Abstract: [Objective] This study aimed to optimize the pretreatment steps of the microbiological rapid detection method for coliforms in frozen desserts using the Soleris microbiological rapid detection system,with the results of the traditional national standard plate counting method for coliforms as the calibration dataset,and to establish a regression calibration curve for coliforms,providing efficient technical support for microbiological detection of frozen desserts.[Method] First,the pretreatment processes for dairy-containing and non-dairy frozen desserts were optimized and unified: 25 g of melted and mixed samples were placed in 225 mL of sterile diluent,and the pH was adjusted to 6.7±0.2 if necessary.Then,the Soleris microbiological rapid detection system and the traditional national standard plate counting method （GB 4789.3—2016） were used to simultaneously detect coliforms in frozen desserts,and the applicability of the Soleris system was verified by comparing the detection results of the two methods.[Result] The detection rate of the Soleris microbiological rapid detection system for 30 samples with different pollution levels reached 100%,with no false positive or false negative results. The regression calibration curve of the detection results was y=-0.405 3x+6.061 9 （R²=0.869 6）,indicating a significant negative correlation between the Soleris detection time and the number of coliforms. Repeatability verification showed that the maximum intra-group relative standard deviation was only 3.4%,indicating good system stability.[Conclusion] On the basis of maintaining the characteristic of "negative correlation between positive warning time and the number of target microorganisms",the optimized Soleris microbiological rapid detection system significantly shortens the detection cycle （warning time 7.2~11.9 h）,effectively reduces the detection workload and improves the detection efficiency,which can meet the needs of rapid quality control in the production process of frozen desserts.

Key words: frozen desserts, coliforms, Soleris microbiological rapid detection system, pretreatment optimization, regression calibration curve