Testing the generalizability and effectiveness of deep learning models among clinics: sperm detection as a pilot study.

BACKGROUND: Deep learning has been increasingly investigated for assisting clinical in vitro fertilization (IVF). The first technical step in many tasks is to visually detect and locate sperm, oocytes, and embryos in images. For clinical deployment of such deep learning models, different clinics use different image acquisition hardware and different sample preprocessing protocols, raising the concern over whether the reported accuracy of a deep learning model by one clinic could be reproduced in another clinic. Here we aim to investigate the effect of each imaging factor on the generalizability of object detection models, using sperm analysis as a pilot example. METHODS: Ablation studies were performed using state-of-the-art models for detecting human sperm to quantitatively assess how model precision (false-positive detection) and recall (missed detection) were affected by imaging magnification, imaging mode, and sample preprocessing protocols. The results led to the hypothesis that the richness of image acquisition conditions in a training dataset deterministically affects model generalizability. The hypothesis was tested by first enriching the training dataset with a wide range of imaging conditions, then validated through internal blind tests on new samples and external multi-center clinical validations. RESULTS: Ablation experiments revealed that removing subsets of data from the training dataset significantly reduced model precision. Removing raw sample images from the training dataset caused the largest drop in model precision, whereas removing 20x images caused the largest drop in model recall. by incorporating different imaging and sample preprocessing conditions into a rich training dataset, the model achieved an intraclass correlation coefficient (ICC) of 0.97 (95% CI: 0.94-0.99) for precision, and an ICC of 0.97 (95% CI: 0.93-0.99) for recall. Multi-center clinical validation showed no significant differences in model precision or recall across different clinics and applications. CONCLUSIONS: The results validated the hypothesis that the richness of data in the training dataset is a key factor impacting model generalizability. These findings highlight the importance of diversity in a training dataset for model evaluation and suggest that future deep learning models in andrology and reproductive medicine should incorporate comprehensive feature sets for enhanced generalizability across clinics.

CP-Net: Instance-aware part segmentation network for biological cell parsing.

Instance segmentation of biological cells is important in medical image analysis for identifying and segmenting individual cells, and quantitative measurement of subcellular structures requires further cell-level subcellular part segmentation. Subcellular structure measurements are critical for cell phenotyping and quality analysis. For these purposes, instance-aware part segmentation network is first introduced to distinguish individual cells and segment subcellular structures for each detected cell. This approach is demonstrated on human sperm cells since the World Health Organization has established quantitative standards for sperm quality assessment. Specifically, a novel Cell Parsing Net (CP-Net) is proposed for accurate instance-level cell parsing. An attention-based feature fusion module is designed to alleviate contour misalignments for cells with an irregular shape by using instance masks as spatial cues instead of as strict constraints to differentiate various instances. A coarse-to-fine segmentation module is developed to effectively segment tiny subcellular structures within a cell through hierarchical segmentation from whole to part instead of directly segmenting each cell part. Moreover, a sperm parsing dataset is built including 320 annotated sperm images with five semantic subcellular part labels. Extensive experiments on the collected dataset demonstrate that the proposed CP-Net outperforms state-of-the-art instance-aware part segmentation networks.

Modeling and Compensation of Positioning Error in Micromanipulation.

In order to improve the positioning accuracy of the micromanipulation system, a comprehensive error model is first established to take into account the microscope nonlinear imaging distortion, camera installation error, and the mechanical displacement error of the motorized stage. A novel error compensation method is then proposed with distortion compensation coefficients obtained by the Levenberg-Marquardt optimization algorithm combined with the deduced nonlinear imaging model. The compensation coefficients for camera installation error and mechanical displacement error are derived from the rigid-body translation technique and image stitching algorithm. To validate the error compensation model, single shot and cumulative error tests were designed. The experimental results show that after the error compensation, the displacement errors were controlled within 0.25 µm when moving in a single direction and within 0.02 µm per 1000 µm when moving in multiple directions.

A Nitrocellulose Paper-Based Multi-Well Plate for Point-of-Care ELISA.

Low-cost diagnostic tools for point-of-care immunoassays, such as the paper-based enzyme-linked immunoassay (ELISA), have become increasingly important, especially so in the recent COVID-19 pandemic. ELISA is the gold-standard antibody/antigen sensing method. This paper reports an easy-to-fabricate nitrocellulose (NC) paper plate, coupled with a desktop scanner for ELISA, which provides a higher protein immobilization efficiency than the conventional cellulose paper-based ELISA platforms. The experiments were performed using spiked samples for the direct ELISA of rabbit IgG with a limit of detection (LOD) of 1.016 µg/mL, in a measurement range of 10 ng/mL to 1 mg/mL, and for the sandwich ELISA of sperm protein (SP-10) with an LOD of 88.8 ng/mL, in a measurement range of 1 ng/mL to 100 µg/mL. The described fabrication method, based on laser-cutting, is a highly flexible one-step laser micromachining process, which enables the rapid production of low-cost NC paper-based multi-well plates with different sizes for the ELISA measurements.

Microengineered platforms for characterizing the contractile function of in vitro cardiac models.

Emerging heart-on-a-chip platforms are promising approaches to establish cardiac cell/tissue models in vitro for research on cardiac physiology, disease modeling and drug cardiotoxicity as well as for therapeutic discovery. Challenges still exist in obtaining the complete capability of in situ sensing to fully evaluate the complex functional properties of cardiac cell/tissue models. Changes to contractile strength (contractility) and beating regularity (rhythm) are particularly important to generate accurate, predictive models. Developing new platforms and technologies to assess the contractile functions of in vitro cardiac models is essential to provide information on cell/tissue physiologies, drug-induced inotropic responses, and the mechanisms of cardiac diseases. In this review, we discuss recent advances in biosensing platforms for the measurement of contractile functions of in vitro cardiac models, including single cardiomyocytes, 2D monolayers of cardiomyocytes, and 3D cardiac tissues. The characteristics and performance of current platforms are reviewed in terms of sensing principles, measured parameters, performance, cell sources, cell/tissue model configurations, advantages, and limitations. In addition, we highlight applications of these platforms and relevant discoveries in fundamental investigations, drug testing, and disease modeling. Furthermore, challenges and future outlooks of heart-on-a-chip platforms for in vitro measurement of cardiac functional properties are discussed.

Advances in sperm analysis: techniques, discoveries and applications.

Infertility affects one in six couples worldwide, and fertility continues to deteriorate globally, partly owing to a decline in semen quality. Sperm analysis has a central role in diagnosing and treating male factor infertility. Many emerging techniques, such as digital holography, super-resolution microscopy and next-generation sequencing, have been developed that enable improved analysis of sperm motility, morphology and genetics to help overcome limitations in accuracy and consistency, and improve sperm selection for infertility treatment. These techniques have also improved our understanding of fundamental sperm physiology by enabling discoveries in sperm behaviour and molecular structures. Further progress in sperm analysis and integrating these techniques into laboratories and clinics requires multidisciplinary collaboration, which will increase discovery and improve clinical outcomes.

A microdevice platform for characterizing the effect of mechanical strain magnitudes on the maturation of iPSC-Cardiomyocytes.

The use of human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) as an in vitro model of the heart is limited by their structurally and functionally immature phenotypes. During heart development, mechanical stimuli from in vivo microenvironments are known to regulate cardiomyocyte gene expression and maturation. Accordingly, protocols for culturing iPSC-CMs have recently incorporated mechanical or electromechanical stimulation to induce cellular maturation in vitro; however, the response of iPSC-CMs to different mechanical strain magnitudes is unknown, and existing techniques lack the capability to dynamically measure changes to iPSC-CM contractility in situ as maturation progresses. We developed a microdevice platform which applies cyclical strains of varying magnitudes (5%, 10%, 15% and 20%) to a monolayer of iPSC-CMs, coincidentally measuring contractile stress during mechanical stimulation using fluorescent nanobeads embedded in the microdevice's suspended membrane. Cyclic strain was found to induce circumferential cell alignment on the actuated membranes. In situ contractility measurements revealed that cyclic stimulation gradually increased cardiomyocyte contractility during a 10-day culture period. The contractile stress of iPSC-CM monolayers was found to increase with a higher strain magnitude and plateaued at 15% strain. Cardiomyocyte contractility positively correlated with the elongation of sarcomeres and an increased expression of ß-myosin heavy chain (MYH7) in a strain magnitude-dependent manner, illustrating how mechanical stress can be optimized for the phenotypic and proteomic maturation of the cells. iPSC-CMs with improved maturity have the potential to create a more accurate heart model in vitro for applications in disease modeling and therapeutic discovery.

Electrical impedance-based contractile stress measurement of human iPSC-Cardiomyocytes.

Heart failure fundamentally results from loss of cardio myocyte contractility. Developing new methods that quantify the contractile stress of the human cardiomyocyte would facilitate the study of the molecular mechanism of heart failure and advance therapy development, to improve the current five year survival for these patients. The measurement of cellular electrical impedance measurement was recently applied to monitor cardiomyocyte beating rate and rhythm, for the study at cellular maturation, and for drug screening. However, due to the lack of a quantified relationship between the impedance signal and contractile stress, change of cardiomyocyte contractile stress cannot genuinely be quantified from impedance measurements. Here, we report the first quantitative relationship between contractile stress and impedance, which enables the accurate prediction of cardiomyocyte contractility using impedance signals. Through simultaneous measurement of beating human iPSC-cardiomyocytes using impedance spectroscopy and atomic force microscopy, a power-law relationship between impedance and contractile stress was established with a confidence level of 95%. The quantitative relationship was validated using pharmacology known to alter cardiomyocyte contractility and beating (verapamil, using clinically relevant concentrations of 0.05 µM, 0.10 µM, and 0.15 µM). The contractile stress values as measured by AFM were 9.04 ± 0.14 kPa (0.05 µM), 7.72 ± 0.11 kPa (0.10 µM) and 6.23 ± 0.17 kPa (0.15 µM), and as predicted by impedance using the derived power-law relationship were 9.39 kPa, 7.76 kPa, and 6.05 kPa with a relative error of 3.73%. Our power-law relationship is the first to describe a quantitative correlation between contractile stress and impedance, broadening the application of electrical impedance measurement for characterizing complex cardiac functions (beating rate, beating rhythm and contractile stress).

Magnetic Measurement and Stimulation of Cellular and Intracellular Structures.

From single-pole magnetic tweezers to robotic magnetic-field generation systems, the development of magnetic micromanipulation systems, using electromagnets or permanent magnets, has enabled a multitude of applications for cellular and intracellular measurement and stimulation. Controlled by different configurations of magnetic-field generation systems, magnetic particles have been actuated by an external magnetic field to exert forces/torques and perform mechanical measurements on the cell membrane, cytoplasm, cytoskeleton, nucleus, intracellular motors, etc. The particles have also been controlled to generate aggregations to trigger cell signaling pathways and produce heat to cause cancer cell apoptosis for hyperthermia treatment. Magnetic micromanipulation has become an important tool in the repertoire of toolsets for cell measurement and stimulation and will continue to be used widely for further explorations of cellular/intracellular structures and their functions. Existing review papers in the literature focus on fabrication and position control of magnetic particles/structures (often termed micronanorobots) and the synthesis and functionalization of magnetic particles. Differently, this paper reviews the principles and systems of magnetic micromanipulation specifically for cellular and intracellular measurement and stimulation. Discoveries enabled by magnetic measurement and stimulation of cellular and intracellular structures are also summarized. This paper ends with discussions on future opportunities and challenges of magnetic micromanipulation in the exploration of cellular biophysics, mechanotransduction, and disease therapeutics.