| Somatic cell count of the fresh milk is an important index to measure the milk yield and quality, and it is the important standard to diagnosis of dairy cow mastitis. Therefore, the research content is the testing method of SCC in milk, which is characteristics of accurate, fast, low cost, online detection.This topic uses HL-60 cells to instead the milk somatic cells to detect object, with microfluidic electrode for detecting sensor, according to the capacitance effect of intact cells in the electric field to improve the sensitivity of cell content detection by microfluidic sensor structure design. The research content is the test response characteristics under different test conditions using electrochemical impedance spectroscopy test technology, and studies on the important role in improving the prediction accuracy of cell content based on the main test parameters model method.The mainly research content and research results are divided into the following parts. First of all, the technical scheme of cell quantitative detection by using microfluidic electrochemical impedance sensor to realize. The microfluidic sensor of two electrode is designed with the microprocessing technology to fabricate the gold electrode of diameter 30μm and 60μm on glass substrate, with 100μm channel width and depth PDMS to proceed plasma bonding to package to get the microfluidic sensor, and with the 10.88ms/cm KCl solution to calibration the electrode constant under normal temperature. Secondly, the micro-porous filter membrane method is used to remove to the original cell in fresh milk samples and HL-60 cell is configured to milk samples of different gradient, and the impedance response information of sensor is obtained by using electrochemical impedance spectroscopy test technology, and the impedance spectroscopy of two gradient milk sample are obvious differences in the condition of 500 mv, 70~530Hz and 30℃. Third, the effect of temperature and different excitation are researched in the detection of somatic cell count, and the impedance spectroscopy of 30℃ is compared with 10℃ more smooth and stable, the test results in the condition of 500 mv and 70~530Hz is better than the test results in the condition of other frequencies, 100 mV and 300 mV. Finally, BP neural network model is established with the resistance and reactance as input, and the model can primely predict the milk somatic cell content. The investigative results provide an effective method for rapid detection of somatic cell content in milk. |