The study found that intermachine reproducibility was good to excellent for most OCT parameters, with many ICCs over 0.90. A few metrics, like disc width and nasal/temporal RNFL thickness, showed lower reliability, with ICCs in the 0.81–0.87 range, but still within clinically acceptable limits.
They did spot small systematic biases. Cup depth and some axial measures tended to be 1.4–3.1% smaller on the Spectralis, while some lateral measures (like cup and disc widths) were 3.5–8% larger on the Envisu. Still, the percentage differences and CVs were low, so these biases probably won’t alter clinical decisions in most cases.
- Excellent reproducibility for most optic nerve metrics (ICCs > 0.90).
- Lower but acceptable reliability for disc width and nasal/temporal RNFL (ICCs 0.81–0.87).
- Small, systematic biases between devices, but with low variation and minimal clinical impact.
- Scale-factor correction successfully aligned measurements across platforms.
Qualitative grading and usability
Expert qualitative grading for signs of IH achieved 100% intermachine agreement. That’s a pretty robust and transferable grading scheme.
Surprisingly, an untrained novice reached similar performance after just a quick training session. That says a lot about the usability and accessibility of the qualitative assessment protocol in different clinical settings.
Clinical implications and future directions
The findings support the feasibility and reliability of both table-top Spectralis and handheld Envisu OCT systems for evaluating optic nerve parameters in children with craniosynostosis. This interchangeability could have a real impact on clinical practice and research integration, making non-invasive IH screening more accessible and paving the way for multicenter studies that pool data from different device platforms.
- Data pooling across handheld and table-top OCT can enhance statistical power and surveillance programs for IH in craniosynostosis.
- Short imaging times and high success rates make routine screening in pediatric clinics more practical, reducing the need for invasive intracranial pressure measurements.
- Intermachine agreement encourages collaborative trials and long-term follow-up across institutions using different OCT technology.
Training and accessibility
The fact that an untrained novice could achieve similar grading performance after brief instruction shows real promise for scalable training programs. Non-specialist staff could help with IH monitoring even in resource-limited settings.
Conclusion
Kids with craniosynostosis can use both Spectralis table‑top OCT and Envisu C2300 handheld OCT to check optic nerve and peripapillary features. Both machines work well and give similar results.
Honestly, that’s pretty promising for clinics and researchers. Non-invasive OCT might help catch intracranial hypertension sooner, without jumping straight to invasive procedures.
Here is the source article for this story: Feasibility and reproducibility of handheld and table-mounted optical coherence tomography in children with craniosynostosis