Breast Volume Measurement by Recycling the Data Obtained From 2 Routine Modalities, Mammography and Magnetic Resonance Imaging

PATIENTS AND METHODS

Breast volume measurement using MRI

Breast 3.0-T MRI scans were conducted in a prone position. A medical imaging system (AQI Viewer, Terarecon Inc, Foster City, CA) was used for MRI volume measurement. An outline of the mammary gland was plotted manually using about 15 slices per breast. The area of interest was integrated into a 3D image by the imaging system, and the breast volume was calculated using a volumetric tool (Fig 1). Measurements were performed using axial and sagittal views, and the average was taken to be the MRI breast volume (V_MRI). The time required for each measurement was recorded.

Figure 1. Breast volume measurement by magnetic resonance imaging. (a) Outline of breast tissue. (b) Reconstructed 3-dimensional image and calculation of the breast volume.

Mammographic volume measurement

Breast volume based on MMG data was calculated using the formula proposed by Kalbhen et al⁴: V_MMG = (π/4) × Breast height × Breast width × Compression thickness in craniocaudal MMG. This calculation is based on the assumption that the shape of the compressed breast in the craniocaudal projection can be regarded as a half-elliptical cylinder (Fig 2).

Figure 2. Breast volume measurement by mammography. The breast volume is calculated using a formula reported by Kalbhen et al.4VMMG indicates breast volume measured using mammography; H, breast height (cm); W, breast width (cm); and C, compression thickness (cm).

RESULTS

The mean age of the patients was 53.34 years (range, 31-84 years). The mean value of V_MMG and V_MRI was 325.26 ± 217.36 mL and 318.47 ± 199.4 mL, respectively, and that of V_sp measured intraoperatively was 318.86 ± 181.82 mL. The mean errors were 64.98 ± 62.71 mL for V_MMG and 27.94 ± 26.97 mL for V_MRI. Scatter plots of V_MRI and V_MMG with V_sp are shown in Figure 3 respectively. Both methods showed linear associations with V_sp, with a slightly higher Pearson correlation coefficient for V_MRI (0.982; 95% confidence interval [CI], 0.960-0.992; P < .05) than for V_MMG (0.911; 95% CI, 0.802-0.961; P < .05). The average time required for measurement using MRI and MMG was 160 and 115 seconds, respectively.

Figure 3. Scatter plot of VMRI and Vsp. The correlation coefficient (r) was 0.982 (P < .05). VMRI indicates breast volume measured using magnetic resonance imaging; Vsp, volume of mastectomy specimen.

DISCUSSION

Accurate and objective measurement of breast volume is helpful for choosing an appropriate breast implant and determining the transplanted volume of autologous tissue in breast reconstruction surgery. Several methods for measuring breast volume have been described, but most require further examination, time, cost, and an extra burden on patients. MMG has been commonly used for breast cancer diagnosis for many years, and in the past 2 decades, MRI has been increasingly used for diagnosis and evaluation of tumor extent in breast cancer, as recommended by various national standards and guidelines.^1-3 MRI screening for high-risk patients is recommended in the guidelines updated in 2008 by the American Cancer Society because MRI is more sensitive than MMG. More recently, MRI has been recommended as the first choice for estimation of tumor extent in breast cancer, especially in Japan.

In the current study, we evaluated the accuracy of breast volume measurement using MMG and MRI because use of these routine modalities reduces the financial, physical, and mental burden on patients and the data can be obtained easily in a clinical setting. Breast MRI is performed in a prone position, with the breast outlined to grasp the extent and characteristics of the tumor. The contrast of breast tissue and skin is depicted clearly, which allows for simple plotting of the outline of breast tissue. Our results showed a strong linear association between breast volumes of surgical specimens and those measured by 3D MRI. The 3D MRI model of the breast can be easily developed, and the breast volume calculations are simple and rapid and based on routine preoperative data. The time taken to calculate the breast volume with MRI data was only 160 seconds, which indicates that the method is simple for both patients and medical staff.

We added 2 refinements to previous measurements of breast volume using MRI.^6,7 First, our study is prospective. Second, the surgical specimen volume was measured with a more accurate method in accordance with Archimedes’ principle and Newton's third law.⁵ This is in contrast to methods in which the surgical specimen volume is obtained by simply dividing the weight by a constant density of breast tissue⁶ or by assuming an approximate weight.⁷ The proportion of breast glandular tissue and fatty tissue is likely to vary among individuals. Therefore, our method provides stronger evidence that breast volume measurement with 3D MRI is accurate. Nevertheless, MRI has a disadvantage of requiring specific analysis software.

Our findings showed that breast volume measurement with MMG is accurate despite the lower accuracy than MRI. We used a relatively simple formula based on a half-elliptical cylinder model among several proposed formulas.^4,8,9 Given that the breast volume calculation from MMG is simple, is not highly variable among individuals, and can be performed without extra time, cost, and discomfort for patients, MMG is a candidate modality to provide information for planning reconstructive breast surgery when MRI or volume analysis software is unavailable.

To the best of our knowledge, this is the first study of breast volume measurement using data from routine modalities of MMG and MRI. Estimating breast volume using a combination of these two modalities may facilitate more accurate and simple breast reconstruction.