Analysis of Clear Corneal Incision Architecture with Anterior Segment Spectral-Domain OCT Theodore Leng, MD, Jianhua Wang, MD, PhD, Sonia H. Yoo, MD, Brandon.

Analysis of Clear Corneal Incision Analysis of Clear Corneal Incision Architecture with Anterior Segment Architecture with Anterior Segment

Spectral-Domain OCTSpectral-Domain OCT

Theodore Leng, MD, Jianhua Wang, MD, PhD, Sonia Theodore Leng, MD, Jianhua Wang, MD, PhD, Sonia H. Yoo, MD, Brandon Lujan, MD, Aizhu Tao, MD,H. Yoo, MD, Brandon Lujan, MD, Aizhu Tao, MD,

Gavriil Tsechpenakis, PhDGavriil Tsechpenakis, PhD

Bascom Palmer Eye Institute, University of Miami Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL USAMiller School of Medicine, Miami, FL USA

The authors have no financial disclosures. This presentation discusses the use of an experimental The authors have no financial disclosures. This presentation discusses the use of an experimental medical device that has not yet been approved by the FDA.medical device that has not yet been approved by the FDA.

PurposePurposeTo use a prototype AS-SDOCT device to To use a prototype AS-SDOCT device to image bi- and tri-planar CCIs constructed for image bi- and tri-planar CCIs constructed for phacoemulsification and to reconstruct those phacoemulsification and to reconstruct those wounds in three-dimensional (3D) space wounds in three-dimensional (3D) space

Advantages of AS-SDOCT• Higher axial resolution• Faster acquisition speed

• Allows for 3D scans Wound features can be analyzed across the full

dimensions of the incision Ability to reconstruct wound in 3D

Materials and MethodsMaterials and MethodsA prototype 1310 nm wavelength AS-SDOCT instrument was constructed and mounted A prototype 1310 nm wavelength AS-SDOCT instrument was constructed and mounted onto a conventional slit lamp for imaging of the anterior segment. The device had an axial onto a conventional slit lamp for imaging of the anterior segment. The device had an axial resolution of 8 resolution of 8 μμm and was able to acquire real-time two-dimensional images at 14 m and was able to acquire real-time two-dimensional images at 14 frames/second and full 3D datasets in approximately 7 seconds. AS-SDOCT datasets of frames/second and full 3D datasets in approximately 7 seconds. AS-SDOCT datasets of 100 B-scans, each consisting of 512 A-scans, were acquired from each patient on post-100 B-scans, each consisting of 512 A-scans, were acquired from each patient on post-operative day one after uncomplicated cataract extraction by phacoemulsification. Each operative day one after uncomplicated cataract extraction by phacoemulsification. Each 3D scan consisted of a 6 x 6 x 3 mm volume of data. The experimental protocol was 3D scan consisted of a 6 x 6 x 3 mm volume of data. The experimental protocol was approved by an institutional review board and all patients underwent an informed consent approved by an institutional review board and all patients underwent an informed consent process and signed a consent form. process and signed a consent form.

The Prototype AS-SDOCT DeviceThe Prototype AS-SDOCT Device

The device (black box) as seen mounted on a conventional slit lampThe device (black box) as seen mounted on a conventional slit lamp

ResultsResultsEn face En face image of the corneaimage of the cornea

B-scan depicting a CCI (arrow)B-scan depicting a CCI (arrow)

Please see movie file Please see movie file for 3D manipulation of for 3D manipulation of datadata

The incision is traced on individual B-scans to create a surface The incision is traced on individual B-scans to create a surface depiction of the CCI geometrydepiction of the CCI geometry

Data AnalysisData Analysis• Compared surgical technique to actual wound Compared surgical technique to actual wound

shape on OCT scansshape on OCT scans• Analyzed wounds for characteristics possibly Analyzed wounds for characteristics possibly

associated with ingress of fluid and associated with ingress of fluid and endophthalmitisendophthalmitis• 13 “bi-planar” wounds scanned13 “bi-planar” wounds scanned• 10 “tri-planar” wounds scanned10 “tri-planar” wounds scanned

Loss of CoaptationLoss of CoaptationBi-PlanarBi-Planar• 1 of 131 of 13

Tri-PlanarTri-Planar• 0 of 10 0 of 10

GeometryGeometryBi-PlanarBi-Planar• 3 of 13 (23%) had 3 of 13 (23%) had

true bi-planar true bi-planar geometrygeometry• Remainder had Remainder had

curvilinear curvilinear

Tri-PlanarTri-Planar• 6 of 10 (60%) had true 6 of 10 (60%) had true

tri-planar geometrytri-planar geometry• Of the remainder:Of the remainder:

Combined Bi and Tri (1)Combined Bi and Tri (1) Pure bi-planar (1)Pure bi-planar (1) Curvilinear (2)Curvilinear (2)

Wound GapeWound Gape

Epithelial SideEpithelial Side• 1 of 13 bi-planar 1 of 13 bi-planar

• 1 of 10 tri-planar 1 of 10 tri-planar

Endothelial SideEndothelial Side• 1 of 13 bi-planar 1 of 13 bi-planar

• 5 of 10 tri-planar 5 of 10 tri-planar

MisalignmentMisalignment

Bi-PlanarBi-Planar• Epithelial - NoneEpithelial - None• Endothelial - 2 of 13 Endothelial - 2 of 13

Tri-PlanarTri-Planar• Epithelial – NoneEpithelial – None• Endothelial – 1 of 10Endothelial – 1 of 10

Stromal EdemaStromal Edema

Bi-PlanarBi-Planar• 3 of 133 of 13• Greater in Roof - 2Greater in Roof - 2• Greater in Floor - 1 Greater in Floor - 1

Tri-PlanarTri-Planar• 4 of 104 of 10• Greater in Roof - 3Greater in Roof - 3• Variable throughout Variable throughout

width of wound - 1width of wound - 1

Descemet’s DetachmentsDescemet’s Detachments

Bi-PlanarBi-Planar• 9 of 139 of 13

Tri-PlanarTri-Planar• 2 of 102 of 10

ConclusionsConclusions• Tri-planar techniques were likely to result in true tri-planar geometry on Tri-planar techniques were likely to result in true tri-planar geometry on

AS-SDOCT scansAS-SDOCT scans

• Tri-planar CCIs had a higher incidence of wound gape than bi-planar Tri-planar CCIs had a higher incidence of wound gape than bi-planar CCIsCCIs

• Neither technique had a high rate of loss of coaptationNeither technique had a high rate of loss of coaptation

• Both techniques had an equal rate of misalignment and stromal edemaBoth techniques had an equal rate of misalignment and stromal edema

• Bi-planar technique was more likely to result in Descemet’s Bi-planar technique was more likely to result in Descemet’s detachmentsdetachments

Future Directions• Improved wound reconstructions• Calculation of wound surface area (which may be related to

leakage and rates of endophthalmitis)• Measurement of wound lengths and angles• Decrease image acquisition time to reduce motion artifiact