The rapid and sensitive detection of specific protein biomarkers is essential for early diagnosis and monitoring of diseases. Magnetic digital microfluidics (MDMF), capable of integrating and miniaturizing chemiluminescence immunoassay (CLIA), offers significant advantages for biomarker detection, including high flexibility, high efficiency, and low sample consumption. Magnetic bead manipulation, as a key component of MDMF, plays a critical role in improving detection sensitivity and accuracy within a short analysis time. In this study, we proposed an automated MDMF-based CLIA system and developed a compact analyzer, enabling faster and more sensitive biomarker detection through optimized magnetic bead manipulation. The system employs a film-sandwiched microfluidic chip, along with a specially designed spliced magnet to generate a stronger and more concentrated magnetic force, significantly enhancing the control over the motion of magnetic beads. By using the magnet to drive magnetic beads through water-in-oil droplets, the system automates the entire CLIA workflow. Through theoretical analysis, simulations, and experiments, we validate the system's advantages in magnetic bead aggregation and transfer, achieving a high magnetic bead collection rate of 97.07% with excellent cleaning effect after three cleaning rounds. The detection of cardiac troponin I (cTnI) can be completed within 10 min, with a limit of detection (LOD) of 2.09 pg mL−1, a detection range of 0.01 to 20 ng mL−1, and a strong linear correlation with conventional CLIA (R2 = 0.9961). This system achieves rapid, sensitive, and automated CLIA, making it a promising platform for clinical diagnostics.
Robotic cochlear implantation is an effective way to restore the hearing of hearing-impaired patients, and facial nerve recognition is the key to the operation. However, accurate facial nerve segmentation is a challenging task, mainly for two key issues: (1) the facial nerve area is very small in image, and there are many similar areas; (2) low contrast of the border between the facial nerve and the surrounding tissues increases the difficulty. In this work, we propose an end-to-end neural network, called FNSegNet, with two stages to solve these problems. Specifically, in the coarse segmentation stage, we first adopt three search identification modules to capture small objects by expanding the receptive field from high-level features and combine an effective pyramid fusion module to fuse. In the refine segmentation stage, we use a decoupling optimization module to establish the relationship between the central region and the boundary details of facial nerve by decoupling the boundary and center area. Meanwhile, we feed them into a spatial attention module to correct the conflict regions. Extensive experiments on the challenging dataset demonstrate that the proposed FNSegNet significantly improves the segmentation accuracy (0.858 on Dice, 0.363 mm on 95% Hausdorff distance), and reduces the computational complexity (13.33G on FLOPs, 9.86M parameters).
The magnetic separation and cleaning module, as a core component of the fully automated chemiluminescence immunoassay (CLIA) analyzer, encounters issues including high magnetic bead loss rate, long cleaning time, and poor cleaning effect. Based on a simulation analysis using COMSOL, we proposed a novel magnetic separation and cleaning module applied to a fully automated CLIA analyzer. The module adopted a method of arranging spliced rectangular magnets on opposite sides, where the same polarity faced each other, as well as a three-stage magnetic bead collection method. With the proposed method, the total cleaning process can be accomplished within 225 s; the total magnetic bead loss rate over three rounds of cleaning is 6.03%, whereas that of traditional instruments is 25.85%; the coefficient of variation (CV) of the magnetic bead loss rate is less than 0.5%; effective cleaning of free markers is achieved under various sample conditions. Compared with traditional CLIA instruments, this method comprehensively improves key performance indicators of the magnetic separation and cleaning module, providing a reference for similar modules in fully automated CLIA analyzers and positively impacting the accuracy of CLIA for the detection of disease biomarkers.
Driven by the rising demand for point-of-care testing (POCT) in aging societies, accurate temperature regulation of reaction solutions has become a core technical bottleneck for miniaturized chemiluminescence immunoassay systems, since conventional indirect control strategies inevitably produce systematic deviations. To tackle this challenge, we present an integrated solution that couples multiphysics simulation, data-driven temperature estimation modeling, and embedded hardware design. We constructed a COMSOL heat transfer model to analyze the thermal performance of the microfluidic chip. Meanwhile, a grey wolf optimization (GWO) enhanced long short-term memory (LSTM) network was developed to infer the unmeasured actual reaction solution temperature based on accessible parameters, including heating voltage, ambient temperature and substrate temperature. The obtained temperature estimation was then fed back to a fuzzy PID controller for closed-loop regulation. Experimental results demonstrated that the GWO-LSTM model limited the estimation error within 0.3 °C, and the steady-state temperature control accuracy reached ±0.2 °C or higher under fluctuating ambient conditions and diverse initial states. For cardiac troponin I (cTnI) detection, the proposed system shortened the incubation duration and reduced the coefficient of variation from 10.77% to 2.69%. This work addresses the key bottleneck restricting precise temperature control in microfluidic chemiluminescence analyzers, which provides robust technical support for the development of next-generation high-performance POCT instruments.