A deep learning based method for automatic analysis of high-throughput droplet digital PCR images

Automatic detection of fluorescent droplets for droplet digital PCR: a device capable of processing multiple microscope images

Droplet digital PCR (ddPCR) is recognized as a high-precision method for nucleic acid quantification, extensively utilized in biomedical research and clinical diagnostics. This technique employs microfluidic technology to partition the nucleic acid-containing reaction mixture into discrete droplets for amplification, achieving absolute quantification by identifying and enumerating the number of fluorescent droplets. The accuracy of droplet quantification is pivotal to the success of the assay. However, current image-processing tools are operationally complex, and commercial instruments are costly. Moreover, the designed algorithms exhibit a need for enhanced accuracy and are often restricted to use by trained personnel with specific microscopy equipment. In response to these challenges, we introduce an automated device (A-MMD), capable of detecting fluorescent droplets in ddPCR images captured by multiple microscopes. The device integrates three distinct algorithms tailored for the image processing of Laser Scanning Confocal Microscopy (LSCM), inverted microscopy, and self-assembled microscopy. Experimental validation using λ DNA demonstrated a 100.00% identification rate for positive droplets across all three image types, and the average identification rates for total droplets being 99.27% for LSCM, 98.96% for inverted microscopy, and 99.08% for self-assembled microscopy. Furthermore, the A-MMD is equipped with a user-friendly interface (UI) that streamlines the operational process, enabling non-specialists to efficiently perform droplet detection tasks. Our device not only has good environmental adaptability and identification accuracy, but also significantly reduces costs and operational complexity. It offers an economical, efficient, and user-friendly solution for ddPCR image analysis, thereby further propelling the advancement and application of nucleic acid detection technology.

Deep-qGFP: A Generalist Deep Learning Assisted Pipeline for Accurate Quantification of Green Fluorescent Protein Labeled Biological Samples in Microreactors

Absolute quantification of biological samples provides precise numerical expression levels, enhancing accuracy, and performance for rare templates. Current methodologies, however, face challenges-flow cytometers are costly and complex, whereas fluorescence imaging, relying on software or manual counting, is time-consuming and error-prone. It is presented that Deep-qGFP, a deep learning-aided pipeline for the automated detection and classification of green fluorescent protein (GFP) labeled microreactors, enables real-time absolute quantification. This approach achieves an accuracy of 96.23% and accurately measures the sizes and occupancy status of microreactors using standard laboratory fluorescence microscopes, providing precise template concentrations. Deep-qGFP demonstrates remarkable speed, quantifying over 2000 microreactors across ten images in just 2.5 seconds, with a dynamic range of 56.52-1569.43 copies µL-1 . The method demonstrates impressive generalization capabilities, successfully applied to various GFP-labeling scenarios, including droplet-based, microwell-based, and agarose-based applications. Notably, Deep-qGFP is the first all-in-one image analysis algorithm successfully implemented in droplet digital polymerase chain reaction (PCR), microwell digital PCR, droplet single-cell sequencing, agarose digital PCR, and bacterial quantification, without requiring transfer learning, modifications, or retraining. This makes Deep-qGFP readily applicable in biomedical laboratories and holds potential for broader clinical applications.

A ddPCR platform based on a microfluidic chip with a dual-function flow-focusing structure for sample-to-result DNA quantification analysis

Droplet digital PCR (ddPCR) is a powerful method for absolute nucleic acid quantification with high precision and accuracy. However, complicated operational steps have hampered the use and diffusion of ddPCR. Therefore, an automated, easy-to-use, low-sample-consumption, and portable ddPCR platform is urgently needed. This paper proposes a microfluidic ddPCR platform based on a microfluidic chip that can realize the sample-to-result function by switching the rotary valve, achieving the dual function of the flow-focusing structure for droplet generation and readout. Sample, generation oil, and analysis oil were pre-added to the reservoirs. Droplets were generated due to focusing flow, and after passing through the integrated temporary storage bin in the rotary valve, the droplets and oil subsequently entered the collecting tube, improving the droplet-to-oil volume ratio for enhanced thermal cycle performance. Droplets with an average diameter of 107.44 μm and a CV of 2.38% were generated using our chip under the optimal pressures. High-performance thermal cycling was achieved through improvements of the droplet-to-oil volume ratio of the sample, the integrated heating lid, the pure copper heating base, and the temperature-controlling algorithm. Gradient quantification experiments were conducted for the HER2 and CEP17 genes extracted from breast cancer cells, yielding strong linear correlations with R2 values of 0.9996 for FAM and 0.9989 for CY5. Moreover, pronounced linearity was obtained between the detected concentrations of HER2 and CEP17, indicated by a slope of 1.0091 and an R2 of 0.9997, signifying consistent HER2 : CEP17 ratios across various sample dilutions. The outcomes of the quantitative analysis, encompassing the dynamic range and the consistency of the HER2 : CEP17 ratio using our ddPCR platform, meet the standards required for breast cancer assessment and therapy. Our ddPCR platform is automated, portable, and capable of stable droplet generation, high-efficiency amplification, realization of the sample-to-result function based on dual-function flow-focusing structure, and accuracy absolute quantification, underscoring its significant potential for ddPCR analysis in clinical diagnostics.