荧光 PCR 多规格耗材自动对焦算法及装置研究

Objective Real^-time fluorescent quantitative polymerase chain reaction (PCR) is the gold standard for nucleic acid detection of infectious diseases and plays an irreplaceable role in many fields, including disease prevention and control, clinical diagnosis, food safety, and inspection and quarantine. However, PCR detection encounters significant problems. Changes in the specifications of consumables affect the collection of fluorescent signals. Traditional manual focusing has many drawbacks, including cumbersome operation, frequent calibration, low efficiency, increased labor intensity, and poor image acquisition accuracy. When the consumables change slightly, it is difficult to adapt to multiple consumable specifications. Although there have been research achievements in autofocus technology both at home and abroad, the research conducted on autofocus imaging of fluorescent PCR with multiple specifications of consumables is still limited. Methods In this study, an optical detection system was constructed for the acquisition of 96^-flux fluorescent PCR images. Firstly, the fluorescence imaging situations were observed under different cycle numbers. The performances of evaluation methods, such as the Tenengrad gradient, Laplacian gradient, image variance, Laplacian inhomogeneity, and Sobel operator, were compared during the process of coarse focusing of fluorescent PCR images to determine the appropriate definition evaluation index. Based on the autofocus optical system and the selected definition evaluation function, the image data in the process of the autofocus algorithm were collected, and coarse, fine, and horizontal calibration were performed. After the focusing was completed, a series of image processing operations were performed on the original image in turn, including binarization processing, detection of circular targets by the Hough circle transform, constructing a mask, and calculating the fluorescence intensity by extracting the gray values of the pixels in the fluorescent region. Finally, FAM fluorophore solutions and DNA samples with different gradient concentrations were prepared, and fluorescent quantitative PCR detection experiments were conducted using consumables of different specifications (0.2 mL and 0.1 mL polypropylene PCR8 strip tubes). The Ct values of the two types of consumables were compared. Results and Discussions The Sobel operator algorithm demonstrates optimal performance during the coarse calibration process. This algorithm accurately identifies the position near the ideal working distance, and its calculation results yield obvious level and gradient changes in the entire focal length range for the sharpness interpretation, which is identified as the most suitable sharpness evaluation index. After coarse calibration, fine calibration, horizontal calibration, and image processing, the image meets the requirements for automatic extraction of fluorescent signals; accordingly, the accuracy of image processing and fluorescent detection was improved considerably. When reading the same fluorescein with 0.1 mL and 0.2 mL PCR tubes, it is found that there is no significant difference in the average fluorescence intensity. The quantification of HBV DNA shows that there are minor Ct value differences regarding the tested nucleic acid concentrations, ranging from 0 to 0.32. It is shown that the autofocus fluorescence detection device and algorithm developed can satisfy the needs of fluorescence signal detection at different consumable specifications. Conclusions The autofocus algorithm proposed in this study can replace manual focusing, meets the imaging requirements of multispecification consumables, greatly improves the focusing accuracy, and enhances the image’s resolution/contrast. This performance guarantees the accurate acquisition of fluorescence signals, reduces the detection error, and improves the reliability of the results. Accordingly, this provides an efficient and reliable solution for fluorescence PCR imaging, which will be conducive to the future development of fluorescence PCR technology.