Clinical assessment of breast symmetry and aesthetic outcome: can 3D imaging be the gold standard?

Figures & data

Figure 1. Pre-marking one of the anatomical landmarks (the xiphoid process) prior imaging.

Figure 2. (a) Analysis in VECTRA Analysis Module, measurement of the breast area of interest (turquoise) for calculations of the mean distance between the two breast surfaces ($d_{\mathit{RMS}}$) as an expression for breast symmetry and (b) volume measurement of the left breast against the interpolated virtual chest wall (brown). Skin marks adjusted in relation to the xyz-plane.

Figure 3. Illustration of the coordinates and data used to obtain the mean distance between two breast surfaces ($d_{\mathit{RMS}}$) and the characteristic diameter ($d_{\mathrm{b}}$) of the left (blue) and the right (maroon) breast, calculated from the measured volumes of the breasts in VECTRA Analysis Module. $d_{\mathit{RMS}}$ is calculated based on measuring the shortest distance between more than 1000 points on one surface to the corresponding points on the other surface.

Table 1. Clinical data of the study sample.

Variable	No cancer	Cancer
	n (%)	n (%)
Number of women	36	22
Age at risk-reducing surgery (years)
Range	26.1–62.4	30.2–63.7
Mean	44.2	43.4
Median	43.8	44.2
Age at 3D-imaging (years)
Range	40.7–71.1	42.1–73.1
Mean	56.8	60.9
Median	55.2	55.1
BRCA mutation status
BRCA1/BRCA2/BRCAX^a	24 (67)	15 (68)
No mutation or unknown	12 (33)	7 (32)
BMI
<18.5	2 (6)	1 (5)
18.5 to <25	23 (64)	15 (68)
25–30	3 (8)	4 (18)
≥30	2 (6)	1 (5)
Missing	6 (17)	1 (5)
Bilateral prophylactic salpingo-oophorectomy
Yes	22 (61)	13 (59)
No	14 (39)	9 (41)
Type of breast cancer
In situ		4 (18)
Invasive		17 (77)
Missing		1 (5)
Type of breast cancer surgery prior risk-reducing surgery
Breast conserving cancer surgery		10 (45)
Mastectomy (cancer surgery)		12 (55)
Radiotherapy
Yes		15 (68)
No		6 (27)
Missing		1 (5)
Chemotherapy
Yes		15 (68)
No		5 (23)
Missing		2 (9)
Endocrine therapy
Yes		11 (50)
No		7 (32)
Missing		4 (18)
Unanticipated reoperations after risk-reducing breast surgery
Yes	17 (47)	14 (64)
No	16 (44)	8 (36)
Missing	3 (8)

^a BRCAX: women with family history of breast cancer screened negative for BRCA1/BRCA2.

Table 2. Estimates of agreement and reproducibility of breast symmetry ($d_{\mathit{RMS}}$), left breast volume ($V_L$), and right breast volume ($V_R$), in two different postures, 45° arm abduction and hands placed on hips.

	dRMS,45 (mm)	dRMS,Hip (mm)	VL,45 (cm^3)	VL,Hip (cm^3)	VR,45 (cm^3)	VR,Hip (cm^3)
One-way random effects ANOVA
Observer 1
Mean [SD]	7.1 [2.9]	7.1 [2.8]	282 [102]	281 [90]	281 [110]	289 [107]
Range	2.2–15.7	2.4–16.4	102–580	81–526	62–644	56–565
${\sigma }_{\mathit{patient}}^2$	5.99	5.79	9352	6700	11284	10214
${\sigma }_{\mathit{error}}^2$	2.17	1.83	1217	1408	922	1283
Repeatability coefficient (95% CI)	4.1 (3.6–4.6)	3.8 (3.3–4.3)	97 (85–110)	104 (91–118)	84 (74–96)	99 (87–113)
Intra-observer reproducibility (95% CI)	0.73 (0.63–0.82)	0.76 (0.66–0.84)	0.88 (0.83–0.92)	0.83 (0.75–0.88)	0.92 (0.89–0.95)	0.89 (0.83–0.93)
Observer 2
Mean [SD]	8.0 [3.4]	8.2 [3.3]	337 [129]	346 [139]	323 [116]	331 [120]
Range	2.3–19.6	2.5–19.2	47–685	22–830	79–766	46–849
${\sigma }_{\mathit{patient}}^2$	9.76	9.44	15271	18485	12534	12954
${\sigma }_{\mathit{error}}^2$	1.61	1.73	1674	1068	1015	1521
Repeatability coefficient (95% CI)	3.5 (3.1–4.0)	3.6 (3.2–4.1)	113 (100–129)	91 (80–103)	88 (78–100)	108 (95–123)
Intra-observer reproducibility (95% CI)	0.86 (0.79–0.91)	0.85 (0.77–0.90)	0.90 (0.85–0.94)	0.95 (0.92–0.96)	0.93 (0.89–0.95)	0.89 (0.84–0.93)
Two-way mixed effects ANOVA
${\sigma }_{\mathit{patient}}^2$	6.18	5.66	9132	8478	6910	7757
${\sigma }_{\mathit{observer}}^2$	1.70	1.96	3180	4115	4999	3827
${\sigma }_{\mathit{error}}^2$	1.89	1.78	1445	1238	968	1402
Bias (95% CI)	–0.9 (–1.5 to −0.3)	–1.1 (–1.7 to −0.5)	–55 (–77 to −32)	–65 (–89 to −40)	–42 (–69 to −15)	–42 (–66 to −18)
95% limits of agreement	–6.2 to 4.3	–6.5 to 4.3	–243 to 134	–268 to 138	–256 to 172	–242 to 158
Intra-observer reproducibility (95% CI)	0.81 (0.74–0.86)	0.81 (0.74–0.86)	0.89 (0.85–0.93)	0.91 (0.88–0.94)	0.92 (0.90–0.95)	0.89 (0.85–0.92)
Inter-observer reproducibility (95% CI)	0.63 (0.50–0.75)	0.60 (0.46–0.73)	0.66 (0.53–0.78)	0.61 (0.46–0.75)	0.54 (0.37–0.70)	0.60 (0.44–0.73)

Table 3. Estimates of volume-shape-symmetry ($\mathit{VSS}$) agreement and reproducibility, based on raw data measurements from the three-dimensional surface images including the variance components, presented for two postures (arms at 45° and hands placed on hips).

	VSS45		VSSHip
	Observer 1	Observer 2	Observer 1	Observer 2
Mean [SD]	0.929 [0.029]	0.923 [0.034]	0.930 [0.028]	0.923 [0.032]
Range	0.828–0.975	0.806–0.977	0.820–0.977	0.788–0.980
One-way random effects ANOVA
${\sigma }_{\mathit{patient}}^2$	0.0006	0.0010	0.0006	0.0009
${\sigma }_{\mathit{error}}^2$	0.0002	0.0001	0.0002	0.0002
Repeatability coefficient (95% CI)	0.041 (0.036–0.046)	0.033 (0.029–0.038)	0.038 (0.034–0.043)	0.035 (0.030–0.039)
Intra-observer reproducibility (95% CI)	0.74 (0.63–0.83)	0.87 (0.81–0.92)	0.76 (0.66–0.84)	0.85 (0.78–0.90)
Two-way mixed effects ANOVA
${\sigma }_{\mathit{patient}}^2$	0.0006		0.0006
${\sigma }_{\mathit{observer}}^2$	0.0002		0.0002
${\sigma }_{\mathit{error}}^2$	0.0002		0.0002
Bias (95% CI)	0.006 (–0.0004 to 0.012)		0.007 (0.001–0.013)
95% limits of agreement	–0.046 to 0.060		–0.037 to 0.050
Intra-observer reproducibility (95% CI)	0.82 (0.75–0.87)		0.81 (0.75–0.86)
Inter-observer reproducibility (95% CI)	0.62 (0.48–0.74)		0.61 (0.47–0.73)

Figure 4. (a) Bland–Altman’s plots of the inter-observer reproducibility comparing observer 1 in relation to observer 2 with corresponding limits of agreement (upper and lower dashed lines) for breast symmetry measurements with arms at 45° ($\mathrm{ }{d}_{\mathit{RMSi},45}$), (b) breast symmetry with hands on the hips ($d_{\mathit{RMSi},\mathit{Hip}}$), (c) left breast volume at 45° ($V_{Li,45}$), (d) left breast volume with hands on the hips ($V_{Li,\mathit{Hip}}$), (e) right breast volume at 45° ($V_{Ri,45}$), and (f) right breast volume with hands on the hips ($V_{Ri,\mathrm{ }\mathit{Hip}}$). $i=1$ for measurements by observer 1, $i=2$ for measurements by observer 2.

Table 4. Spearman’s rank correlation (ρ) between breast symmetry ($d_{\mathit{RMS}}$) and breast volume differences ($\Delta V=\mathrm{left}–\text{right breast volume}$) of measurements from six 3D surface images of 58 women captured in two different postures (45° and hands placed on hips) obtained by two observers.

	Observer 1		Observer 2
	Spearman’s ρ	p	Spearman’s ρ	p
Arms at 45°
Image 1	0.138	0.303	–0.090	0.502
Image 2	0.315	0.016	–0.110	0.412
Image 3	0.043	0.752	0.138	0.302
Hands on hips
Image 4	0.330	0.011	–0.041	0.760
Image 5	0.128	0.338	–0.004	0.979
Image 6	–0.015	0.912	–0.140	0.294