CAJ | 학술논문

目的：针对近年来市场上抗生素鸡肉事件的频繁发生，提出一种可以快速、准确地检测出鸡肉中四环素残留量的模型，并对其进行健康风险评估。方法首先，利用近红外光谱仪采集肉鸡的光谱信息，通过多元散射校正法进行预处理。接着，使用区间偏最小二乘回归分析方法来建立预测模型，对四环素进行定量分析。然后，根据人体摄入四环素残留量数据计算出膳食的暴露量，从而引入食品安全指数指标，对鸡肉进行健康风险评估。最后，利用SOM人工神经网络对其实用性加以验证。结果风险评估数学模型结合国家标准将鸡肉带来的风险分为：高、中、低3类，等级划分达到95.5%的正确率。结论本实验研究方法可以应用到其他相关的农产品检测和食品检测中，以保障食品安全。
목적：침대근년래시장상항생소계육사건적빈번발생，제출일충가이쾌속、준학지검측출계육중사배소잔류량적모형，병대기진행건강풍험평고。방법수선，이용근홍외광보의채집육계적광보신식，통과다원산사교정법진행예처리。접착，사용구간편최소이승회귀분석방법래건립예측모형，대사배소진행정량분석。연후，근거인체섭입사배소잔류량수거계산출선식적폭로량，종이인입식품안전지수지표，대계육진행건강풍험평고。최후，이용SOM인공신경망락대기실용성가이험증。결과풍험평고수학모형결합국가표준장계육대래적풍험분위：고、중、저3류，등급화분체도95.5%적정학솔。결론본실험연구방법가이응용도기타상관적농산품검측화식품검측중，이보장식품안전。
Objective To put forward a model with detection of tetracycline residues in chicken and to carry on the health risk assessment for antibiotic chicken events frequently occured in the market in recent years. Methods Based on the multiple scattering correction method for pretreatment, chicken spectral information was collected by near infrared spectrometer. The forecast model was built by interval partial least squares regression method to analyze the tetracycline quantitatively. Then, according to the data of intake by human body dietary with tetracycline residues, the dietary exposure was calculated so as to introduce food safety index indicators to assess the health risk from the chicken. Finally, the practicality of that was verified by using SOM neural network. Results Based on risk assessment model combined with the national standard, the chicken risks divided into high, medium and low levels, the accuracy rate of hierarchy could reach to 95.5%. Conclusion This method can be applied to test other related agricultural products and food industry to guarantee food security.