Deep learning method for cell count from transmitted-light microscope

Automatic cell counting provides an effective tool for medical research and diagnosis.Currently,cell counting can be completed by transmitted-light microscope,however,it requires expert knowledge and the counting accuracy which is unsatisfied for overlapped cells.Further,the image-translation-based detection method has been proposed and the potential has been shown to accomplish cell counting from transmitted-light microscope,automatically and effectively.In this work,a new deep-learning(DL)-based two-stage detection method(cGAN-YOLO)is designed to further enhance the performance of cell counting,which is achieved by combining a DL-based fluorescent image translation model and a DL-based cell detection model.The various results show that cGAN-YOLO can effectively detect and count some different types of cells from the acquired transmitted-light microscope images.Compared with the previously reported YOLO-based one-stage detection method,high recognition accuracy(RA)is achieved by the cGAN-YOLO method,with an improvement of 29.80%.Furthermore,we can also observe that cGAN-YOLO obtains an improvement of 12.11%in RA compared with the previously reported image-translation-based detection method.In a word,cGAN-YOLO makes it possible to implement cell counting directly from the experimental acquired transmitted-light microscopy images with high flexibility and performance,which extends the applicability in clinical research.

A rapid and visual detection method for Crimean-Congo hemorrhagic fever virus by targeting S gene

Crimean-Congo hemorrhagic fever(CCHF)is a zoonotic disease caused by the CCHF virus(CCHFV),which is primarily transmitted by ticks(Lorenzo Juanes et al.2023).It is an emerging disease that occurs sporadically in Africa,Asia,and Europe,with a high morbidity and mortality rate,as high as 30%in humans(Ceylan et al.2013).CCHFV,belonging to genus Nairovirus,family Bunyaviridae,was first identified in the Congo in the 1960s.

Assessment of cortical bone fatigue using coded nonlinear ultrasound

Bone fatigue accumulation is a factor leading to bone fracture,which is a progressive process of microdamage deteriorating under long-term and repeated stress.Since the microdamage of the early stage in bone is difficult to be investigated by linear ultrasound,the second harmonic generation method in nonlinear ultrasound technique is employed in this paper,which is proved to be more sensitive to microdamage.To solve the deficiency that the second harmonic component is easily submerged by noise in traditional nonlinear measurement,a weighted chirp coded sinusoidal signal was applied as the ultrasonic excitation,while pulse inversion is implemented at the receiving side.The effectiveness of this combination to improve the signal-to-noise ratio has been demonstrated by in vitro experiment.Progressive fatigue loading experiments were conducted on the cortical bone plate in vitro for microdamage generation.There was a significant increase in the slope of the acoustic nonlinearity parameter with the propagation distance(increased by 8%and 24%respectively)when the bone specimen was at a progressive level of microdamage.These results indicate that the coded nonlinear ultrasonic method might have the potential in diagnosing bone fatigue.

Amplitude modulation excitation for cancellous bone evaluation using a portable ultrasonic backscatter instrumentation

The ultrasonic backscatter(UB)has the advantage of non-invasively obtaining bone density and structure,expected to be an assessment tool for early diagnosis osteoporosis.All former UB measurements were based on exciting a short single-pulse and analyzing the ultrasonic signals backscattered in bone.This study aims to examine amplitude modulation(AM)ultrasonic excitation with UB measurements for predicting bone characteristics.The AM multiple lengths excitation and backscatter measurement(AM-UB)functions were integrated into a portable ultrasonic instrument for bone characterization.The apparent integrated backscatter coefficient in the AM excitation(AIB_(AM))was evaluated on the AM-UB instrumentation.The correlation coefficients of the AIB_(AM) estimating volume fraction(BV/TV),structure model index(SMI),and bone mineral density(BMD)were then analyzed.Significant correlations(|R|=0.82-0.93,p<0.05)were observed between the AIB_(AM),BV/TV,SMI,and BMD.By growing the AM excitation length,the AIB_(AM) values exhibit more stability both in 1.0-MHz and 3.5-MHz measurements.The recommendations in AM-UB measurement were that the avoided length(T1)should be lower than AM excitation length,and the analysis length(T2)should be enough long but not more than AM excitation length.The authors conducted an AM-UB measurement for cancellous bone characterization.Increasing the AM excitation length could substantially enhance AIB_(AM) values stability with varying analyzed signals.The study suggests the portable AM-UB instrument with the integration of real-time analytics software that might provide a potential tool for osteoporosis early screening.

Ultrasonic backscatter characterization of cancellous bone using a general Nakagami statistical model

The goal of this study is to analyze the statistics of the backscatter signal from bovine cancellous bone using a Nakagami model and to evaluate the feasibility of Nakagami-model parameters for cancellous bone characterization. Ultrasonic backscatter measurements were performed on 24 bovine cancellous bone specimens in vitro and the backscatter signals were compensated for the frequency-dependent attenuation prior to the envelope detection. The statistics of the backscatter envelope were modeled using the Nakagami distribution. Our results reveal that the backscatter envelope mainly followed pre-Rayleigh distributions, and the deviations of the backscatter envelope from Rayleigh distribution decreased with increasing bone density. The Nakagami shape parameter(i.e., m) was significantly correlated with bone densities(R = 0.78–0.81, p < 0.001) and trabecular microstructures(|R| = 0.46–0.78, p < 0.05). The scale parameter(i.e.,?) and signal-to-noise ratio(SNR) also yielded significant correlations with bone density and structural features. Multiple linear regressions showed that bone volume fraction(BV/TV) was the main predictor of the Nakagami parameters,and microstructure produced significantly independent contribution to the prediction of Nakagami distribution parameters,explaining an additional 10.2% of the variance at most. The in vitro study showed that statistical parameters derived with Nakagami model might be useful for cancellous bone characterization, and statistical analysis has potential for ultrasonic backscatter bone evaluation.

Novel discovery of schisandrin A regulating the interplay of autophagy and apoptosis in oligoasthenospermia by targeting SCF/c-kit and TRPV1 via biosensors

Oligoasthenospermia is the primary cause of infertility.However,there are still enormous challenges in the screening of critical candidates and targets of oligoasthenospermia owing to its complex mechanism.In this study,stem cell factor(SCF),c-kit,and transient receptor potential vanilloid 1(TRPV1)biosensors were successfully established and applied to studying apoptosis and autophagy mechanisms.Interestingly,the detection limit reached 2.787×10^(-15)g/L,and the quantitative limit reached 1.0×10^(-13)g/L.Furthermore,biosensors were used to investigate the interplay between autophagy and apoptosis.Schisandrin A is an excellent candidate to form a system with c-kit similar to SCF/c-kit with a detection constant(K_(D))of 5.701×10^(-11)mol/L,whereas it had no affinity for SCF.In addition,it also inhibited autophagy in oligoasthenospermia through antagonizing TRPV1 with a K_(D) of up to 4.181×10^(-10)mol/L.In addition,in vivo and in vitro experiments were highly consistent with the biosensor.In summary,high-potency schisandrin A and two potential targets were identified,through which schisandrin A could reverse the apoptosis caused by excessive autophagy during oligoasthenospermia.Our study provides promising insights into the discovery of effective compounds and potential targets via a well-established in vitro-in vivo strategy.

Research on Potential Pathogenesis of Advanced Diabetic Nephropathy based on Bioinformatic Analysis

Objective:Through the bioinformatic analysis of gene chips related to advanced diabetic nephropathy in the GEO database,the key genes of advanced diabetic nephropathy are screened,whose biological functions and signal pathways are predicted as well.Methods:The gene chips related to advanced diabetic nephropathy from the GEO expression profile database was downloaded,and the differentially expressed genes in patients with advanced diabetic nephropathy and normal people were analyzed through R.For the screened differentially expressed genes,the biological function of GO and the enrichment analysis of KEGG signal pathway were used to predict their biological functions and related signal pathways.In addition,a protein-protein interaction network was constructed,so as to screen core pathogenic genes utilizing STRING database and Cytoscape.Results:By analyzing the chip GSE142025,301 differential genes were obtained,including 197 up-regulated genes and 104 down-regulated genes.Both GO annotation and enrichment analysis suggested that differential genes were mainly involved in immune-inflammatory response and cytokine action.Furthermore,KEGG pathway analysis suggested that the most important pathway related to advanced diabetic nephropathy was MAPK signaling pathway.Through protein-protein interaction network and module analysis,C3,CCR2,CCL19,and SAA1 were selected as the core sites of the interaction.Conclusions:Differential genes participate in the pathogenesis of advanced diabetic nephropathy through the KEGG pathway,the immune inflammatory response and cytokine action,which provides new ways for the diagnosis and treatment of advanced diabetic nephropathy.

The Protocol of Ultrasonic Backscatter Measurements of Musculoskeletal Properties

This study aims to introduce the protocol for ultrasonic backscatter measurements of musculoskeletal properties based on a novel ultrasonic backscatter bone diagnostic(UBBD)instrument.Dual-energy X-ray absorptiometry(DXA)can be adopted to measure bone mineral density(BMD)in the hip,spine,legs and the whole body.The muscle and fat mass in the legs and the whole body can be also calculated by DXA body composition analysis.Based on the proposed protocol for backscatter measurements by UBBD,ultrasonic backscatter signals can be measured in vivo,deriving three backscatter parameters[apparent integral backscatter(AIB),backscatter signal peak amplitude(BSPA)and the corresponding arrival time(BSPT)].AIB may provide important diagnostic information about bone properties.BSPA and BSPT may be important indicators of muscle and fat properties.The standardized backscatter measurement protocol of the UBBD instrument may have the potential to evaluate musculoskeletal characteristics,providing help for promoting the application of the backscatter technique in the clinical diagnosis of musculoskeletal disorders(MSDs),such as osteoporosis and muscular atrophy.

Screening COVID-19 from chest X-ray images by an optical diffractive neural network with the optimized F number

The COVID-19 pandemic continues to significantly impact people's lives worldwide, emphasizing the critical need for effective detection methods. Many existing deep learning-based approaches for COVID-19 detection offer high accuracy but demand substantial computing resources, time, and energy. In this study, we introduce an optical diffractive neural network(ODNN-COVID), which is characterized by low power consumption, efficient parallelization, and fast computing speed for COVID-19 detection. In addition, we explore how the physical parameters of ODNN-COVID affect its diagnostic performance. We identify the F number as a key parameter for evaluating the overall detection capabilities. Through an assessment of the connectivity of the diffractive network, we established an optimized range of F number, offering guidance for constructing optical diffractive neural networks. In the numerical simulations, a three-layer system achieves an impressive overall accuracy of 92.64% and 88.89% in binary-and threeclassification diagnostic tasks. For a single-layer system, the simulation accuracy of 84.17% and the experimental accuracy of 80.83% can be obtained with the same configuration for the binary-classification task, and the simulation accuracy is 80.19% and the experimental accuracy is 74.44% for the three-classification task. Both simulations and experiments validate that the proposed optical diffractive neural network serves as a passive optical processor for effective COVID-19 diagnosis, featuring low power consumption, high parallelization, and fast computing capabilities. Furthermore, ODNN-COVID exhibits versatility, making it adaptable to various image analysis and object classification tasks related to medical fields owing to its general architecture.

Virtual Fluorescence Translation for Biological Tissue by Conditional Generative Adversarial Network

Fluorescence labeling and imaging provide an opportunity to observe the structure of biological tissues,playing a crucial role in the field of histopathology.However,when labeling and imaging biological tissues,there are still some challenges,e.g.,time-consuming tissue preparation steps,expensive reagents,and signal bias due to photobleaching.To overcome these limitations,we present a deep-learning-based method for fluorescence translation of tissue sections,which is achieved by conditional generative adversarial network(cGAN).Experimental results from mouse kidney tissues demonstrate that the proposed method can predict the other types of fluorescence images from one raw fluorescence image,and implement the virtual multi-label fluorescent staining by merging the generated different fluorescence images as well.Moreover,this proposed method can also effectively reduce the time-consuming and laborious preparation in imaging processes,and further saves the cost and time.

Fluo-Fluo translation based on deep learning

Fluorescence microscopy technology uses fluorescent dyes to provide highly specific visualization of cell components,which plays an important role in understanding the subcellular structure.However,fluorescence microscopy has some limitations such as the risk of non-specific cross labeling in multi-labeled fluorescent staining and limited number of fluo-rescence labels due to spectral overlap.This paper proposes a deep learning-based fluorescence to fluorescence[Flu0-Fluo]translation method,which uses a conditional generative adversarial network to predict a fluorescence image from another fluorescence image and further realizes the multi-label fluorescent staining.The cell types used include human motor neurons,human breast cancer cells,rat cortical neurons,and rat cardiomyocytes.The effectiveness of the method is verified by successfully generating virtual fluorescence images highly similar to the true fluorescence images.This study shows that a deep neural network can implement Fluo-Fluo translation and describe the localization relationship between subcellular structures labeled with different fluorescent markers.The proposed Fluo-Fluo method can avoid non-specific cross labeling in multi-label fluorescence staining and is free from spectral overlaps.In theory,an unlimited number of fluorescence images can be predicted from a single fluorescence image to characterize cells.