Spectral Analysis in Ex-vivo between Tumor and Normal Liver Tissues

• Main Authors: Ya-Yu Lin, 林亞妤
• Other Authors: Yin Chang
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/11419098225887047153

碩士 === 國立陽明大學 === 生物醫學工程學系 === 105 === Most cases of liver cancer are hepatocellular carcinoma (HCC) and can be multiple deposited in the liver. The primary diagnostic and imaging methods of locating liver cancer include ultrasound, computed tomography (CT), magnetic resonance imaging (MRI) and angiography. However, most of these methods are not available or can only provide 2D images of the target during surgical resection, which is the optimal treatment for resectable liver tumors. Since the malignant tumor typically has irregular shapes and margins, we assume it would be beneficial for enhancing the efficiency and accuracy by using a real-time optical biopsy to help determine the area of liver tumors during surgery. Therefore, we generated an idea of collecting and analyzing optical spectra from cancer and healthy liver tissues with an optical probe. Theoretically, the optical properties could be different between cancer tissues and normal tissues due to the dissimilarity in morphological and functional characteristics. This study is an IRB collaboration with Taipei Veterans General Hospital to acquire human liver cancer tissues from surgical resection for data collection. Diffuse reflectance spectroscopy (DRS) were employed to characterize spectral signals for these tissue samples through a probe of optic fiber embedded needle and covered a broadband of wavelength from 200 nm to 1100 nm with a single light source. Spectra of tumorous and normal liver tissues from 15 patients were acquired and analyzed. Both normalized spectrum and spectrum to system background ratio data showed significant differences between cancer and normal tissue, especially at around 515 nm and 615 nm, and 755 nm respectively. The spectral intensities at these specific wavelengths can be utilized as classifiers to determine whether the contacting tissue is cancerous. If the detecting accuracy has improved and appropriate sensitivity and specificity have chosen in the future, this system can be helpful for liver cancer diagnosis or treatment due to its minimally invasive and real-time distinguishing characteristics.