SAMIRIX

SAMIRIX performs semi-automatic intraretinal segmentation of spectral-domain OCT macular 3D volume scans to quantify intraretinal layer thicknesses for research on neurodegenerative and neuroinflammatory diseases affecting the optic nerve and retina.

Key Features:

• Modular pipeline: Composed of modular components including import filters and a segmentation module for intraretinal analysis.
• Import filters: Tailored for optical coherence tomography (OCT) data, specifically spectral-domain OCT macular 3D volume scans from the Heidelberg Engineering Spectralis OCT.
• Segmentation algorithm: Integrates a third-party segmentation algorithm for intraretinal layer delineation.
• Manual correction: Supports manual correction of automatic segmentation results with reported intra-class correlation coefficient > 0.99 across all layers.
• Batch processing: Supports batch operations for processing multiple OCT volumes.
• Thickness computation: Computes intraretinal layer thicknesses including macular thickness (MT), macular retinal nerve fiber layer (mRNFL), ganglion cell and inner plexiform layer (GCIPL), and inner nuclear layer (INL) within a 6 mm ETDRS circle.
• Normative data generation: Produces and stores normative retinal layer thickness values while accounting for age, sex, and refractive error.

Scientific Applications:

• Disease research: Quantification of retinal layer changes in neurodegenerative and neuroinflammatory conditions, including multiple sclerosis.
• Normative reference generation: Establishment of normative macular intraretinal layer thicknesses from a cross-sectional cohort of 218 healthy volunteers (144 females, 74 males, mean age 36.5 ± 12.3 years), reporting MT 313.70 ± 12.02 μm, mRNFL 39.53 ± 3.57 μm, GCIPL 70.81 ± 4.87 μm, and INL 35.93 ± 2.34 μm.
• Reference dataset provision: Provides a normative database usable as a reference in studies that lack internal normative data.

Methodology:

Import filters for OCT data are applied to spectral-domain OCT macular 3D volume scans from the Heidelberg Engineering Spectralis OCT, a third-party segmentation algorithm is used for layer delineation, manual corrections are applied to segmentation results, batch processing is supported, and intraretinal layer thicknesses are computed within a 6 mm ETDRS circle.