Complex Classification Research Articles

An invitation to the promise constraint satisfaction problem

The study of the complexity of the constraint satisfaction problem (CSP), centred around the Feder-Vardi Dichotomy Conjecture, has been very prominent in the last two decades. After a long concerted effort and many partial results, the Dichotomy Conjecture has been proved in 2017 independently by Bulatov and Zhuk. At about the same time, a vast generalisation of CSP, called promise CSP, has started to gain prominence. In this survey, we explain the importance of promise CSP and highlight many new very interesting features that the study of promise CSP has brought to light. The complexity classification quest for the promise CSP is wide open, and we argue that, despite the promise CSP being more general, this quest is rather more accessible to a wide range of researchers than the dichotomy-led study of the CSP has been.

TYPOLOGICAL CHARACTERISTICS OF MULTIFUNCTIONAL SPORTS COMPLEXES (Part I)

The article gives a classification of sports and modern multifunctional sports complexes. The terminology of the concept of multifunctional sports complex has been unified. The analysis of modern domestic and foreign multifunctional sports complexes is carried out that allows to define a number of obvious tendencies of development of the sports environment. The construction of sports facilities is included in the general plans of cities in the form of microdistricts, districts and cities. A study of the development of modern sports complexes. Trends in the development of the spatial structure of the sports environment have been identified. The sports infrastructure of Ukraine is analyzed for the possibility of holding international sports competitions. It is determined that according to the existing norms and rules, all multifunctional sports facilities can be divided into main, auxiliary and devices-devices for spectators. The classification of multifunctional sports complexes by location in the structure of the city is given. Multifunctional buildings are, if possible, adapted for various sports. In order to determine the role and place of multifunctional sports complexes in the structure of the city, sports centers available in natural and urban areas were analyzed. The basic requirements to be presented to modern sports complexes at designing are defined. The work consists of two parts. In the first part, the development of the sports environment is studied and the classification of modern sports complexes is given. In the second part, the features of the typological characteristics of multifunctional sports complexes are formulated.

Randomized Trial on the Effects of a Group EMDR Intervention on Narrative Complexity and Specificity of Autobiographical Memories: A Path Analytic and Supervised Machine-Learning Study.

Narratives of autobiographical memories may be impaired by adverse childhood experiences, generating narrative fragmentation and increased levels of perceived distress. Eye movement desensitization and reprocessing (EMDR) proved to be an effective treatment to overcome traumatic experiences and to promote coherent autobiographical narratives. However, the specific mechanisms by which EMDR promotes narrative coherence remains largely unknown. We conducted a randomized controlled pilot trial (ClinicalTrials.gov Identifier NCT05319002) in a non-clinical sample of 27 children recruited in a primary school. Participants were randomly assigned to the experimental and control groups. The experimental group underwent a three-week group EMDR intervention. Subjective unit of distress (SUD), validity of cognition (VoC), classification of autobiographical memories, narrative complexity and specificity were assessed before and after the group EMDR intervention. The group EMDR intervention was able to improve SUD and VoC scales, narrative complexity and specificity, and promoted the classification of autobiographical memories as relational. The path analysis showed that SUD was able to predict VoC and narrative specificity, which, in turn, was able to predict both narrative complexity and the classification of autobiographical memories as relational. Machine-learning analysis showed that random tree classifier outperformed all other models by achieving a 93.33% accuracy. Clinical implications are discussed.

Clinical Routine and Necessary Advances in Soft Tissue Tumor Imaging Based on the ESSR Guideline: Initial Findings.

Soft tissue sarcomas are malignant diseases with a complex classification and various histological subtypes, mostly clinically inconspicuous appearance, and a rare occurrence. To ensure safe patient care, the European Society of Musculoskeletal Radiology (ESSR) issued a guideline for diagnostic imaging of soft tissue tumors in adults in 2015. In this study, we investigated whether implementation of these guidelines resulted in improved MRI protocol and report quality in patients with soft tissue sarcomas in our cancer center. All cases of histologically confirmed soft tissue sarcomas that were treated at our study center from 2006 to 2018 were evaluated retrospectively. The radiological reports were examined for their compliance with the recommendations of the ESSR. Patients were divided into two groups, before and after the introduction of the 2015 ESSR guidelines. In total, 103 cases of histologically confirmed sarcomas were studied. The distribution of, age, gender, number of subjects, performing radiology, and MRI indication on both groups did not show any significant differences. Only using the required MRI sequences showed a significant improvement after the introduction of the guidelines (p = 0.048). All other criteria, especially the requirements for the report of findings, showed no improvement. The guidelines of the European Society for Musculoskeletal Radiology are not regularly followed, and their establishment did not consistently improve MRI quality in our study group. This poses a risk for incorrect or delayed diagnosis and, ultimately, therapy of soft tissue tumors. However, this study is the first of its kind and involves a limited collective. A European-wide multicenter study would be appreciated to confirm these results.

Evaluation of Japanese Teaching Quality Based on Deep Neural Network

The 21st century is an era of rapid development of information and frequent international exchanges, and Japanese language teaching has received increasing attention. Because of this, colleges and universities are now focused on improving the quality of Japanese education, both now and in the future. We need to boost the whole management of teaching quality, notably the assessment of instructors’ teaching quality, in order to improve teaching quality. However, because a number of factors influence the quality of instruction, and each factor’s weight varies, the evaluation results are difficult to express in a mathematical analytical formula, resulting in a complex nonlinear classification problem that traditional classification methods cannot solve well. As a new technology, as a result of the artificial neural networks (ANNs) fundamental qualities, it has been extensively applied in different evaluation issues for pattern recognition, nonlinear classification, and other research. This subject introduces the optimized deep neural network theory into Japanese teaching quality evaluation and completes the following work: (1) the algorithm of discrete Hopfield neural network is introduced in detail, and the neural network theory is introduced into teaching evaluation. (2) Then, based on the evaluation data of teachers’ teaching quality in a school, a large number of simulation experiments and training were carried out, and a neural network model for evaluation of teachers’ teaching effect was constructed and designed. Experiments reveal that the neural network model proposed in this paper is a nonlinear mapping method, which increases the evaluation’s dependability and makes the outcomes more effective and objective.

Poland syndrome: a new classification system based on a retrospective analysis of 74 cases.

III.

Hopping-Transport Mechanism for Reconfigurable Logic in Disordered Dopant Networks

The rise of artificial intelligence has led to an explosion in demand for computing power---demand that soon will be insatiable using conventional CMOS-based computing hardware. Thus there is a worldwide quest for unconventional hardware that can replace or complement CMOS-based gear. This study contributes to that quest simulations of how reconfigurable logic can be realized in disordered dopant networks in a semiconductor. This paper reveals the operating principles, based on variable-range hopping of charges between randomly located dopants, that underlie the performance of complex classification tasks by these networks.

Current Global Land Systems Classifications: Comparison of Methods and Outputs

The anthropogenic impact on the functioning of natural systems and the concept of Anthropocene as a period of the human domination of the Earth has been widely discussed in literature in the past few decades. Consequently, several land systems classifications have been developed on a global scale to capture the diversity, intensity, and spatial distribution of the human suppression of natural stratification. This review presents the comparison of the most widely used complex global classifications, incorporating both natural conditions and the human influence on nature. Methods, input data, the number and type of output categories as well as their geographical extent and distribution are described and compared. The review will help potential users to find differences between available classifications and choose the right one for a particular use.

Land and Property Complex: its Essence and Approaches to Management

The article studies the essence of the economic category ‘land-property complex’ in respect to a business entity – enterprise, analyzes theoretical approaches to its management and shows approbation of methodology of strategic management at the agricultural enterprise. Differences between property and land-property complexes were identified and classification of land-property complexes was provided for management. By using PEST-analysis the author estimates the degree of macro-environment impact on management of land-property complex at the agricultural enterprise. The author developed the forecast economic-mathematic model of net present value cost of the land-property complex element, which was adjusted for specificity of agricultural enterprise work. Understanding of the essence of the notion ‘land-property complex’, its characteristic features and elements and adequate application of tools of strategic management allow us to resolve problems in this field in accounting and governance. In this research the author used such academic methods and cognition principles as analysis, synthesis, systemacity, comparison, economic and mathematic modeling.

Radiomics and artificial intelligence in malignant uterine body cancers: Protocol for a systematic review.

Uterine body cancers (UBC) are represented by endometrial carcinoma (EC) and uterine sarcoma (USa). The clinical management of both is hindered by the complex classification of patients into risk classes. This problem could be simplified through the development of predictive models aimed at treatment tailoring based on tumor and patient characteristics. In this context, radiomics represents a method of extracting quantitative data from images in order to non-invasively acquire tumor biological and genetic information and to predict response to treatments and prognosis. Furthermore, artificial intelligence (AI) methods are an emerging field of translational research, with the aim of managing the amount of data provided by the various -omics, including radiomics, through the process of machine learning, in order to promote precision medicine. The aim of this protocol for systematic review is to provide an overview of radiomics and AI studies on UBCs. A systematic review will be conducted using PubMed, Scopus, and the Cochrane Library to collect papers analyzing the impact of radiomics and AI on UBCs diagnosis, prognostic classification, and clinical outcomes. The PICO strategy will be used to formulate the research questions: What is the impact of radiomics and AI on UBCs on diagnosis, prognosis, and clinical results? How could radiomics or AI improve the differential diagnosis between sarcoma and fibroids? Does Radiomics or AI have a predictive role on UBCs response to treatments? Three authors will independently screen articles at title and abstract level based on the eligibility criteria. The risk of bias and quality of the cohort studies, case series, and case reports will be based on the QUADAS 2 quality assessment tools. PROSPERO registration number: CRD42021253535.

(Retracted) Image recognition technology of crop diseases based on neural network model fusion

The neural network is an important model of machine learning that utilizes a hierarchical structure composed of multiple neurons to fit complex nonlinear classification surfaces. Deep learning is a new direction in the field of machine learning. Learning the inherent laws and representation levels of sample data, the information obtained during the learning process plays a great role in interpreting data such as images and speech. The crop disease image recognition technology based on the neural network model is studied to develop a system with an improved accuracy. First, we analyze the background and significance of image recognition technology applied to crops; then, on the basis of understanding the general background, the pattern recognition method based on the image is deeply studied; and, finally, the experimental design of crop disease image recognition based on convolutional neural network (CNN) is tested. The experimental results show that, in the problem of disease identification, compared with a single CNN model, the proposed model further improves the highest verification accuracy. The integration of DenseNet and the visual geometry group model has the best effect on cucumber and rice disease identification, and the verification accuracy can reach 96.24%.

A 107-gene Nanostring assay effectively characterizes complex multiomic gastric cancer molecular classification in a translational patient-derived organoid model.

4049 Background: Multi-omics profiling of gastric cancer (GC) has produced numerous molecular classification systems. However, widespread clinical implementation and testing of molecular subtypes are currently limited. Here, we develop, validate and implement a custom Nanostring assay capable of allocating GC molecular subtypes to clinical specimens in a translational patient-derived organoid model. Methods: Using publicly available whole-transcriptome data, machine learning models were developed to predict GC molecular subtypes from 376 Cancer Genome Atlas (TCGA) and 1797 Tumour Microenvironment Score (TME) patients. Models were generated using feature selection with 10-fold nested cross-validation. GC biopsies from 10 local patients were preserved in paraffin (tumour) and established as an organoid culture (organoid). Gene expression was measured using Nanostring. The allocation of molecular subtypes was internally and externally validated using gold-standard reference features in public databases comprising 2202 GC patients and 10 tumour-organoid pairs, respectively. We evaluated the concordance of tumour-organoid molecular subtypes and explored the correlation between subtype scores and in-vitro chemotherapy response. Results: Classification models for TCGA (57 genes) and TME (50 genes) predicted subtypes with an accuracy ± standard deviation of 89.46% ± 0.04 and 89.33% ± 0.02, respectively. Subtype assignment of microsatellite instability (MSI) in reference to capillary electrophoresis was found to have 99.3% [95% CI 97.4-99.9, n = 277] internal and 100% [95% CI 83.2-100, n = 20] external accuracy. In reference to Epstein-Barr Virus (EBV) in-situ hybridization, EBV type internal and external accuracy was 98.7% [95% CI 97.4-99.5, n = 552] and 100% [95% CI 83.2-100, n = 20], respectively. TCGA Genomically Stable (GS) scores followed a previously characterized enrichment of diffuse-type histology compared to intestinal-type in internal and external cohorts (Dunn’s Test, p < 0.0001 and p < 0.05, n = 1471 and n = 15, respectively). Statistically similar subtype scores (Paired Wilcoxon, p > 0.05) were found for tumour-organoid pairs. Discordance occurred in three tumour-organoid pairs. In-vitro Drug Sensitivity Score was not statistically efficacious in any molecular subtype, but Pearson correlation identified increasing efficacy with increasing EBV and MSI scores. Conclusions: Patient-derived organoids generally recapitulate the molecular subtype of parent tumours; however, in specific cases, subtype discordance occurs. Although additional external validation is required, our 107 gene assay effectively captures multi-omics classification systems in GC and allows future inquiry into the prognostic and therapeutic implications of these molecular subtypes.

Fusion genes as diagnostic and predictive biomarkers for tumor

Structural variants, including chromosomal rearrangements and translocations, may result in the formation of fusion genes. Glioma is the most frequent brain tumor among adults. Due to complex tumor classifications, characterization of recurrence, inadequate sample size and controversial mechanisms of tumor malignancy, clinical strategies have not been developed for almost 30 years. Fusion gene is one of the strong driver events in glioma tumorigenesis and has provided fundamental insights into the disease mechanisms. This review elucidates the literature on the discovery of fusion genes, the development of detection techniques, and their clinical implementations. In conclusion, fusion genes are important diagnostic and predictive biomarkers for brain tumors.

Impact of CT Enhanced Syntax Score (Ctesx) on Grading Complexity of Coronary Artery Disease

Purpose: The SYNTAX Score is a crucial tool to grade angiographic complexity and risk stratification of patients for revascularization. Excessive coronary calcification can lead to adverse outcomes after percutaneous coronary intervention (PCI), and it represents an essential part of SYNTAX score calculation. The sensitivity in detecting any CAC by CT is much higher than in ICA. Our purpose is to evaluate the impact of CT-enhanced calcium detection on SYNTAX score in the evaluation of complex coronary anatomy.
 Methodology: This study was conducted at Beni-suef University Hospital and National Heart Institute from December 2019 to January 2021. Fifty consecutive male patients were enrolled, Their Coronary CT Angiography showed significant MVD; then, conventional coronary angiography was done. The additive value, rather than the comparative value, of adding Calcification from CT to CA syntax score (CT enhanced SYNTAX (CTeSx)), was assessed regarding its impact on the classification of complexity.
 Findings: Using a CTeSX Vs CA Sx results in Lower Possibilities of Low Syntax group (39% vs 24%) (P-value <0.001) and a higher Possibilities of High Syntax score (23% vs 52% P-value 0.004
 Recommendation: CTeSX results in a significant change in the percentage of all complexity categories with a higher possibility to categorize complex anatomy.

The effects on classifier performance of 2D discrete wavelet transform analysis and whale optimization algorithm for recognition of power quality disturbances

Power Quality (PQ) is becoming more and more important day by day in the electric network. Signal processing, pattern recognition and machine learning are increasingly being studied for the automatic recognition of any disturbances that may occur during the generation, transmission, and distribution of electricity. There are three main steps to identify the PQ disturbances. These include the use of signal processing methods to calculate the features representing the disturbances, the selection of those that are more useful than these feature sets to prevent the creation of a complex classification model, the creating a classification model that recognizes multiple classes using the selected feature subsets. In this study, one-dimensional (1D) PQ disturbances signals are transformed into two-dimensional (2D) signals, 2D discrete wavelet transforms (2D-DWT) are used to extract the features. The features are extracted by using the wavelet families such as Daubechies, Biorthogonal, Symlets, Coiflets and Fejer-Korovkin in 2D-DWT to analyze PQ disturbances. Whale Optimization Algorithm (WOA) and k-nearest neighbor (KNN) classifier determine the feature subsets. Then, WOA and k nearest neighbor (KNN) classifier are used to determine the feature group. By using KNN and Support Vector Machines (SVM) classification methods, Classifier models that distinguish PQ disturbances are formed. The main aim of the study is to determine the features derived from 2D wavelet coefficients for different wavelet families and to determine which of them has a better classification performance to distinguish PQ disturbances signals. At the same time, different classification methods are simulated and a model which can classify PQ disturbances signals with high performance is created. Also, the generated models are analysed for their performance in terms of different noise levels (40 dB, 30 dB, 20 dB). The result of this simulation study shows that the model developed to classify PQ disturbances is superior to conventional models and other 2D signal processing methods in the literature. In addition, it was concluded that the proposed method can cope better with noisy signals by low computational complexity and higher classification rate.

Risk-stratified posthepatectomy pathways based upon the Kawaguchi–Gayet complexity classification and impact on length of stay

BackgroundThe Kawaguchi–Gayet classification is a validated system to stratify open liver resections by complexity and postoperative complications. We hypothesized that Kawaguchi–Gayet classification could be used to create and implement risk-stratified posthepatectomy pathways to reduce length of stay and variation in care. MethodsClinicopathologic data from hepatectomy patients (1/2017–6/2020) were abstracted from a prospective database. All open hepatectomies were assigned to groups based on 2 levels of Kawaguchi–Gayet classification, and corresponding risk-stratified posthepatectomy pathways were created to decrease length of stay by 1day compared to patients who were historically treated without a pathway: low–intermediate risk (open Kawaguchi–Gayet I/II) and high risk (open Kawaguchi–Gayet III). Outcomes were compared between periods before ("PRE"; 1/1/2017–9/30/2019) and after ("POST"; 10/1/2019–6/30/2020) implementation. ResultsAmong 487 open hepatectomies (PRE: 374, POST: 113), 55.0% (n=268) were low–intermediate risk and 45.0% (n=219) were high risk. Major complications were similar PRE/POST: low–intermediate risk (PRE: 7.8%, POST: 9.4%, P=.681) and high risk (PRE: 18.9%, POST 10.0%, P=0.139). Risk-stratified posthepatectomy pathway implementation reduced median length of stay for both low–intermediate risk (4 to 3.5days, P=.009) and high risk (5 to 4days, P=0.022) patients. Risk-stratified posthepatectomy pathways decreased length of stay variation, reflected in mean and standard deviation for all patients (PRE 5.5±7.5 vs POST 4.4±2.8days). There was no difference in 90-day readmission rates between PRE (12.6%) and POST (8.8%) periods (P=.278). ConclusionThe creation and implementation of risk-stratified posthepatectomy pathways reduced length of stay without increasing readmissions after hepatectomy. These generalizable risk-stratified posthepatectomy pathways preoperatively stratify patients a priori into pathways for individualized preoperative discussions on realistic postoperative complications and length of stay expectations.

Multiscale depth of anaesthesia prediction for surgery using frontal cortex electroencephalography.

Hypnotic and sedative anaesthetic agents are employed during multiple medical interventions to prevent patient awareness. Careful titration of agent dosing is required to avoid negative side effects; the accuracy thereof may be improved by Depth of Anaesthesia Monitoring. This work investigates the potential of a patient specific depth monitoring prediction using electroencephalography recorded neural oscillation from the frontal lobe of 10 patients during sedation, where a comparison of the prediction accuracy was made across five different approaches to post‐processing; Noise Assisted‐Empirical Mode Decomposition, the Raw Signal, Linear Series Decomposition Learner, Deep Wavelet Scattering and Deep Learning features. These methods towards anaesthesia depth prediction were investigated using the Bispectral Index as ground truth, where it was seen that the Raw Signal, enhanced feature set and a low complexity classification model (Linear Discriminant Analysis) provided the best classification accuracy, in the region of 85.65 % ±10.23 % across the 10 subjects. Subsequent work in this area would now build on these results and validate the best performing methods on a wider cohort of patients, investigate means of continuous DoA estimation using regressions, and also feature optimisation exercises in order to further streamline and reduce the computation complexity of the designed model.

Development of the Revision Hip Complexity Classification using a modified Delphi technique.

AimsThe aim of this modified Delphi process was to create a structured Revision Hip Complexity Classification (RHCC) which can be used as a tool to help direct multidisciplinary team (MDT) discussions of complex cases in local or regional revision networks.MethodsThe RHCC was developed with the help of a steering group and an invitation through the British Hip Society (BHS) to members to apply, forming an expert panel of 35. We ran a mixed-method modified Delphi process (three rounds of questionnaires and one virtual meeting). Round 1 consisted of identifying the factors that govern the decision-making and complexities, with weighting given to factors considered most important by experts. Participants were asked to identify classification systems where relevant. Rounds 2 and 3 focused on grouping each factor into H1, H2, or H3, creating a hierarchy of complexity. This was followed by a virtual meeting in an attempt to achieve consensus on the factors which had not achieved consensus in preceding rounds.ResultsThe expert group achieved strong consensus in 32 out of 36 factors following the Delphi process. The RHCC used the existing Paprosky (acetabulum and femur), Unified Classification System, and American Society of Anesthesiologists (ASA) classification systems. Patients with ASA grade III/IV are recognized with a qualifier of an asterisk added to the final classification. The classification has good intraobserver and interobserver reliability with Kappa values of 0.88 to 0.92 and 0.77 to 0.85, respectively.ConclusionThe RHCC has been developed through a modified Delphi technique. RHCC will provide a framework to allow discussion of complex cases as part of a local or regional hip revision MDT. We believe that adoption of the RHCC will provide a comprehensive and reproducible method to describe each patient’s case with regard to surgical complexity, in addition to medical comorbidities that may influence their management.Cite this article: Bone Jt Open 2022;3(5):423–431.

Refinement of the Intensive Care Unit Complexity Assessment and Monitoring to Ensure Optimal Outcomes III Acuity Tool.

The Intensive Care Unit (ICU) Complexity Assessment and Monitoring to Ensure Optimal Outcomes (CAMEO) acuity tool quantifies patient acuity in terms of nursing cognitive workload complexity. The aim of this study was to refine the ICU CAMEO II acuity tool. An expert panel of nursing staff from 4 pediatric ICUs convened to refine the CAMEO II across a large, freestanding children's hospital in the United States. This study used a modified Delphi technique. Through a series of 4 Delphi rounds, the expert panel identified Domains of Care and nursing care items that were suitable to be collapsed or bundled. The number of Domains of Care decreased from 18 to 10. Each of the expert panel members then completed the ICU CAMEO II tool and the newly revised tool, ICU CAMEO III, on 5 to 10 patients. Sixty completed ICU CAMEO II tools, and ICU CAMEO III tools were available for comparison. The average difference of the 2 tools' total scores was 5 points (minimum, 4; maximum, 7). The level of agreement between the 2 tools by CAMEO Complexity Classification level (I-V) was 90%. The ICU CAMEO III acuity tool is a streamlined measure to describe and quantify the acuity of pediatric critical care nursing. Use of this acuity measure will support projection of staffing models, staffing assignments, and benchmarking across pediatric ICUs. Further research is underway to validate the CAMEO III for multisite use.

FPGA-Based BNN Architecture in Time Domain with Low Storage and Power Consumption

With the increasing demand for convolutional neural networks (CNNs) in many edge computing scenarios and resource-limited settings, researchers have made efforts to apply lightweight neural networks on hardware platforms. While binarized neural networks (BNNs) perform excellently in such tasks, many implementations still face challenges such as an imbalance between accuracy and computational complexity, as well as the requirement for low power and storage consumption. This paper first proposes a novel binary convolution structure based on the time domain to reduce resource and power consumption for the convolution process. Furthermore, through the joint design of binary convolution, batch normalization, and activation function in the time domain, we propose a full-BNN model and hardware architecture (Model I), which keeps the values of all intermediate results as binary (1 bit) to reduce storage requirements by 75%. At the same time, we propose a mixed-precision BNN structure (model II) based on the sensitivity of different layers of the network to the calculation accuracy; that is, the layer sensitive to the classification result uses fixed-point data, and the other layers use binary data in the time domain. This can achieve a balance between accuracy and computing resources. Lastly, we take the MNIST dataset as an example to test the above two models on the field-programmable gate array (FPGA) platform. The results show that the two models can be used as neural network acceleration units with low storage requirements and low power consumption for classification tasks under the condition that the accuracy decline is small. The joint design method in the time domain may further inspire other computing architectures. In addition, the design of Model II has certain reference significance for the design of more complex classification tasks.