Set Of Groups Research Articles

Artificial intelligence automated surgical phases recognition in intraoperative videos of laparoscopic pancreatoduodenectomy.

Laparoscopic pancreatoduodenectomy (LPD) is one of the most challenging operations and has a long learning curve. Artificial intelligence (AI) automated surgical phase recognition in intraoperative videos has many potential applications in surgical education, helping shorten the learning curve, but no study has made this breakthrough in LPD. Herein, we aimed to build AI models to recognize the surgical phase in LPD and explore the performance characteristics of AI models. Among 69 LPD videos from a single surgical team, we used 42 in the building group to establish the models and used the remaining 27 videos in the analysis group to assess the models' performance characteristics. We annotated 13 surgical phases of LPD, including 4 key phases and 9 necessary phases. Two minimal invasive pancreatic surgeons annotated all the videos. We built two AI models for the key phase and necessary phase recognition, based on convolutional neural networks. The overall performance of the AI models was determined mainly by mean average precision (mAP). Overall mAPs of the AI models in the test set of the building group were 89.7% and 84.7% for key phases and necessary phases, respectively. In the 27-video analysis group, overall mAPs were 86.8% and 71.2%, with maximum mAPs of 98.1% and 93.9%. We found commonalities between the error of model recognition and the differences of surgeon annotation, and the AI model exhibited bad performance in cases with anatomic variation or lesion involvement with adjacent organs. AI automated surgical phase recognition can be achieved in LPD, with outstanding performance in selective cases. This breakthrough may be the first step toward AI- and video-based surgical education in more complex surgeries.

“The offline part was the essence of it” – the social history of iWiW, the Hungarian online social network

iWiW was one of the first Hungarian online social networks (OSNs) and the largest ever. The site was launched in 2002, and at its peak in 2010 it had 4.5 million members, more than two-thirds of all internet users in the country. The site was shut down in 2014 by its then-owner, Hungarian Telekom. As the first socio-historical account of the site, based on oral history interviews, this paper aims to shed light on a lesser-known chapter of local digital media history by providing insights into the development of iWiW. This examination includes a focus on the social meanings and values that underpinned the OSN’s development strategies, which were contested by a changing set of relevant social groups. The paper shows how the debates of the five main relevant social groups (entertainers, early enthusiasts, entrepreneurs, mass users, and corporate entrepreneurs) shaped the development of the site during the OSNs three main eras. In each era, the site had different goals, technological features, business and development strategies, favoured and represented by the most influential relevant social groups at the time. In this sense, the "offline part" of the network has always played a decisive role in the history of iWiW, from the beginning to the end.

Relationship between immune phenotype and treatment selection of Chemo-IO vs. IO-only in TPS-high NSCLC using hypothetical test-and-control group generation based on survival data extracted from phase III trials.

e13569 Background: Immunotherapy has become the treatment of choice for the first-line treatment of unresectable NSCLC with high PD-L1 tumor proportion score (TPS). However, the optimal treatment regimen, particularly when to add chemotherapy, remains unclear. To evaluate whether an immune microenvironment biomarker could help guide the treatment regimen, we have explored computational hypothetical analysis applying historical proportion of inflamed immune phenotype (IIP) to published outcome data of clinical trials. Methods: R package “IPDfromKM” was used to reconstruct raw time-to-event outcomes from Kaplan-Meier (KM) curves. Data from KN-189 and KN-407 for anti-PD1 with chemotherapy (Chemo-IO), and KN-024 and KN-042 for anti-PD1 (IO-only) in the first-line treatment of NSCLC were included for analysis. Overall survival (OS) data were computed from KM of pts with PD-L1 TPS of ≧ 50% from these studies (KN-189 n=132, KN-407 n=73, KN-024 n=154, KN-042 n=299). It was assumed that Lunit SCOPE IO classified pts into IIP or non-IIP at a ratio of 57% vs. 43%, based on previous analysis. One hundred hypothetical sets of Chemo-IO and IO-only groups satisfying hazard ratio (HR) of IIP to non-IIP as 0.81 and 0.61 (± 0.025) while maintaining the proportion of IIP were randomly generated. A total of 10,000 (100 x 100) combinations of cohorts were analyzed to test our hypothesis of superiority of Chemo-IO over IO-only in non-IIP as well as non-superiority of Chemo-IO over IO-only in IIP. Results: Patient-level OS data for a total of 205 pts with Chemo-IO and 453 pts with IO-only were reconstructed. The median OS was favorable in Chemo-IO vs IO-only (HR 0.73, 95% confidence interval [CI] 0.54-0.99, p = 0.045). From 10,000 hypothetical combinations of cohorts, the median HR of Chemo-IO compared to IO-only was 0.61 (interquartile range [IQR] 0.60-0.63) in non-IIP group, whereas that was 0.86 (IQR 0.84-0.89) in IIP group (Table). The upper 95% bound of HR was < 1 for 8,611 combinations of non-IIP groups, but > 1 for all possible combinations of IIP groups. Given a non-superiority margin of 1.37(=1/0.73), 8,306 combinations were positive for the upper 95% bound of HR < 1.37. Overall, 6,917 matched combinations supported our hypothesis. Conclusions: From the computational analysis, 69.17% cases of hypothetical combinations support that Chemo-IO would be superior to IO-only in non-IIP group, whereas Chemo-IO would not be superior to IO-only in IIP group, implying a potential value of confirmatory biomarker analysis in providing evidence for regimen selection. [Table: see text]

Totally deranged elements of almost simple groups and invariable generating sets

AbstractBy a classical theorem of Jordan, every faithful transitive action of a non‐trivial finite group has a derangement (an element with no fixed points). The existence of derangements with additional properties has attracted much attention, especially for faithful primitive actions of almost simple groups. In this paper, we show that an almost simple group can have an element that is a derangement in every faithful primitive action, and we call these elements totally deranged. In fact, we classify the totally deranged elements of all almost simple groups, showing that an almost simple group contains a totally deranged element only if the socle of is or with . Using this, we classify the invariable generating sets of a finite simple group of the form where and , answering a question of Garzoni. As a final application, we classify the elements of almost simple groups that are contained in a unique maximal subgroup in the case where is not core‐free, which complements the recent work of Guralnick and Tracey addressing the case where is core‐free.

Extension of the Constantinou and Gani group contribution method with the Tochigi method through an automatic conversion procedure

Group contribution methods (GCMs) are widely employed across various disciplines to estimate compound properties when experimental data is lacking in the literature. While several methods exist, none are comprehensive, as they exhibit gaps either in functional groups or the predicted properties or simply do not offer the required accuracy. As a result, different methods may be necessary to evaluate distinct properties of the same compound. Typically, switching between these models is performed manually due to variations in the sets of functional groups employed. However, with the advancements in computational power and numerical optimum search, there is a growing need to automate the conversion between different group contribution methods. This study presents a procedure to extend the property estimation of the Constantinou and Gani method with vapor pressure by supplementing it with the Tochigi method using an automated group conversion algorithm. The difficulties of automatic conversion procedures, resulting from the differences in the group sets and the shortcomings of the GCMs, are also highlighted. It is also demonstrated that the accuracy of the acentric factor estimation can only be refined to a limited extent by incorporating the Tochigi method, which is, however, indispensable for the several groups where the Constantinou and Gani group contribution values are missing.

Do multi-channel use and online engagement matter for critical literacy and information verification behaviors? Comprehensive comparisons of six generations from before 1954 (war generation) to generation Z (after 1997)

The prevalence of multi-channel information consumption, which involves the simultaneous engagement with multiple information sources across various platforms, underscores the urgent need for research on its effects as it profoundly influences how individuals seek, evaluate, and interact with information, ultimately shaping their information-related choices and behaviors. Using an innovative empirical method to extract a set of latent groups based on individuals' multi-channel practices, the study examined the role of channel plurality relative to how people evaluate and consume information online. Importantly, this research examined how conditions of the critical literacy and practices can be configured to benefit six distinct generational groups, ranging from the youngest generation group (1990s) to the oldest generation (1950s). Based on results of a nationally representative survey (N = 10,032), the current research demonstrated how individuals’ channel plurality contributes to their critical literacy and information verification behaviors. Finally, this study identified differences in information practices manifested in six generational groups. In the end, this study bridged two streams of research, new media literacy and the digital divide, and discussed how different educational and regulatory approaches can be made to encourage critical information consumption in an increasingly digitized network environment.

Glycolysis and Automated Plaque Regrowth Method for Evaluation of Antimicrobial Performance.

This study explored the potential of a new in vitro method in evaluating antiplaque benefits from five sets of antimicrobial systems including cetylpyridinium chloride (CPC), stannous fluoride (SnF2), Listerine essential oil mouthwashes (+/- alcohol), zinc chloride (ZnCl2), and sodium fluoride. (NaF). Gingival dental plaque was collected and propagated using sterilized tryptic soy broth and sucrose, and then allocated into separate glycolysis and regrowth recipes for antiplaque evaluations. Glycolysis measurements (in duplicate) were recorded via pH microelectrode on plaque-treatment samples thermomixed (1200 rpm, 37 °C) for 4 h. For plaque regrowth, optical densities (in duplicate) were automatically collected on plaque-treatment samples using a microplate reader (linear shaking, 37 °C) from baseline to 4 h. Calculations of percent change in pH and optical density were performed and analyzed for each set of antimicrobial treatment groups. Statistical analysis (one-way ANOVA, Student-Newman-Keuls stepwise comparison tests) revealed dose responses and significant differences (p < 0.05) among treatment groups, including between negative and clinically relevant positive controls. This lab method produces results consistent with published clinical observations. This glycolysis and plaque growth method is sensitive to antimicrobial mechanisms of action, and may offer a convenient and clinically relevant screening tool in the evaluation of putative antimicrobial agents and formulations.

Elite Coalitions and Rebel Control in Northern Côte d’Ivoire

ABSTRACT How do armed movements gain civilian compliance and prevent counter-mobilization in localities under their wartime control? This paper argues that rebel capacity to control civilians is contingent on the coalitions formed between rebel mid-level commanders and elite groups embedded in local communities. Rebels are confronted with coalitional choices because local communities are contested and comprised of multiple sets of elite groups. Coalitional choices influence rebel control because they determine whether rebels ally with strong or weak local partners. Where rebel allies are weak, they should confront serious challenges to eliciting civilian compliance. Where rebel allies are stronger, rebels should confront less resistance. I argue that the availability of elites as potential allies for armed movements depends on their pre-conflict proximity to the state. This paper develops this argument through a comparison of three localities governed by the Forces Nouvelles (FN) in Côte d’Ivoire (2002–2011).

Enantioselective synthesis of chiral α,α-dialkyl indoles and related azoles by cobalt-catalyzed hydroalkylation and regioselectivity switch

General, catalytic and enantioselective construction of chiral α,α-dialkyl indoles represents an important yet challenging objective to be developed. Herein we describe a cobalt catalyzed enantioselective anti-Markovnikov alkene hydroalkylation via the remote stereocontrol for the synthesis of α,α-dialkyl indoles and other N-heterocycles. This asymmetric C(sp3)−C(sp3) coupling features high flexibility in introducing a diverse set of alkyl groups at the α-position of chiral N-heterocycles. The utility of this methodology has been demonstrated by late-stage functionalization of drug molecules, asymmetric synthesis of bioactive molecules, natural products and functional materials, and identification of a class of molecules exhibiting anti-apoptosis activities in UVB-irradiated HaCaT cells. Ligands play a vital role in controlling the reaction regioselectivity. Changing the ligand from bi-dentate L6 to tridentate L12 enables CoH-catalyzed Markovnikov hydroalkylation. Mechanistic studies disclose that the anti-Markovnikov hydroalkylation involves a migratory insertion process while the Markovnikov hydroalkylation involves a MHAT process.

Simulations predict preferred Mg2+ coordination in a nonenzymatic primer-extension reaction center

The mechanism by which genetic information was copied prior to the evolution of ribozymes is of great interest because of its importance to the origin of life. The most effective known process for the nonenzymatic copying of an RNA template is primer extension by a two-step pathway in which 2-aminoimidazole-activated nucleotides first react with each other to form an imidazolium-bridged intermediate that subsequently reacts with the primer. Reaction kinetics, structure-activity relationships, and X-ray crystallography have provided insight into the overall reaction mechanism, but many puzzles remain. In particular, high concentrations of Mg2+ are required for efficient primer extension, but the mechanism by which Mg2+ accelerates primer extension remains unknown. By analogy with the mechanism of DNA and RNA polymerases, a role for Mg2+ in facilitating the deprotonation of the primer 3′-hydroxyl is often assumed, but no catalytic metal ion is seen in crystal structures of the primer-extension complex. To explore the potential effects of Mg2+ binding in the reaction center, we performed atomistic molecular dynamics simulations of a series of modeled complexes in which a Mg2+ ion was placed in the reaction center with inner-sphere coordination with different sets of functional groups. Our simulations suggest that coordination of a Mg2+ ion with both O3′ of the terminal primer nucleotide and the pro-Sp nonbridging oxygen of the reactive phosphate of an imidazolium-bridged dinucleotide would help to pre-organize the structure of the primer/template substrate complex to favor the primer-extension reaction. Our results suggest that the catalytic metal ion may play an important role in overcoming electrostatic repulsion between a deprotonated O3′ and the reactive phosphate of the bridged dinucleotide and lead to testable predictions of the mode of Mg2+ binding that is most relevant to catalysis of primer extension.

Gut microbiome alterations in Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS) suggests variation in the gut-brain axis and gut metabolic potential

PANDAS, or Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections is a form of Children’s Post infectious Autoimmune Encephalopathy (CPAE), a neuropsychiatric condition characterized by the abrupt onset of diverse neuropsychiatric manifestations following streptococcal infection. It is an increasingly recognized neuropsychiatric disorder that affects a subset of children whose underlying etiology remains unclear. Gut microbiome plays a crucial role in its host brain development and is implicated throughout the life in neurobehavior and neurological disorders. Emerging evidence suggest gut dysbiosis resulting in altered intestinal flora can impair gut barrier and increase leakiness in the blood-brain barrier leading to inflammation. We aimed to investigate if the microbiome is altered in pediatric patients with PANDAS. Stool (n=19), oral (n=20) and nasal swabs (n=18) were collected from patients diagnosed with PANDAS or healthy controls (HC) and used for 16S analysis. 13 stool samples (7 HC and 6 PANDAS) samples were further used for shogun metagenomics and 17 for untargeted metabolomics (8 HC and 9 PANDAS). 16S rRNA gene sequencing identified significant difference in microbial α diversity, with lower Shannon H’ in PANDAS compared to HC (Welch’s t-test t(16.95)=-2.6, p=0.02), and in β-diversity (PERMANOVA R2=0.08 p=0.04) between PANDAS and HC fecal but not throat and nasal samples. Shotgun metagenomics identified a significant reduction in gene diversity (Welch’s t-test t(9.9)=-3.06, p=0.01) and changes in various KEGG ortholog modules involved in gut metabolism and gut-brain axis. Significant differences were also detected in metabolite composition (PERMANOVA R2=0.10, p=0.02). We also identified a set of microbial groups and metabolites enriched or depleted in CPAE patients, involved in different aspects of host physiology and gut metabolism. Our results show that PANDAS patients have an altered gut microbial community, which may exacerbate their behavioral symptoms and it should be taken into consideration in unveiling the complex pathogenesis and selection of treatment options. Arizona Department of Health Services RFGA2022-010-23 (PK) and ADHS17-00007403 (SR, MD); The Alex Manfull Fund (PK). This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

MORE ON GALOIS COHOMOLOGY, DEFINABILITY, AND DIFFERENTIAL ALGEBRAIC GROUPS

AbstractAs a continuation of the work of the third author in [5], we make further observations on the features of Galois cohomology in the general model theoretic context. We make explicit the connection between forms of definable groups and first cohomology sets with coefficients in a suitable automorphism group. We then use a method of twisting cohomology (inspired by Serre’s algebraic twisting) to describe arbitrary fibres in cohomology sequences—yielding a useful “finiteness” result on cohomology sets.Applied to the special case of differential fields and Kolchin’s constrained cohomology, we complete results from [3] by proving that the first constrained cohomology set of a differential algebraic group over a bounded, differentially large, field is countable.

Engineering 3D Floquet Codes by Rewinding

Floquet codes are a novel class of quantum error-correcting codes with dynamically generated logical qubits arising from a periodic schedule of noncommuting measurements. We utilize the interpretation of measurements in terms of condensation of topological excitations and the rewinding of measurement sequences to engineer new examples of Floquet codes. In particular, rewinding is advantageous for obtaining a desired set of instantaneous stabilizer groups on both toric and planar layouts. Our first example is a Floquet code with instantaneous stabilizer codes that have the same topological order as the three-dimensional (3D) toric code(s). This Floquet code also exhibits a splitting of the topological order of the 3D toric code under the associated sequence of measurements, i.e., an instantaneous stabilizer group of a single copy of the 3D toric code in one round transforms into an instantaneous stabilizer group of two copies of the 3D toric code up to nonlocal stabilizers in the following round. We further construct boundaries for this 3D code and argue that stacking it with two copies of the 3D subsystem toric code allows for a transversal implementation of the logical non-Clifford controlled-controlled-Z gate. We also show that the coupled-layer construction of the X-cube Floquet code can be modified by a rewinding schedule such that each of the instantaneous stabilizer codes is finite depth equivalent to the X-cube model up to toric codes; the X-cube Floquet code exhibits a splitting of the X-cube model into a copy of the X-cube model and toric codes under the measurement sequence. Our final 3D example is a generalization of the 2D Floquet toric code on the honeycomb lattice to three dimensions, which has instantaneous stabilizer codes with the same topological order as the 3D fermionic toric code. Published by the American Physical Society 2024

Digraphs of power maps over finite nilpotent groups

Given a finite group G and an integer t, let G(φt/G) be the directed graph with vertex set G and directed edges g→gt,g∈G. We prove that nilpotent groups are precisely the class of groups yielding a special symmetry on the graphs G(φt/G) that has appeared several times in the literature (considering functions over various algebraic structures). We then explore the question on when two finite nilpotent groups yield the same digraphs as t runs over Z: this defines an equivalence relation ∼φ. We completely describe the set of integers M in which the relation ∼φ is equivalent to the isomorphism of groups when one restricts to the set of nilpotent groups of order M. Our proofs rely on arguments from number theory and combinatorics.

Colloquial Emotive Lexicon of Russian in Aspect of Polyparadigmality: Lexicographic Perspective (Thesaurus Dictionary “Emotion Alphabet”)

The article is dedicated to the study of colloquial emotive lexicon of the Russian in the aspect of polyparadigmality. The peculiarities of its categorization, conceptualization, and interpretation are examined from the perspectives of various scientific approaches. The analysis of colloquial emotive lexicon in the Russian is conducted in comparison with a multitude of normative literary lexicon presented in the thesaurus dictionary ‘Emotion Alphabet.’ Together, they comprise 11,210 units of analysis, with 2,073 lexemes belonging to colloquial emotive lexicon. It is considered as a lexical-semantic space in a spatial dimension, with its internal structure constructed in a horizontal-vertical dimension at the intersection of lexical sets of colloquial emotive lexicon of different nature. The study employs a complex of linguistic methods, including statistical description, structuralsemantic, functional-semantic, denotativeideographic, linguistic-cognitive, and other methods of analysis. This allows for presenting colloquial emotive lexicon as an overlapping set of various groups of lexemes: denotative-ideographic, cognitive-derivational, semantic-functional, and functional-stylistic. Trends and cognitive strategies in the formation of this set are identified in comparison with literary language. The results of the study will be used in the future for creating a thesaurus dictionary of colloquial emotive lexicon.

Inherited retinal dystrophies and orphan designations in the European Union.

Inherited Retinal Dystrophies (IRD) are diverse rare diseases that affect the retina and lead to visual impairment or blindness. Research in this field is ongoing, with over 60 EU orphan medicinal products designated in this therapeutic area by the Committee for Orphan Medicinal Products (COMP) at the European Medicines Agency (EMA). Up to now, COMP has used traditional disease terms, like retinitis pigmentosa, for orphan designation regardless of the product's mechanism of action. The COMP reviewed the designation approach for IRDs taking into account all previous Orphan Designations (OD) experience in IRDs, the most relevant up to date scientific literature and input from patients and clinical experts. Following the review, the COMP decided that there should be three options available for orphan designation concerning the condition: i) an amended set of OD groups for therapies that might be used in a broad spectrum of conditions, ii) a gene-specific designation for targeted therapies, and iii) an occasional term for products that do not fit in the above two categories. The change in the approach to orphan designation in IRDs caters for different scenarios to allow an optimum approach for future OD applications including the option of a gene-specific designation. By applying this new approach, the COMP increases the regulatory clarity, efficiency, and predictability for sponsors, aligns EU regulatory tools with the latest scientific and medical developments in the field of IRDs, and ensures that all potentially treatable patients will be included in the scope of an OD.

Metal-free construction of diverse 1,2,4-triazolo[1,5-a]pyridines on water

A transition-metal- and oxidant-free amination/cyclization reaction to access 1,2,4-triazolo[1,5-a]pyridines was realized in water by using amino diphenylphosphinate as amino source. A broad array of readily accessible N-(pyridyl)amides could be converted into the products featuring a diverse set of functional groups. The sustainable methodology was successfully applied to the late-stage functionalization of natural products and drugs.

Automated classification of giant virus genomes using a random forest model built on trademark protein families

Viruses of the phylum Nucleocytoviricota, often referred to as “giant viruses,” are prevalent in various environments around the globe and play significant roles in shaping eukaryotic diversity and activities in global ecosystems. Given the extensive phylogenetic diversity within this viral group and the highly complex composition of their genomes, taxonomic classification of giant viruses, particularly incomplete metagenome-assembled genomes (MAGs) can present a considerable challenge. Here we developed TIGTOG (Taxonomic Information of Giant viruses using Trademark Orthologous Groups), a machine learning-based approach to predict the taxonomic classification of novel giant virus MAGs based on profiles of protein family content. We applied a random forest algorithm to a training set of 1531 quality-checked, phylogenetically diverse Nucleocytoviricota genomes using pre-selected sets of giant virus orthologous groups (GVOGs). The classification models were predictive of viral taxonomic assignments with a cross-validation accuracy of 99.6% at the order level and 97.3% at the family level. We found that no individual GVOGs or genome features significantly influenced the algorithm’s performance or the models’ predictions, indicating that classification predictions were based on a comprehensive genomic signature, which reduced the necessity of a fixed set of marker genes for taxonomic assigning purposes. Our classification models were validated with an independent test set of 823 giant virus genomes with varied genomic completeness and taxonomy and demonstrated an accuracy of 98.6% and 95.9% at the order and family level, respectively. Our results indicate that protein family profiles can be used to accurately classify large DNA viruses at different taxonomic levels and provide a fast and accurate method for the classification of giant viruses. This approach could easily be adapted to other viral groups.

Autism spectrum disorder detection with kNN imputer and machine learning classifiers via questionnaire mode of screening.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder. ASD cannot be fully cured, but early-stage diagnosis followed by therapies and rehabilitation helps an autistic person to live a quality life. Clinical diagnosis of ASD symptoms via questionnaire and screening tests such as Autism Spectrum Quotient-10 (AQ-10) and Quantitative Check-list for Autism in Toddlers (Q-chat) are expensive, inaccessible, and time-consuming processes. Machine learning (ML) techniques are beneficial to predict ASD easily at the initial stage of diagnosis. The main aim of this work is to classify ASD and typical developed (TD) class data using ML classifiers. In our work, we have used different ASD data sets of all age groups (toddlers, adults, children, and adolescents) to classify ASD and TD cases. We implemented One-Hot encoding to translate categorical data into numerical data during preprocessing. We then used kNN Imputer with MinMaxScaler feature transformation to handle missing values and data normalization. ASD and TD class data is classified using Support vector machine, k-nearest-neighbor (KNN), random forest (RF), and artificial neural network classifiers. RF gives the best performance in terms of the accuracy of 100% with different training and testing data split for all four types of data sets and has no over-fitting issue. We have also examined our results with already published work, including recent methods like Deep Neural Network (DNN) and Convolution Neural Network (CNN). Even using complex architectures like DNN and CNN, our proposed methods provide the best results with low-complexity models. In contrast, existing methods have shown accuracy upto 98% with log-loss upto 15%. Our proposed methodology demonstrates the improved generalization for real-time ASD detection during clinical trials.

Convolutional Neural Network-Driven Impedance Flow Cytometry for Accurate Bacterial Differentiation.

Impedance flow cytometry (IFC) has been demonstrated to be an efficient tool for label-free bacterial investigation to obtain the electrical properties in real time. However, the accurate differentiation of different species of bacteria by IFC technology remains a challenge owing to the insignificant differences in data. Here, we developed a convolutional neural networks (ConvNet) deep learning approach to enhance the accuracy and efficiency of the IFC toward distinguishing various species of bacteria. First, more than 1 million sets of impedance data (comprising 42 characteristic features for each set) of various groups of bacteria were trained by the ConvNet model. To improve the efficiency for data analysis, the Spearman correlation coefficient and the mean decrease accuracy of the random forest algorithm were introduced to eliminate feature interaction and extract the opacity of impedance related to the bacterial wall and membrane structure as the predominant features in bacterial differentiation. Moreover, the 25 optimized features were selected with differentiation accuracies of >96% for three groups of bacteria (bacilli, cocci, and vibrio) and >95% for two species of bacilli (Escherichia coli and Salmonella enteritidis), compared to machine learning algorithms (complex tree, linear discriminant, and K-nearest neighbor algorithms) with a maximum accuracy of 76.4%. Furthermore, bacterial differentiation was achieved on spiked samples of different species with different mixing ratios. The proposed ConvNet deep learning-assisted data analysis method of IFC exhibits advantages in analyzing a huge number of data sets with capacity for extracting predominant features within multicomponent information and will bring about progress and advances in the fields of both biosensing and data analysis.