目的：研究乳腺肿块的超微血管三维立体成像（smart three-dimensional superb microvascular imaging，Smart 3D SMI）血流分布特点，探讨乳腺癌血流分布模式的三维图像特征。方法：研究经病理证实的145例乳腺癌患者的Smart 3D SMI微血流声像图特征，并结合血流显示区域相应的超声造影（contrast-enhanced ultrasound, CEUS）图像特征，评估乳腺癌的血流分布模式特征。结果：145例乳腺癌患者的Smart 3D SMI检查血流分布模式中，血流最丰富的球形型血流分布模式最多见（58例，40%），其次是分枝型（32例，22%），点线型最少（11例，8%）。进一步比较CEUS与Smart 3D SMI图像上病变区血流分布模式特征，结果显示，CEUS图像均显示有超声对比剂灌注。Smart 3D SMI图像显示为点线型的患者中有约55%在CEUS图像中显示存在粗大血流，而在显示为分枝型、球形型、扩散型、混合型的患者中，CEUS图像显示存在粗大血流的比例分别为75%、72%、78%、82%。Smart 3D SMI各型（点线型、分枝型、球形型、扩散型、混合型）患者，在CEUS图像中显示为无法辨识清晰的分支血流的比例分别为82%、78%、67%、60%和58%。Smart 3D SMI分枝型和球形型的病灶在CEUS上较多表现为血流显示区高强度灌注，与点线型相比差异有统计学意义（P=0.018、0.006）。Smart 3D SMI点线型、分枝型及球形型病灶的灌注区多位于肿块内部，仅少数病例在病灶周边也可见到较强的放射状灌注。结论：乳腺癌的Smart 3D SMI血流分布评估中常见“球形型”和“分枝型”，从超声微血流三维检测的角度提供了乳腺癌微血流的血供分布区域及分布特征，表现出显著的三维形态学特征，为乳腺癌血流分布模式的深入研究提供了更多的信息。

Objective: To study the characteristics of Smart 3D SMI blood flow distribution in breast mass and evaluate the three-dimensional image characteristics of blood flow distribution pattern in breast cancer. Methods: Smart 3D SMI features of 145 pathologically confirmed breast cancer were analyzed. The characteristics of blood flow distribution pattern of breast cancer was evaluated in coordination with the characteristics of contrast enhanced ultrasound (CEUS) pattern of corresponding area. Results: Of the Smart 3D SMI of 145 breast cancer patients, the spherical blood flow distribution pattern with abundant blood flow was the most frequent (58 cases, 40%), then was the branched pattern (32 cases, 22%), and dot-line pattern was the least (11 cases, 8%).Comparing the blood flow distribution pattern of CEUS and Smart 3D SMI, CEUS showed that all cases had contrast agent perfusion. Of the patients with dot-line pattern of Smart 3D SMI, 55% showed rough and rich blood flow, while in branched pattern, spherical pattern, expansion pattern and mixed pattern, 75%, 72%, 75% and 52% showed rough and rich blood flow, respectively. Of the various Smart 3D SMI pattern, CEUS showed that in 82%, 75%, 67%, 60% and 58% of dot-line pattern, branched pattern, spherical pattern, expansion pattern and mixed pattern, respectively, the branched blood flow could not been clearly delineated. The perfusion intensity of branched and spherical pattern was higher than that of dot-line pattern, and the difference was significant(P=0.018, 0.006). The perfusion area was mostly located in interior of the mass in dot-line pattern, branched pattern and spherical pattern, and only a few cases had radial perfusion at periphery of the mass. Conclusions: Smart 3D SMI technology is commonly used in the evaluation of ultrasound micro-blood flow distribution of breast mass, spherical pattern and branched pattern are the frequently seen pattern. The three-dimensional morphological characteristics can afford more information regarding the blood flow distribution pattern of breast cancer.