Machine Learning Models Research Articles

Chemometric approaches for discriminating manufacturers of Korean handmade paper using infrared spectroscopy

The objective of this study was to identify the manufacturer of Hanji, Korean handmade paper widely used in conservation science. To achieve this, machine learning models utilizing attenuated total reflectance–infrared spectroscopy (ATR–IR) were developed to assess the robustness and effectiveness of the computed models. Principal component analysis (PCA), partial least squares–discriminant analysis (PLS–DA), decision tree (DT), and k-NN models were constructed using IR spectral data, with the spectral region between 1800 and 1500 cm⁻1 identified as the critical input variable through Variable Importance in Projection (VIP) scores. The transformation of the obtained spectra into second derivative spectra proved beneficial in this key spectral region, leading to significant improvements in model performance. Additionally, the application of DBSCAN for outlier detection was effective in refining the dataset, further enhancing the performance of the models. Specifically, the k-NN model, when applied to the selected variables and preprocessed with the second derivative transformation, achieved an F1 score of 0.92. These findings underscore the importance of focusing on the 1800–1500 cm⁻1 spectral range and applying outlier detection techniques, such as DBSCAN, to enhance the robustness and accuracy of the Hanji classification models by eliminating the influence of atypical data points.

A robust deep learning approach for identification of RNA 5-methyluridine sites

RNA 5-methyluridine (m5U) sites play a significant role in understanding RNA modifications, which influence numerous biological processes such as gene expression and cellular functioning. Consequently, the identification of m5U sites can play a vital role in the integrity, structure, and function of RNA molecules. Therefore, this study introduces GRUpred-m5U, a novel deep learning-based framework based on a gated recurrent unit in mature RNA and full transcript RNA datasets. We used three descriptor groups: nucleic acid composition, pseudo nucleic acid composition, and physicochemical properties, which include five feature extraction methods ENAC, Kmer, DPCP, DPCP type 2, and PseDNC. Initially, we aggregated all the feature extraction methods and created a new merged set. Three hybrid models were developed employing deep-learning methods and evaluated through 10-fold cross-validation with seven evaluation metrics. After a comprehensive evaluation, the GRUpred-m5U model outperformed the other applied models, obtaining 98.41% and 96.70% accuracy on the two datasets, respectively. To our knowledge, the proposed model outperformed all the existing state-of-the-art technology. The proposed supervised machine learning model was evaluated using unsupervised machine learning techniques such as principal component analysis (PCA), and it was observed that the proposed method provided a valid performance for identifying m5U. Considering its multi-layered construction, the GRUpred-m5U model has tremendous potential for future applications in the biological industry. The model, which consisted of neurons processing complicated input, excelled at pattern recognition and produced reliable results. Despite its greater size, the model obtained accurate results, essential in detecting m5U.

Identifying prognostic factors in patients with cardiogenic shock undergoing VA-ECMO treatment: insights from ECLS-SHOCK using machine learning

Abstract Background/Introduction Venoarterial extracorporeal membrane oxygenation (VA-ECMO) has not been shown to reduce short-term mortality in patients with acute myocardial infarction complicated by cardiogenic shock. However, whether some patients might benefit from VA-ECMO remains to be cleared. Purpose To identify the impact of VA-ECMO therapy on outcomes by leveraging the potential of machine learning. Methods The ECLS-SHOCK trial is a randomized controlled trial where patients with acute myocardial infarction complicated by cardiogenic shock for whom early revascularization was planned were randomly assigned to receive early VA-ECMO plus usual medical treatment or usual medical treatment alone. The primary outcome was death from any cause at 30 days. Extreme gradient boosting (XGBoost) was used to develop a model for the prediction of 30-day all-cause mortality and to assess the influence of VA-ECMO on outcomes. Results Overall, 209 patients were randomized to the VA-ECMO and 208 to control. Baseline characteristics between groups were well balanced (median age 63 years, 81% male, 67% presenting with ST-segment elevation myocardial infarction). At 30-days 100 patients in the VA-ECMO and 102 in the control group were dead. The final XGBoost model contained 27 variables and the 10 most influential variables are shown in Fig. 1. On derivation the model predicted mortality with an AUC of 1.00 and on validation with an AUC of 0.80 (95% CI 0.71 to 0.89), with a sensitivity of 95% (95% CI 88% to 100%) and a specificity of 55% (40% to 69%), resulting in a positive predictive value of 65% and a negative predictive value of 92% at a Youden-derived cut-off. Importantly, neither the VA-ECMO therapy nor its complications were among the most influential factors of the model with a relative importance of 1.1% for VA-ECMO therapy and 0.25% for peripheral vascular complications and bleeding complications even being discarded by the model. Conclusion In patients with acute myocardial infarction complicated by cardiogenic shock the influence of VA-ECMO therapy on 30-day mortality, even when investigated with complex machine learning models, is negligibly low.

MemoriaNova: Optimizing Memory-Aware Model Inference for Edge Computing

In recent years, deploying deep learning models on edge devices has become pervasive, driven by the increasing demand for intelligent edge computing solutions across various industries. From industrial automation to intelligent surveillance and healthcare, edge devices are being leveraged for real-time analytics and decision-making. Existing methods face two challenges when deploying machine learning models on edge devices. The first challenge is handling the execution order of operators with a simple strategy, which can lead to a potential waste of memory resources when dealing with directed acyclic graph structure models. The second challenge is that they usually process operators of a model one by one to optimize the inference latency, which may lead to the optimization problem getting trapped in local optima. We present MemoriaNova, comprising BTSearch and GenEFlow, to solve these two problems. BTSearch is a graph state backtracking algorithm with efficient pruning and hashing strategies designed to minimize memory overhead during inference and enlarge latency optimization search space. GenEFlow, based on genetic algorithms, integrates latency modeling and memory constraints to optimize distributed inference latency. This innovative approach considers a comprehensive search space for model partitioning, ensuring robust and adaptable solutions. We implement BTSearch and GenEFlow and test them on eleven deep-learning models with different structures and scales. The results show that BTSearch can reach 12% memory optimization compared with the widely used random execution strategy. At the same time, GenEFlow reduces inference latency by 33.9% in distributed systems with four-edge devices.

Reinforcement Learning Algorithm for Optimising Durian Irrigation Systems: Maximising Growth and Water Efficiency

This study presents a Reinforcement Learning-based algorithm designed to optimise irrigation for Durio Zibethinus (i.e., durian) trees, aiming to maximise tree growth and reduce water usage. Traditional irrigation methods, as well as current machine learning models, often focus only on soil moisture and weather data, neglecting critical factors like actual tree growth. This study proposed a reinforcement learning irrigation (RL-Irr) algorithm incorporating tree growth stages, soil moisture, and weather conditions to determine precise irrigation needs. The algorithm was developed by calibrating the AQUACROP model using data from actual durian plantations where rain-fed irrigation (rain-fed) was practised. Daily irrigation volumes were calculated based on real-time soil moisture, weather forecasts, and weekly tree growth measurements. The reinforcement learning method was used to optimise irrigation schedules, with rewards based on soil moisture, tree growth, rainfall, and weather conditions. The algorithm was tested using AQUACROP simulations and compared against soil moisture balance irrigation (SMB-Irr) and rain-fed. The results showed that the RL-Irr reduced water use by up to 75 percent while maintaining tree growth. These findings suggest the algorithm could significantly improve water efficiency in durian farming, though real-world applications should consider potential model limitations.

A new fusion neural network model and credit card fraud identification.

Credit card fraud identification is an important issue in risk prevention and control for banks and financial institutions. In order to establish an efficient credit card fraud identification model, this article studied the relevant factors that affect fraud identification. A credit card fraud identification model based on neural networks was constructed, and in-depth discussions and research were conducted. First, the layers of neural networks were deepened to improve the prediction accuracy of the model; second, this paper increase the hidden layer width of the neural network to improve the prediction accuracy of the model. This article proposes a new fusion neural network model by combining deep neural networks and wide neural networks, and applies the model to credit card fraud identification. The characteristic of this model is that the accuracy of prediction and F1 score are relatively high. Finally, use the random gradient descent method to train the model. On the test set, the proposed method has an accuracy of 96.44% and an F1 value of 96.17%, demonstrating good fraud recognition performance. After comparison, the method proposed in this paper is superior to machine learning models, ensemble learning models, and deep learning models.

Exploring IRGs as a Biomarker of Pulmonary Hypertension Using Multiple Machine Learning Algorithms

Background: Pulmonary arterial hypertension (PAH) is a severe disease with poor prognosis and high mortality, lacking simple and sensitive diagnostic biomarkers in clinical practice. This study aims to identify novel diagnostic biomarkers for PAH using genomics research. Methods: We conducted a comprehensive analysis of a large transcriptome dataset, including PAH and inflammatory response genes (IRGs), integrated with 113 machine learning models to assess diagnostic potential. We developed a clinical diagnostic model based on hub genes, evaluating their effectiveness through calibration curves, clinical decision curves, and ROC curves. An animal model of PAH was also established to validate hub gene expression patterns. Results: Among the 113 machine learning algorithms, the Lasso + LDA model achieved the highest AUC of 0.741. Differential expression profiles of hub genes CTGF, DDR2, FGFR2, MYH10, and YAP1 were observed between the PAH and normal control groups. A diagnostic model utilizing these hub genes was developed, showing high accuracy with an AUC of 0.87. MYH10 demonstrated the most favorable diagnostic performance with an AUC of 0.8. Animal experiments confirmed the differential expression of CTGF, DDR2, FGFR2, MYH10, and YAP1 between the PAH and control groups (p < 0.05); Conclusions: We successfully established a diagnostic model for PAH using IRGs, demonstrating excellent diagnostic performance. CTGF, DDR2, FGFR2, MYH10, and YAP1 may serve as novel molecular diagnostic markers for PAH.

Antibonding valence states induce low lattice thermal conductivity in metal halide semiconductors

Reduction of phonon mediated thermal transport properties, i.e., lattice thermal conductivity (κL), of semiconductors can strongly affect the performance of thermoelectrics and optoelectronics. Although extrinsic routes to reduce κL have been achieved through selective scattering of phonons via doping, alloying, and hierarchical nano-structuring, semiconductors with intrinsically low κL have recently gained widespread attention due to their ability to decouple electronic and phonon transports. While innate low κL in crystalline semiconductors is a desired requirement to achieve high performance thermoelectrics, the solar upconversion efficiency of photovoltaics based on metal halide perovskites (MHPs) have been shown to increase due to their ultralow κL through the hot-phonon bottleneck effect. Therefore, understanding the microscopic mechanisms underlying ultralow κL in crystalline semiconductors is extremely important. Several structural factors that are intrinsic to a material have been shown to strongly influence the reduction of κL. Among them, the presence of rattling atoms, lone-pair electrons, and large lattice anharmonicity have been widely studied. Here, we bring out yet another largely unexplored intrinsic characteristic of materials related to the filled antibonding valence states (AVS) near the Fermi level, which are shown to induce low κL in crystalline compounds. We focus our review on an emerging class of compounds–metal halide semiconductors including MHPs and investigate the interplay between structures, chemical bonding and κL, carefully curating from literature a list of 33 compounds having different structure dimensionality with known κL. We established a universal connection between the elastic moduli, speeds of sound, and κL with the presence of AVS just below the Fermi level. We found that large peak in the AVS correlates positively with lower values of elastic moduli, speeds of sound, and κL, providing antibonding states based design criteria of low-κL compounds. Furthermore, we discuss different synthesis strategies, which are crucial for experimental realization of ultralow κL through structure manipulation. Additionally, we outline how chemical bonding data can be utilized in machine learning models for predictive modeling of κL. We hope that our approach of understanding low-κL through the viewpoint of chemical bonding theory would encourage exploration of phonon transport properties in other families of materials having filled AVS that can provide further insights on the structure-bonding-property relationships aiding novel materials design approaches.

Personalized prediction model specific for sudden cardiac death in the population: a machine learning study based on EHR

Abstract Background Predicting sudden cardiac death (SCD) in the general population remains a significant challenge, particularly at the individual level. SCD and myocardial infarction (MI) share similar pattern of characteristics and risk factors, making their specific prediction highly difficult. As a result, the current prediction models exhibit low specificity. Methods To estimate the specific risk of SCD over three months, we trained a machine learning model on electronic health record (EHR) data representing 8,566,229 drug prescriptions and 801,352 of hospital diagnoses up to five years prior to SCD. The data were obtained from a French cohort of 12,338 SCD and from a cohort of 12,338 controls from 2011 to 2015. We then validated the results on two external cohorts: one temporal cohort in the same area between 2016 and 2020 with 11,620 SCD cases and 11,620 controls and one geographical cohort from the USA with 892 SCD cases and 892 controls from 2013 to 2021. In order to address the specificity of our prediction model for SCD, SCD cases collected between 2016 and 2020 were also matched with a group of 35 000 controls who had myocardial infarction, with no SCD, following the same protocol as for the controls issued from the general population. Findings: The analysis reveals three distinct patterns of distribution for SCD cases, controls, and MI cases. We observed a semi linear increase of SCD cases over the deciles of predicted risk, and our model detected 24% (2,908) of SCD cases with a predicted risk exceeding 90% (highest decile). We observed the opposite among the controls that decreased linearly with increasing deciles of predicted risk: most of the controls were accurately identified in the lowest deciles, and there were only 214 controls (2%) in the highest decile. Interestingly, we observed a relatively stable distribution of MI cases across the deciles, with 10% of MI cases in the highest decile (Figure 1). Interpretation: We developed and validated a personalized prediction model that is able to identify subjects at high-risk for SCD specifically. Figure 1: Histogram of predicted risks for sudden cardiac death cases, controls and myocardial infarction cases in the validation cohort.Figure 1

Integrated Analysis of Gut microbiome, Inflammation and Neuroimaging Features Supports the Role of Microbiome-Gut-Brain Crosstalk in Schizophrenia

Abstract Background and Hypothesis Gut microbiota has been implicated in the pathogenesis of schizophrenia (SZ) and relevant changes in the brain, but the underlying mechanism remains elusive. This study aims to investigate the microbiota-gut-brain crosstalk centered on peripheral inflammation in SZ patients. Study Design We recruited a cohort of 182 SZ patients and 120 healthy controls (HC). Multi-omics data, including fecal 16S rRNA, cytokine data, and neuroimaging data, were collected and synthesized for analysis. Multi-omics correlations and mediation analyses were utilized to determine the associations of gut microbiome with inflammatory cytokines and neuroimaging characteristics. Additionally, machine learning models for effective SZ diagnosis were separately generated based on gut microbial and neuroimaging data. Study Results Gut microbial dysbiosis, characterized by a decrease in butyrate-producing bacteria and an increase in pro-inflammatory bacteria, has been identified in SZ patients. These key microbial taxa were associated with increased inflammatory cytokines, potentially through mediating lipid metabolic pathways such as steroid biosynthesis and linoleic acid metabolism. Further analysis revealed altered microbial genera to be correlated with disrupted gray matter volume and regional homogeneity in SZ patients. Importantly, certain inflammatory cytokines mediated the relationship between the SZ-enriched genus Succinivibrio and aberrant activity of anterior cingulate cortex and left inferior temporal gyrus in the SZ group. Moreover, the classification model based on gut microbial data showed comparable efficacy to the model based on brain functional signatures in SZ diagnosis. Conclusions This study presents evidence for the dysregulated microbiota-gut-brain axis in SZ and emphasizes the central role of peripheral inflammation.

Gut microbiota predict retinopathy in patients with diabetes: A longitudinal cohort study

The gut microbiota has emerged as an independent risk factor for diabetes and its complications. This research aimed to delve into the intricate relationship between the gut microbiome and diabetic retinopathy (DR) through a dual approach of cross-sectional and prospective cohort studies. In our cross-sectional study cross-sectional investigation involving ninety-nine individuals with diabetes, distinct microbial signatures associated with DR were identified. Specifically, gut microbiome profiling revealed decreased levels of Butyricicoccus and Ruminococcus torques group, alongside upregulated methanogenesis pathways among DR patients. These individuals concurrently exhibited lower concentrations of short-chain fatty acids in their plasma. Leveraging machine learning models, including random forest classifiers, we constructed a panel of microbial genera and genes that robustly differentiated DR cases. Importantly, these genera also demonstrated significant correlations with dietary patterns and the molecular profiles of peripheral blood mononuclear cells. Building upon these findings, our prospective cohort study followed 62 diabetes patients over a 2-year period to assess the predictive value of these microbial markers. The results underlined the panel’s efficacy in predicting DR incidence. By stratifying patients based on the predictive genera and metabolites identified in the cross-sectional phase, we established significant associations between reduced levels of Butyricicoccus, plasma acetate, and increased susceptibility to DR. This investigation not only deepens our understanding of how gut microbiota influences DR but also underscores the potential of microbial markers as early indicators of disease risk. These insights hold promise for developing targeted interventions aimed at mitigating the impact of diabetic complications.Key points• Microbial signatures are differed in diabetic patients with and without retinopathy• DR-related taxa are linked to dietary habits and transcriptomic profiles• Lower abundances of Butyricicoccus and acetate were prospectively associated with DR

Early inflammation as a footprint of increased mortality risk in infants living with HIV from three African countries

In this work our aim was to identify early biomarkers in plasma samples associated with mortality in children with perinatal HIV treated early in life, to potentially inform early intervention targeting this vulnerable group. 20/215 children (9.3%) with perinatal HIV, enrolled within 3 months of age died prematurely within the first year of the study, despite early ART initiation. Using a propensity score, we selected 40 alive study participants having similar clinical and virological records compared to the deceased group. 13 HIV unexposed (HU) healthy children were additionally used as controls. Baseline plasma samples were analyzed using a targeted proteomic approach, and to assess pathogen-associated and damage-associated molecular patterns (PAMPs, DAMPs) levels. Data from deceased participants were compared to both control groups, with multivariate logistic regression models used to evaluate the association between mortality and plasma proteins. We developed a machine learning model to predict mortality risk, finding that IL-6 and CXCL11 not only were higher in deceased children than Matched-children with HIV (p < 0.001 and p = 0.0034) but also predictive of mortality (accuracy of 77%); levels of PAMPs were higher in deceased children (p = 0.0016). Thus, measuring early inflammatory biomarkers, particularly IL-6, could help mortality risk prediction and potentially guide targeted interventions.

An ensemble machine learning approach for predicting heart failure risk in acute coronary syndrome patients: heartguardai

Abstract Background Heart failure (HF) constitutes a major post-event morbidity in patients experiencing Acute Coronary Syndrome (ACS), underscoring the imperative for precise risk stratification methodologies. Traditional prognostic models, although beneficial, are limited by their reliance on a narrow scope of variables, thereby overlooking the comprehensive analytical potential afforded by extensive clinical datasets. Aims To develop "HeartGuardAI," an ensemble machine learning model that predicts HF risk in ACS patients. By incorporating a wide array of variables including ejection fraction (EF), glomerular filtration rate (GFR), body mas index(BMI), male sex, this model seeks to offer a more practical tool for clinical decision-making. Methods We queried ACTION-ACS a large-scale multicenter database of 3335 patients at 10-year follow-up. We employed a feature selection process informed by Random Forest Classifier to identify the most predictive variables. An ensemble model combining the strengths of Random Forest and Gradient Boosting methodologies was then developed. The model's predictive performance was validated through Stratified K-Fold cross-validation, focusing on Area Under the Curve (AUC) and confidence intervals to assess its efficacy. The model's predictive performance was validated through Stratified K-Fold cross-validation, focusing on Area Under the Curve (AUC) and confidence intervals to assess its efficacy. Results Demonstrating a mean AUC of 0.76 ± 0.036, "HeartGuardAI" showed a high capability to predict HF risk in post-ACS patients and emphasizes the significance of incorporating multifactorial clinical assessments Conclusions HF risk can be predicted by a handy tool developed with advanced machine learning methods. Future studies are needed to prospectively validate this algorithm.Graphic 1Graphic 2

Artificial intelligence in cardiology: a machine learning model for supraventricular tachycardia discrimination

Abstract Background Distinguishing atrioventricular nodal reentrant tachycardia (AVNRT) from orthodromic atrioventricular reentrant tachycardia (AVRT) presents a diagnostic challenge, with clinical and electrocardiogram (ECG) features playing a crucial role in differential diagnosis. Purpose The aim of this study was to devise and validate a predictive algorithm that utilises basic clinical and ECG parameters to accurately classify AVNRT versus AVRT. Furthermore, this algorithm was integrated into a user-friendly, web-based tool to support clinical decision-making. Methods We conducted a prospective analysis of 704 patients (median age: 53.6 years; 59.7% female) with electrophysiologically confirmed AVNRT (n=535) or AVRT (n=169). A comprehensive evaluation was performed using a structured questionnaire tailored to arrhythmia-specific symptoms. The study utilised advanced machine learning techniques for classification, employing feature selection methods to enhance predictive accuracy. The analysis included 19 demographic, clinical, and ECG variables, adopting an 80-20% split for training and internal validation. The algorithm's performance was evaluated against a model comprising all variables (Full Model), expert ECG interpretations (ExpertECG), objective ECG criteria (ECG model), and a baseline dummy classifier, with efficacy assessed by the Area Under the Receiver Operating Characteristic curve (AUC) and the Area Under the Precision-Recall Curve (AUPRC). Results The optimised model, utilising the XGBoost algorithm with six critical variables (initial age at first episode, biological sex, presence of neck palpitations, pseudoR wave in lead V1/pseudoS wave in inferior leads, visibility of retrograde P-waves, and tachycardia cycle length), exhibited superior performance, achieving an AUC of 0.916 and an AUPRC of 0.823. This model significantly outperformed other comparative models in terms of predictive accuracy (Figure 1). Implementation of the model into a web-based interface was accomplished using Shiny, enhancing its accessibility for clinical application (Figure 2). Conclusions This study illustrates the efficacy of employing machine learning algorithms to utilise bedside clinical and ECG variables for accurately differentiating between AVNRT and AVRT. The creation of a web-based calculator marks a significant advancement, empowering clinicians, especially non-specialists, with a potent tool for providing personalised diagnostic insights. Future research should focus on external validation to ascertain the model's generalisability and its applicability in broader clinical settings.Figure 1Figure 2

Importance and limits of cardiovascular risk factors on the prediction of SCD using machine learning approach

Abstract Background SCD is mainly due to lethal ventricular arrhythmias and occurs often among patients with ischemic heart disease. SCD and ischemic heart disease share a similar pattern of cardiovascular risk factors (CVRF). Commonly developed risk scores have been focused on cardiac-related risk factors. We aimed to investigate whether non-CVRF variables could enhance predictive performance beyond standard CVRF of SCD. Method We compared 2 different strategies for variable inclusion. First, we built a prediction model containing EHR data restricted to cardiovascular diseases and risk factors (CVD MR model) that occurred up to 5 years before SCD. We selected all available main cardiovascular variables as surrogate markers for coronary artery disease, stroke, diabetes, hypertension, smoking status, obesity, lipid disorders and chronic renal failure. On the other hand, our global approach (ALL MR model) includes all medical records codes, without any prior selection. To estimate the risk of SCD over three months, we trained a machine learning model on EHR data representing 8,566,229 drug prescriptions and 801,352 hospital diagnoses up to five years prior to SCD. The data were obtained from a cohort of 12,338 SCD in France and 12,338 controls from 2011 to 2015. We then validated the results on two external cohorts: one temporal in the same area between 2016 and 2020 with 11,620 SCD and 11,620 controls and one geographical from the USA with 892 SCD and 892 controls from 2013 to 2021. Results The CVD MR + ML model (CatBoost algorithm) yielded moderate performances with an AUC of 0.68 (95%CI 0.65-0.68), a sensitivity of 45%.(95%CI 42-47) and a specificity of 82% (95%CI 80-84). We found that the EHR model with all medical records (All MR + ML model) offered better performances. In the derivation cohort, this model achieved an AUC of 0.80 (95% CI: 0.78–0.82) with a sensitivity of 67% (95% CI 64-69), a specificity of 80% (95% CI 78-82) (Figure 1). The logistic regression (All MR + LR) applied on all medical records performed better than the CatBoost algorithm when restricted to CVD medical records (Figure 1). Conclusion The inclusion of all potential variables beyond the usual CVRF significantly improves the performances of SCD prediction models, independently of the methods (AI and logistic regression). Figure 1 : Receiver operating characteristic curves of the different models (ALL MR + ML, ALL MR + LR, CVD MR + ML and CVD MR + LR) in the derivation cohort and two validation cohorts (temporal and geographical)Figure 1.ROC curves

Integrating Machine Learning for Enhanced Agricultural Productivity: A Focus on Bananas and Arecanut in the Context of India’s Economic Growth

AbstractAgriculture is one of the sectors that has an important impact, taking into account the problem of sufficient food supply on a global level. The process of predicting the yield of crops is among the most challenging undertakings in the agricultural industry. Agriculture is the main source of income for most developing nations. The purpose of the study is to investigate the significant role that agriculture plays in boosting India's economic growth. Additionally, the research considers the challenges posed by a growing population and a changing environment in terms of agricultural production and food security. The research focuses on analysing the complex characteristics of the agricultural industry, with a particular emphasis on the nutritional importance of tropical fruits, notably bananas and arecanut. These fruits are well-known for their vital nutrients and their role in ensuring world food security. This study acknowledges the importance of sustainable agriculture practices and incorporates sophisticated machine learning algorithms as dynamic tools to forecast crop yields and enhance decision-making processes throughout the crop development cycle. The main aim of this study is to create strong machine learning models and statistical techniques that can accurately predict crop yield by combining a variety of environmental parameters, then assess which models outperform each other. Assist yield projections may provide governments and policymakers with valuable information to make well-informed choices about food security, import–export policies, and resource allocation. It facilitates national- and regional-level food supply planning. The validation method utilises important metrics like R square (R2), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). This present research adds to the continuing discussion on using creative methods to promote sustainable agricultural growth and ensure food security.

Predicting diabetes prognosis using machine learning techniques in a Hungarian clinical database

Abstract Background Chronic conditions such as type 2 diabetes mellitus have great impact on patients’ quality of life. Although clinical databases provide a perfect base for research aimed at improving diabetes care, analyzing such databases requires extensive pre-processing, mainly due to the large amount of unstructured data. Our study aimed to present the steps of generating a dataset from a large clinical database, and to apply machine learning-based analytical techniques regarding. Methods Data of the Clinical Center of University of Debrecen was used. To structure the unstructured data, regular expressions and natural language processing methods were used. The main machine learning models were as follows: Gradient Boosting Machines to predict the risk of complication development; Long Short-Term Memory Networks to forecast future health outcomes. All analysis and procedures were done using Python. Results The database contains approximately 1600 tables, with a total size of 1.9 terabytes, where the largest table is 21.07 gigabytes with approximately 44.64 million rows. The final dataset consisted of 40,332 patients. Most variables originate from the unstructured data, including complications and comorbidities of diabetes, as well as physical and laboratory parameters. Related to laboratory parameters, the number of measurements and the median value for every half-year were available. The diagnosis time of the complications’ occurrence is also presented. Machine learning methods were more accurate compared with traditional statistical methods in predicting the prognosis (p &amp;lt; 0.05). Discussion Our research highlights the importance of clinical data in chronic disease management. There are challenges in pre-processing and managing datasets, but machine learning-based methods are very efficient not only in extracting useful information from unstructured data but also in predicting the prognosis and identifying potential intervention points for better care. Key messages • Natural language processing can be used to obtain useful information from unstructured clinical data. • Using machine learning techniques on clinical data could improve diabates care.

Machine learning for prediction of incident heart failure and heart failure hospitalisation: multi-cohort development, validation and association with clinical phenotypes

Abstract Background Delayed heart failure (HF) diagnosis is associated with adverse outcomes and costs of hospitalisation.(1,2) Whilst previous tools have been developed to predict incident HF,(3) risk of HF incidence may correlate poorly with risk of HF hospitalisation, which may be more important in targeting diagnostics. Purpose To develop and validate a novel decision support tool for prediction of incident HF and HF hospitalisation, and investigate association of score with HF with reduced ejection fraction (HFrEF) and preserved ejection fraction (HFpEF). Methods We developed the models (FIND-HF) in UK primary care health record data from individuals aged ≥40 years without known HF (Jan 2, 1998 to Feb 28, 2022), randomly divided into training (80%) and testing (20%) datasets. We evaluated logistic regression and supervised machine learning models for prediction of 1- and 5- year HF and HF hospitalisation risk in the testing dataset with internal bootstrap validation. Models were externally validated for HF hospitalisation in data from a Taiwanese university hospital. Association of FIND-HF score with HF phenotypes of HFpEF and HFrEF was assessed in two prospective cohorts, the MATCH registry of patients diagnosed with HFrEF and the FIND-AF cohort of individuals at higher predicted risk of atrial fibrillation who underwent echocardiography. Results Of 565 284 UK individuals (mean age 68.9 (SD 12.2) years, 49.2% women), incidence of HF and HF hospitalisation was 1.1% and 0.5% at 1-year, and 6.6% and 3.4% at 5-years, respectively. Prediction performance was superior for machine learning algorithms compared with logistic regression, with best performance for XGBoost (area under receiver operating characteristic curve (AUROC) at 1 and 5-years: 0.755 and 0.723 for incident HF, and 0.738 and 0.711 for incident HF hospitalization). On external validation in 106,026 individuals in Taiwan (mean age 64.7 (SD 10.6) years, 52.8 % women), HF hospitalisation incidence was 0.71%, and 2.46% at 1- and 5- years, respectively, and discrimination performance was excellent (AUROC at 1- and 5-years: 0.834 and 0.843). When the FIND-HF score was applied to individuals in the MATCH cohort (n = 133, mean age 69.0 (SD 11.7) years, 38.3% women) and FIND-AF cohort (n = 82, mean age 71.6 (SD 7.5) years, 50.0% women), the distribution of score to clinical phenotypes demonstrated a binomial distribution (Figure 1), and likelihood of HFpEF diagnosis correlated with the numeric score (Figure 2). Conclusions The machine learning FIND-HF decision support tool can accurately identify individuals at risk of incident HF and HF hospitalisation, to enable new approaches for early detection and prevention. The variation of score amongst high risk individuals with different HF clinical phenotypes emphasizes the distinct pathophysiological pathways but shared adverse outcome profile across the HF syndrome.Figure 1Figure 2

Predicting the severity of valvular heart diseases using ai integrated with ECG and anthropometric data

Abstract Background AI has been used in studies to predict the severity of aortic stenosis (AS) from electrocardiography (ECG), but the application of AI to predict the severity of aortic regurgitation (AR), mitral regurgitation (MR) and tricuspid regurgitation (TR) has not been widely explored. This study aimed to determine whether deep learning can predict the severity of valvular heart disease (AS, AR, MR and TR) assessed by echocardiography using ECG data. Methods and Results We analyzed data from 31,787 adult patients who underwent ECG and echocardiography within three days of each other between 2014 and 2019 at two hospitals. Valvular heart diseases, including AS, AR, MR, and TR, were classified into categories; normal, mild, moderate, and severe; according to guidelines. ECGs, along with patient age, gender, height, and weight, were inputs to a LightGBM machine learning model. The model was trained on 80.6% of the data, with the remaining 19.4% used for validation. The sensitivity for predicting moderate or severe valvular heart disease was 85% for AS, 88% for AR, 85% for MR, and 85% for TR. The negative predictive values for AS, AR, MR, and TR were all 100%. Prediction accuracies for the severity of each disease were AS at 0.857, AR at 0.871, MR at 0.855, and TR at 0.890. Predictions based on AS diagnostic parameters; mean pressure gradient, valve area, and maximum blood flow velocity also showed favorable accuracies at 0.891, 0.821, and 0.877, respectively. Conclusion Integrating AI with standard clinical parameters, including ECG data, height, and weight, showed potential in accurately predicting the severity of various valvular heart diseases. Further researches across diverse populations and settings are necessary to validate the approach's effectiveness and explore its impact on patient care.

Generative adversarial networks (GANs) to produce synthetic 12-lead electrocardiogram signals for specific and rare diseases: a novel, powerful tool towards clinical advancements

Abstract Background Remarkable technological advancements in electrocardiogram (ECG) analysis face challenges due to insufficiently diverse and extensive real-world datasets. This limitation hampers both the development of sophisticated deep learning (DL) models for ECG analysis and the exposure of new physicians to substantial training examples. Purpose To leverage Generative Adversarial Networks (GANs), a novel AI algorithm type, to generate synthetic yet realistic ECG signals resembling specific diseases. Methods We initially employed a general-purpose GAN on more than 21,000 ECG samples from the PTB-XL database, encompassing normal and abnormal ECGs across various diseases. The GAN outputs served as inputs for training "DeceptionECG", a novel GAN model inspired by the InceptionTime architecture (code available online). DeceptionECG focused on generating disease-specific ECG signals, trained on annotated ECGs validated by two electrophysiologists. Post-training, we generated synthetic ECGs for common diseases like atrial fibrillation (AF) and anterior ST-segment elevation myocardial infarction (STEMI), as well as rarer conditions like Wolf-Parkinson White (WPW) syndrome. Authentication involved an independent discriminator DL model and two cardiologists who were asked to distinguish real from synthetic ECGs. Additionally, a DL model, a variant of the InceptionTime architecture, discerned specific diseases in the ECG signals to assess their realism. Finally, we trained a DL model for detecting abnormalities in ECGs with the augmented dataset to evaluate potential performance improvement. Results We generated sets of 5,000 synthetic ECGs for various abnormalities, at a rate of 2,500 12-lead ECG signals per minute. The generated ECGs were indistinguishable from real ones, evidenced by low accuracy (Accuracy, ACC: 48.9%) of the discriminator model and two clinicians (ACC: 56% and 51%), on test sets with anterior STEMI. A DL model trained on real ECGs accurately differentiated between synthetic normal ECGs and those hosting the target disease in nearly 9 out of 10 cases (ACC: 88.7%). Furthermore, the employed DL model detector trained on the augmented dataset, encompassing both real (n=1,589) and synthetic (n=5,000) ECGs, outperformed its counterpart trained solely on real ECGs, when tested on a 1,000-sample set of real signals (ACC: 94.6% vs. 89.3%). Conclusion Our GAN-based approach showcases the potential of generating high-quality synthetic ECGs tailored for specific diseases, overcoming data scarcity challenges, and enhancing the training and performance of machine learning models. Additionally, it holds promise in advancing medical professional training by providing diverse signals with distinct "disease stamps" at massive amounts, addressing privacy concerns associated with extensive data exposure. Further testing is warranted for the applicability of the approach, especially in exceedingly rare conditions with limited available training data.Project workflow and outcomesGAN-generated ECGs: Which is fake?