Joro R, Lääperi A-L, Dastidar P, Järvenpää R, Kuukasjärvi T, Toivonen T, Saaristo R, Soimakallio S
Institute of Measurement and Information Technology, Tampere University of Technology, Tampere, Finland.
Acta Radiol. 2009 Oct;50(8):860-9. doi: 10.1080/02841850903095377.
Dynamic infrared (IR) imaging is an emerging functional imaging modality for the detection of breast cancer without evidence of optimal imaging and diagnostic application.
To evaluate dynamic IR imaging in breast cancer diagnostics by comparing a stepwise diagnostic scheme to digital mammography and postoperative histopathology.
Dynamic IR imaging of breasts was undertaken preoperatively with a long-wave quantum well (QWIP) and two mid-wave photovoltaic (PV) IR cameras in 10 cases (age 34-80 years) with breast cancer size 6-45 mm on mammography. Image stabilization, two-phase frequency analysis, and two image-processing algorithms were applied.
Combining image processing with frequency analysis proved advantageous in detecting breast cancer. The IR imaging process recognized the cancer area independently of tissue density, cancer size, and cancer appearance on mammography. Compared to histopathology, all cancers yielded abnormal analysis results, including one case of ductal carcinoma in situ. Evidence of lymphatic invasion in postoperative histopathology, imaging with PV camera, and image processing with the Wiener filtering combination correlated with highest confidence between normal and cancer tissue measured by the calculated superiority value.
Dynamic IR imaging with image-processing-guided frequency analysis is a promising modality for breast cancer detection and may not have the tissue-dependent limitations of mammography. Our results encourage further work on medical IR imaging and comparison to established breast-imaging modalities.
动态红外（IR）成像是一种新兴的功能成像方式，用于在缺乏最佳成像和诊断应用证据的情况下检测乳腺癌。
通过将逐步诊断方案与数字乳腺摄影及术后组织病理学进行比较，评估动态IR成像在乳腺癌诊断中的作用。
对10例（年龄34 - 80岁）乳腺摄影显示乳腺癌大小为6 - 45 mm的患者，术前使用长波量子阱（QWIP）和两台中波光伏（PV）红外相机进行乳房动态IR成像。应用了图像稳定、两相频率分析和两种图像处理算法。
将图像处理与频率分析相结合在检测乳腺癌方面被证明是有利的。IR成像过程独立于乳腺摄影上的组织密度、癌症大小和癌症外观识别出癌区。与组织病理学相比，所有癌症均产生异常分析结果，包括1例原位导管癌。术后组织病理学中的淋巴浸润证据、使用PV相机成像以及采用维纳滤波组合的图像处理与通过计算优势值测量的正常组织与癌组织之间的最高置信度相关。
具有图像处理引导频率分析的动态IR成像是一种有前景的乳腺癌检测方式，可能不存在乳腺摄影的组织依赖性局限性。我们的结果鼓励在医学IR成像方面开展进一步工作，并与已确立的乳腺成像方式进行比较。
