Background/Objectives: Breast cancer is a significant global health concern, and early detection is crucial for improving patient outcomes. Mammography is widely used but has limitations, particularly for younger women with denser breasts. These include reduced sensitivity, false positives, and radiation risks. This highlights the need for alternative screening methods. In this study, we assess the performance of SAFE (Scan and Find Early), a novel microwave imaging device, in detecting breast cancer in a larger patient cohort. Unlike previous studies that predominantly relied on cross-validation, this study employs a more reliable, independent evaluation methodology to enhance generalizability. Methods: We developed an XGBoost model to classify breast cancer cases into positive (malignant) and negative (benign or healthy) groups. The model was analyzed with respect to key factors such as breast size, density, age, tumor size, and histopathological findings. This approach provides a better understanding of how these factors influence the model’s performance, using an independent evaluation methodology for increased reliability. Results: Our results demonstrate that SAFE exhibits high sensitivity, particularly in dense breasts (91%) and younger patients (83%), suggesting its potential as a supplemental screening tool. Additionally, the system shows high detection accuracy for both small (<2 cm) and larger lesions, proving effective in early cancer detection. Conclusions: This study reinforces the potential of SAFE to complement existing screening methods, particularly for patients with dense breasts, where mammography’s sensitivity is reduced. The promising results warrant further research to solidify SAFE’s clinical application as an alternative screening tool for breast cancer detection.
**背景/目的：** 乳腺癌是全球性的重大健康问题，早期检测对于改善患者预后至关重要。乳腺X线摄影（钼靶）虽广泛应用，但存在局限性，尤其对于乳腺致密的年轻女性，其敏感性降低、假阳性率高且存在辐射风险。这凸显了对替代筛查方法的需求。本研究旨在评估一种新型微波成像设备SAFE（Scan and Find Early）在更大患者队列中检测乳腺癌的性能。与以往主要依赖交叉验证的研究不同，本研究采用了更可靠、独立的评估方法以增强结果的普适性。 **方法：** 我们开发了一个XGBoost模型，用于将乳腺癌病例分类为阳性（恶性）和阴性（良性或健康）组。该模型针对乳房大小、密度、年龄、肿瘤大小和组织病理学结果等关键因素进行了分析。通过采用独立的评估方法，本研究旨在更可靠地理解这些因素如何影响模型的性能。 **结果：** 我们的结果表明，SAFE表现出高敏感性，尤其在致密乳腺（91%）和年轻患者（83%）中，提示其作为补充筛查工具的潜力。此外，该系统对于小病灶（<2 cm）和较大病灶均显示出较高的检测准确性，证明其在早期癌症检测中的有效性。 **结论：** 本研究进一步证实了SAFE作为现有筛查方法补充的潜力，尤其适用于乳腺致密、钼靶敏感性降低的患者。这些积极的结果值得进一步研究，以巩固SAFE作为乳腺癌检测替代筛查工具的临床应用。
肿瘤(癌症)患者之家