基于新型收发一体式探头的裸鼠乳腺癌组织电磁特性数据快速检测方法研究

Rapid detection method for electromagnetic properties of breast cancer tissues in nude mice based on novel transmit-receive integrated probe

  • 摘要: 背景 乳腺癌手术中快速准确辨识切除组织至关重要。基于S参数的生物组织区分方法具有应用潜力,但经典开端同轴探头法需多次校准才能利用S11参数进行辨识,限制了其临床应用。新型收发一体式探头方法无需复杂校准,且可同步获取S11和S21参数,为生物组织快速检测提供了新途径。目的 验证新型收发一体式探头在生物组织快速辨识中的应用潜力。方法 以荷瘤裸鼠为对象,分别使用开端同轴探头和新型收发一体式探头测量皮肤、脂肪、肌肉及肿瘤组织的S 参数,通过幅度差值分析、多频点(5、10、15、20 GHz)统计检验及机器学习算法,评估组织区分能力。进一步利用手术离体标本(包括肿瘤、腺体、脂肪、皮肤及癌旁组织)验证该方法的临床适用性。结果 开端同轴探头对荷瘤裸鼠的测量结果显示,肿瘤与脂肪组织在5 GHz处S11参数无显著差异,与肌肉组织在10、15、20 GHz处亦无差异;而新型收发一体式探头在四个频点下均检测到肿瘤与三类正常组织的S11及S21参数存在显著差异(P<0.001)。机器学习对荷瘤裸鼠及手术离体标本的分析结果表明,联合使用S11和S21参数可提升分类效能,其中S21参数显著增强了对正常或癌旁组织的辨识能力。结论 本研究通过荷瘤裸鼠和手术离体标本实验,系统验证了新型收发一体式探头在生物组织快速区分中的有效性,初步证明其具备临床适用性,为乳腺癌术中实时组织辨识提供了新技术支持。

     

    Abstract: Background Rapid and accurate identification of resected tissues is crucial during breast cancer surgery. The Sparameter- based method for distinguishing biological tissues shows potential, but the classical open-ended coaxial probe method requires multiple calibrations to utilize the S11 parameter for identification, limiting its clinical application. The novel transmitreceive integrated probe method eliminates the need for complex calibration and can simultaneously acquire both S11 and S21 parameters, offering a new approach for rapid biological tissue detection.Objective To validate the potential application of the novel transmit-receive integrated probe in the rapid identification of biological tissues. Methods Using tumor-bearing mice as subjects, S-parameters of skin, fat, muscle, and tumor tissues were measured using both the open-ended coaxial probe and the novel transmit-receive integrated probe. The tissue discrimination capability was evaluated through magnitude difference analysis, multifrequency point (5, 10, 15, and 20 GHz) statistical testing, and machine learning algorithms. Furthermore, the clinical applicability of this method was verified using surgically resected specimens (including tumor, gland, fat, skin, and adjacent normal tissues). Results Measurements on tumor-bearing mice using the open-ended coaxial probe showed no significant difference in S11 parameters between tumor and fat tissues at 5 GHz, nor between tumor and muscle tissues at 10, 15, and 20 GHz. In contrast, the novel transmit-receive integrated probe detected significant differences (P<0.001) in both S11 and S21 parameters between tumor and the three types of normal tissues at all four frequency points. Machine learning analysis results from both tumor-bearing mice and surgically resected specimens indicated that combining S11 and S21 parameters improved classification performance, with the S21 parameter significantly enhancing the ability to identify normal or adjacent normal tissues. Conclusion This study systematically validates the effectiveness of the novel transmit-receive integrated probe for rapid discrimination of biological tissues through experiments on tumor-bearing mice and surgically resected specimens, preliminarily demonstrating its clinical applicability and providing new technical support for real-time tissue identification during breast cancer surgery.

     

/

返回文章
返回