Morphometrical evaluation of the choriocapillaris imaged by swept-source optical coherence tomography angiography

Figures & data

Figure 1 Analyzed images.

Notes: (A) An original image cropped to 800×800 pixels centered at the central fovea from 1,024×1,024 pixels (3×3 mm) images. (B) The image was adjusted by the contrast limited adaptive histogram equalization method in the Fiji package. (C) Vessel segmentation for image B, performed with AngioTool.

Table 1 Parameters in healthy individuals and patients’ groups

	Healthy individuals (N=35)	Patients (N=32)	P-value
Age (y±SD) (range)	27.9±5.8 (22–49)	59.1±12.8 (34–82)	<0.001
Gender (M/F)	13/22	17/15	0.225
SE (D±SD)	−1.23±1.17	−0.86±1.61	0.295

Abbreviations: D, diopters; SE, spherical equivalent.

Table 2 Measurement of flow void by thresholding

		Number of flow voids				Average size (×10^−4 mm^2)				%FV
		Mean	SD	CV	ICC	Mean	SD	CV	ICC	Mean	SD	CV	ICC
HI	First Second	2,552 2,560	192 177	7.5 6.9	0.873	3.20 3.15	1.25 1.16	39.2 36.9	0.950	14.62 14.44	4.57 4.35	31.2 30.1	0.915
P	First Second	2,252 2,252	197 188	8.7 8.3	0.923	4.56 4.64	1.52 1.47	33.3 31.8	0.958	18.29 18.65	4.47 4.44	24.4 23.8	0.957
HI & P	First Second	2,409 2,413	245 238	10.2 9.9	0.938	3.85 3.86	1.54 1.51	39.9 39.1	0.964	16.37 16.45	4.85 4.85	29.6 29.5	0.945

Abbreviations: %FV, total area size of flow void/total measured area; CV, coefficient of variation; HI, healthy individuals; ICC, intraclass correlation; P, patients.

Table 3 Measurement parameters and reproducibility in images treated with local contrast enhancement

	Vessel area (mm^2)					Total vessel length (mm)				Vessel diameter index				Mean lacunarity
	Mean	%	SD	CV	ICC	Mean	SD	CV	ICC	Mean	SD	CV	ICC	Mean	SD	CV	ICC
Healthy individuals, intensity 15, N: 35 eyes ×2=70 images
2	3.566	(65.1)	0.036	1.0	0.711	221.1	4.057	1.8	0.863	16.1	0.204	1.3	0.718	0.0095	0.0006	6.8	0.894
3	3.096	(56.5)	0.030	1.0	0.758	177.7	5.205	2.9	0.860	17.4	0.365	2.1	0.823	0.0162	0.0011	7.0	0.870
4	2.897	(52.9)	0.034	1.2	0.835	138.7	5.235	3.8	0.840	20.9	0.588	2.8	0.829	0.0254	0.0017	6.6	0.833
5	2.807	(51.2)	0.040	1.4	0.836	111.5	5.249	4.7	0.765	25.2	0.897	3.6	0.727	0.0349	0.0022	6.2	0.802
2,3	3.745	(68.5)	0.039	1.0	0.793	216.7	3.908	1.8	0.870	17.3	0.197	1.1	0.592	0.0089	0.0006	7.1	0.912
3,4	3.336	(60.9)	0.036	1.1	0.852	175.1	5.183	3.0	0.877	19.1	0.389	2.0	0.843	0.0155	0.0011	6.9	0.874
2,3,4	3.939	(72.1)	0.041	1.0	0.832	207.8	3.587	1.7	0.869	19.0	0.182	1.0	0.663	0.0085	0.0006	7.2	0.901
Healthy individuals, intensity 20, N: 35 eyes ×2=70 images
2	3.477	(63.5)	0.043	1.2	0.548	220.9	4.075	1.8	0.873	15.7	0.215	1.4	0.664	0.01	0.0007	7.0	0.845
3	3.045	(55.6)	0.034	1.1	0.660	177.2	5.199	2.9	0.853	17.2	0.363	2.1	0.770	0.0167	0.0012	7.2	0.857
4	2.859	(52.2)	0.037	1.3	0.807	138.5	5.242	3.8	0.830	20.7	0.584	2.8	0.812	0.026	0.0018	6.7	0.828
5	2.772	(50.6)	0.043	1.5	0.789	111.7	5.265	4.7	0.777	24.9	0.883	3.6	0.746	0.0358	0.0023	6.3	0.786
Patients, intensity 15, N: 32 eyes ×2=64 images
2	3.460	(63.1)	0.094	2.7	0.646	218.6	6.362	2.9	0.960	15.8	0.500	3.2	0.793	0.0105	0.0013	12.3	0.865
3	3.042	(55.5)	0.056	1.8	0.804	173.7	5.039	2.9	0.897	17.5	0.462	2.6	0.856	0.0179	0.0013	7.2	0.892
4	2.867	(52.3)	0.052	1.8	0.914	135.2	4.400	3.3	0.820	21.2	0.551	2.6	0.769	0.0278	0.0016	5.8	0.887
5	2.788	(50.9)	0.059	2.1	0.882	108.6	4.517	4.2	0.818	25.7	0.801	3.1	0.733	0.0377	0.0021	5.6	0.862
2,3	3.676	(67.1)	0.058	1.6	0.937	208.9	4.840	2.3	0.944	17.6	0.187	1.1	0.817	0.0106	0.001	9.3	0.964
3,4	3.298	(60.2)	0.031	0.9	0.907	167.2	3.997	2.4	0.879	19.7	0.327	1.7	0.822	0.0176	0.001	5.5	0.928
2,3,4	3.864	(70.5)	0.060	1.5	0.932	200.3	4.273	2.1	0.934	19.3	0.190	1.0	0.792	0.0101	0.001	9.4	0.961
Healthy individuals and patients, intensity 15, N: 67 eyes ×2=134 images
2	3.515	(64.2)	0.088	2.5	0.779	219.9	5.410	2.5	0.933	16.0	0.403	2.5	0.808	0.01	0.0011	11.2	0.896
3	3.070	(56.1)	0.052	1.7	0.847	175.8	5.488	3.1	0.892	17.5	0.415	2.4	0.842	0.017	0.0015	8.7	0.921
4	2.883	(52.6)	0.046	1.6	0.899	137.0	5.143	3.8	0.850	21.1	0.590	2.8	0.814	0.0265	0.002	7.6	0.906
5	2.798	(51.1)	0.051	1.8	0.870	110.1	5.107	4.6	0.801	25.4	0.883	3.5	0.746	0.0362	0.0025	7.0	0.878
2,3	3.712	(67.8)	0.060	1.6	0.925	213.0	5.837	2.7	0.951	17.4	0.247	1.4	0.814	0.0097	0.0012	12.2	0.975
3,4	3.318	(60.6)	0.038	1.1	0.903	171.4	6.090	3.6	0.928	19.4	0.492	2.5	0.911	0.0165	0.0015	8.9	0.950
2,3,4	3.903	(71.3)	0.063	1.6	0.934	204.2	5.423	2.7	0.950	19.1	0.249	1.3	0.847	0.0093	0.0012	12.4	0.972

Note: The italicized and bold numbers in the ICC columns indicate values over 0.8.

Abbreviations: CV, coefficient of variation; ICC, intraclass correlation.

Figure 2 Comparisons of the results analyzed using AngioTool, with vessel thickness 3–4 and intensity 15 between healthy individuals and the patients.

Notes: (A) Vessel area. (B) Total vessel length. (C) Vessel diameter index. (D) Mean lacunarity.

Figure 3 Regional variations of morphometrical results in healthy individuals.

Notes: (A) Assignment of the measurement zones. The square image of 800×800 pixels was divided into 64 squares (each square is100×100 pixels), and the 64 squares were assigned to four zones. (B–D) Histograms of the morphometrical measurements of each small square (100×100 pixels) in each zone are shown. (B) Vessel area. (C) Total vessel length. (D) Vessel diameter index. (E) Mean lacunarity. *P<0.05.

Figure 4 Morphometrical evaluation of focal lesions.

Notes: (A) An original image (800×800 pixels) of an eye with resolved central serous retinopathy. Images (B–D) correspond to the 50×50 pixel yellow-lined square B, C, and D on the image (A). Images (B–D) are contrast adjusted by contrast limited adaptive histogram equalization. (E–G) corresponding to (B–D) are segmented images from AngioTool. The measurements were performed with vessel thickness 3–4 and intensity 15. The vessel borders are outlined in yellow. The vessel skeletons are outlined in red. The blue areas represent the branching points of the vessels. (H–J), corresponding to B, C, and D lettered on image (A), are the images resulting from thresholding using Phansalkar’s method. With increasing %FV, VD decreased and VDI increased.
Abbreviations: %FV, the percentage of flow void measured on the image thresholding; VD, vessel density; VDI, vessel diameter index.