Study On Terahertz Spectroscopy And Refection Imaging Technology In Brain Surgery Disease Detection

The terahertz wave is in the transitional zone between electronics and photonics.It has many special properties,such as sensitivity to water,fingerprint characteristics of biomacromolecules and non-ionization.Therefore,it has obvious advantages in biomedical detection.Currently,terahertz time-domain spectroscopy and imaging techniques are widely implemented in the detection of biological tissue lesions.The incidence and mortality of brain diseases such as cerebral ischemia and glioma are extremely high.However,the prior art methods face many problems in the early diagnosis and detection of these diseases.Therefore,the terahertz time-domain spectroscopy and imaging techniques are applied in this paper to conduct in-depth research on the diagnosis and boundary extraction of brain diseases represented by cerebral ischemia and glioma.The main contents and innovations are as follows:1.Transmissive terahertz time-domain spectroscopy was used to detect both fresh and paraffin-embedded cerebral ischemic tissue at different ischemic times.The detection of early cerebral ischemic tissue with ischemia for two hours was achieved.Biosensors based on artificial metamaterial structure were designed and fabricated.The transmittance of the biosensors with different concentrations of PC12 nerve cells implanted for terahertz wave is studied.High sensitivity detection of PC12 nerve cell concentration was achieved by detecting the drift of the sensor resonance peak and the change of the terahertz wave transmittance at the resonance peak.The detection sensitivities were 10 GHz/(10⁴·m L^-1)and 0.49 d B/(10⁴·m L^-1),respectively.2.Considering that the continuous-wave terahertz images normally have poor signal-to-noise ratio and contrast,a hybrid method consisting of the block-matching 3D denoising,fuzzy c-means clustering,morphology operations and‘Canny'edge detection was proposed.The letters were accurately segmented in the continuous-wave terahertz image of a pattern plate with accuracy,sensitivity and specificity of 94.3%,91.6%and 95.5%,respectively.Furthermore,the hybrid method was applied to the terahertz image of ex vivo glioma,and the accurate extraction of tumor regions was achieved.The accuracy,sensitivity and specificity of the algorithm were 95.6%,84.5%and 97.7%,respectively.3.The terahertz-Raman spectroscopy was applied to the detection of glioma.Distinct differences between the glioma tissue and healthy brain tissue could be observed at the intensity,number and Raman frequency of the Raman peak in the terahertz-Raman spectrums.Which might be caused by changes in the structure and content of biomacromolecules during tissue carcinogenesis.