Automatic Segmentation of the Eye in 3D Magnetic Resonance Imaging: A novel Statistical Shape Model for treatment planning of Retinoblastoma

摘要

Purpose: Proper delineation of ocular anatomy in 3D imaging is a big challenge, particularly whenuddeveloping treatment plans for ocular diseases. Magnetic Resonance Imaging (MRI) is nowadaysudutilized in clinical practice for the diagnosis confirmation and treatment planning of retinoblastoma inudinfants, where it serves as a source of information, complementary to the Fundus or Ultrasoundudimaging. Here we present a framework to fully automatically segment the eye anatomy in the MRIudbased on 3D Active Shape Models (ASM), we validate the results and present a proof of concept toudautomatically segment pathological eyes.ududMaterial and Methods: Manual and automatic segmentation were performed on 24 images of healthyudchildren eyes (3.29±2.15 years). Imaging was performed using a 3T MRI scanner. The ASMudcomprises the lens, the vitreous humor, the sclera and the cornea. The model was fitted by firstudautomatically detecting the position of the eye center, the lens and the optic nerve, then aligning theudmodel and fitting it to the patient. We validated our segmentation method using a leave-one-out crossudvalidation. The segmentation results were evaluated by measuring the overlap using the DiceudSimilarity Coefficient (DSC) and the mean distance error.ududResults: We obtained a DSC of 94.90±2.12% for the sclera and the cornea, 94.72±1.89% for theudvitreous humor and 85.16±4.91% for the lens. The mean distance error was 0.26±0.09mm. The entireudprocess took 14s on average per eye.ududConclusion: We provide a reliable and accurate tool that enables clinicians to automatically segmentudthe sclera, the cornea, the vitreous humor and the lens using MRI. We additionally present a proof ofudconcept for fully automatically segmenting pathological eyes. This tool reduces the time needed forudeye shape delineation and thus can help clinicians when planning eye treatment and confirming theudextent of the tumor.