Preoperative Estimation of Deep Inferior Epigastric Perforator Flap Weight Using Synapse Vincent 3-dimensional Analysis: A Feasibility Study

Summary:

Accurate preoperative estimation of flap weight is essential for successful breast reconstruction using the deep inferior epigastric perforator (DIEP) flap. However, current methods relying on physical examination or experience are often subjective. We developed a simple objective method for predicting DIEP flap weight using Synapse Vincent, a widely available 3-dimensional image analysis software program. This prospective study included 40 patients who underwent autologous breast reconstruction using DIEP flaps between January 2020 and September 2023. A semicylindrical 3-dimensional model of the abdominal subcutaneous tissue was constructed from preoperative computed tomography images using Synapse Vincent. The volume of the model was converted to the predicted flap weight by assuming a fat density of 1.0 g/cm³. The predicted and actual intraoperative flap weights were compared using Pearson correlation and linear regression analyses. The mean predicted flap weight was 732 ± 306 g, and the mean actual flap weight was 754 ± 310 g. A strong correlation was observed between predicted and actual weights (r = 0.938, P < 0.0001). The regression formula was as follows: actual weight = 0.9391 × predicted weight + 68.98. The coefficient of determination (R²) was 0.88, and the root mean square percentage error was 17%, indicating high predictive accuracy. Our method using Synapse Vincent allows accurate and practical preoperative estimation of DIEP flap weights. Its ease of use, broad availability, and compatibility with standard computed tomography imaging make it a valuable tool for surgical planning in autologous breast reconstruction.

Takeaways

Question: Can preoperative computed tomography–based 3-dimensional modeling using Synapse Vincent accurately predict deep inferior epigastric perforator flap weights in autologous breast reconstruction?

Findings: In a prospective study of 40 patients, a strong correlation (r = 0.938, R² = 0.88) was observed between predicted and actual flap weights using a semicylindrical 3-dimensional model. The root mean square percentage error was 17%. Most values fell within acceptable limits on a Bland–Altman plot.

Meaning: Synapse Vincent facilitates accurate preoperative estimation of deep inferior epigastric perforator flap weight, thereby supporting surgical planning and informed patient counseling.

INTRODUCTION

The deep inferior epigastric perforator (DIEP) flap has become one of the most commonly used options for autologous breast reconstruction due to the presence of abundant subcutaneous fat and the reliable vascular anatomy of the deep inferior epigastric artery.^1,2 However, it remains challenging to determine preoperatively whether the lower abdomen contains a sufficient volume of tissue to achieve a satisfactory breast mound, especially in lean patients. Accurate preoperative estimation of flap weight is critical for planning a successful reconstruction, yet conventional methods relying on physical examination or 2-dimensional imaging remain unreliable.^3,4

To address this issue, we developed a novel method for predicting DIEP flap weight using Synapse Vincent, a 3-dimensional (3D) image analysis software program.^5,6 This approach involves constructing a simplified 3D model of the designed flap area based on preoperative computed tomography (CT) scans and calculating its volume. By assuming a tissue density of 1.0 g/cm³ for subcutaneous fat, the estimated volume can be directly converted into a predicted flap weight. In this study, we evaluated the correlation between the flap weight predicted using this method and the actual weight of the harvested flap in a series of breast reconstruction cases.

MATERIALS AND METHODS

The study was approved by the ethics committee of Tokyo Women’s Medical University (approval number: 2020-0088). All patients provided written informed consent to participate in the study.

Patient Selection

This prospective study included 40 consecutive patients who underwent autologous breast reconstruction using DIEP flaps at our institution between January 21, 2020, and September 30, 2023. Both immediate and delayed reconstructions were included.

Estimation of DIEP Flap Weight

Preoperative contrast-enhanced abdominal CT scans were obtained for all patients. A 3D model of the abdominal subcutaneous tissue corresponding to the flap design was generated using Synapse Vincent (Fujifilm, Tokyo, Japan).

The modeling process consisted of the following steps (Fig. 1). (See Video [online], which details instructions for estimating DIEP flap weight using Synapse Vincent.)

Fig. 1.

Preoperative volume simulation using SYNAPSE VINCENT. A 3D semicylindrical model of abdominal subcutaneous fat was created based on preoperative contrast-enhanced CT data. The histogram displays the internal volume distribution. The total volume of the selected region is calculated automatically.

Video 1. This video details the instructions for estimation of deep inferior epigastric perforator flap weight using Synapse Vincent.

Download video file ^{(33MB, mp4)}

1. The software was launched, and the patient’s 3D CT data were loaded.

2. Virtual resection was performed to remove any structures cranial and caudal to the vertical dimension of the planned flap, based on the surgical design.

3. Similarly, the data outside the lateral limits of the flap design were removed, resulting in a semicylindrical 3D model that approximated the actual flap shape.

4. Tissue located beneath the subcutaneous fat layer (eg, muscle, fascia, viscera) was manually removed.

5. Finally, the software automatically calculated the volume of the remaining 3D tissue model, which represented the estimated flap volume.

Assuming a density of 1.0 g/cm³ for subcutaneous fat, the predicted flap weight was considered equivalent to the calculated volume in grams. Intraoperative flap weights were measured using a calibrated sterile surgical scale immediately after flap harvest.

Statistical Analysis

Correlation coefficients between the estimated and actual DIEP flap weights were calculated using Pearson correlation analysis. Linear regression analysis was performed with the estimated DIEP flap weight as the explanatory variable and the actual DIEP flap weight as the objective variable to obtain an estimation formula for DIEP flap weight. The coefficient of determination (R²) and root mean square percentage error were also calculated.

Additionally, agreement between estimated and actual DIEP flap weights was assessed using a Bland–Altman plot. The mean difference (bias) and 95% limits of agreement (mean ± 1.96 × SD) were calculated.

The results were expressed as the mean ± SD. All analyses were performed using GraphPad Prism (version 7.02 for Windows; GraphPad Software Inc., La Jolla, CA). P values of less than 0.05 were considered statistically significant.

RESULTS

A total of 40 patients were included in the analysis. The mean age of the patients was 50.4 (range, 42–69) years, and the mean body mass index (BMI) was 23.3 (range, 17.7–32.5) kg/m². Of these cases, 36 underwent immediate reconstruction and 4 underwent delayed reconstruction. The mean predicted flap weight based on the 3D model was 732 ± 306 g, whereas the mean actual flap weight measured intraoperatively was 754 ± 310 g. An example of a harvested DIEP flap is shown in Figure 2. Pearson correlation analysis revealed a strong positive correlation between the predicted and actual flap weights, with a correlation coefficient (r) of 0.938 (P < 0.0001).

Fig. 2.

Intraoperative photograph of a harvested DIEP flap. The actual flap harvested was based on the preoperative design. The flap includes the umbilicus and infraumbilical skin-fat tissue, corresponding to the region simulated in the Synapse Vincent model.

Linear regression analysis yielded the following estimation formula:

$${\displaystyle \begin{array}{l}\text{A}\mathrm{c}\mathrm{t}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{f}\mathrm{l}\mathrm{a}\mathrm{p}\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}(\mathrm{g})\\ =\text{0}.9391\times \mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{c}\mathrm{t}\mathrm{e}\mathrm{d}\mathrm{f}\mathrm{l}\mathrm{a}\mathrm{p}\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}(\mathrm{g})+\mathrm{68.98.}\end{array}}$$

The R² value was 0.88, indicating a high degree of linear association. Root mean square percentage error was 17%, which was considered clinically acceptable (Fig. 3).

Fig. 3.

Pearson correlation analysis. Pearson correlation analysis revealed a strong positive correlation between the predicted and actual flap weights, with a correlation coefficient (r) of 0.938 (P < 0.0001). Linear regression analysis yielded the following estimation formula: actual flap weight (g) = 0.9391 × predicted flap weight (g) + 68.98. The coefficient of determination (R²) was 0.88, indicating a high degree of linear association. The root mean square percentage error was calculated to be 17%, which was considered clinically acceptable.

Bland–Altman analysis showed a mean bias of 24.08 g, with 95% limits of agreement ranging from −220.3 to 268.4 g (SD = 124.7 g). Most values fell within these limits, indicating acceptable agreement between the predicted and actual weights (Fig. 4).

Fig. 4.

Bland–Altman plot comparing estimated and actual DIEP flap weights. The horizontal axis represents the average of the predicted and actual flap weights, and the vertical axis shows the difference (actual − predicted). The solid line indicates the mean bias of 24.08 g. The upper and lower limits of agreement are 268.4 and −220.3 g, respectively, calculated as the mean difference ± 1.96 × SD (SD = 124.7 g). The majority of data points fall within these limits, suggesting good agreement between predicted and actual values.

DISCUSSION

In this prospective study, we demonstrated that using Synapse Vincent to estimate the preoperative weight of the DIEP flap based on CT imaging provided a highly accurate prediction of the actual harvested flap weight, facilitating the assessment of DIEP flap feasibility. Our method enables objective and reproducible estimation using standard preoperative imaging and is simple enough to incorporate into routine clinical workflows.

One key advantage of this method is its accessibility and practicality. Synapse Vincent is commercially available, with a starting price of approximately 5 million yen, and as of 2024, it has been implemented in more than 1400 medical facilities across Japan. Importantly, this method requires only contrast-enhanced abdominal CT images. This flexibility may improve surgical planning efficiency and enhance preoperative patient counseling. Furthermore, by predicting flap volume, it becomes possible to design a smaller flap, potentially reducing postoperative scarring.

Synapse Vincent provides precise control over segmentation and volume measurement without complex programming or artificial intelligence.⁷ However, the current method assumes uniform fat density and excludes skin and dermis, possibly contributing to prediction bias in some patients. Although this was a prospective study at a single institution, the sample size was limited. Owing to the low prevalence of high-BMI patients in Japan, the BMI range was narrow. Furthermore, as predictions were made by a single observer, the impact of case volume and interobserver variability could not be assessed. Future studies will address these limitations by including more cases and multiple observers.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENT

The authors acknowledge the use of Synapse Vincent 3D in this study.