Model Validation Research Articles

Enhancing machine learning models: addressing challenges and future directions

Machine learning is considered as a core of modern artificial intelligence with progressive advancements throughout a spectrum including but not limited to healthcare and finance, natural language processing and self-driving cars. However, several problems remain to affect the efficiency, equal opportunities of users, and adaptability of ML models for an even faster-growing era. The limitations include shortage of high quality and access to training data, model complexity that can lead to overfitting, built in bias of the algorithm, interpretability and finally, the computational density needed for such big data models. These problems posed challenges to translate the knowledge derived from the ML systems into real-world use as well as hindering generalization of ML systems, particularly the medical and legal fields that have requirements of fairness and interpretability. These are the basic issues this journal addresses and provides possible ways of enhancing the performance of the ML models. To mitigate the problem of data deficiency, we present various techniques including data augmentation and transfer learning. To mitigate this issue, we present regularization strategies and methods of model validation. Several prevention methods are also mentioned including biasing of Algorithm and models using adversarial biasing, and Fairness-aware learning methods. Furthermore, we explore the increasing relevance of post-hoc model interpretability such as the SHAP and LIME methods which explain model’s outputs in a more detailed manner. The objective of the present Journal is to support further development of more stable, efficient and fair Machine Learning systems. In this paper, recent developments and long-term solutions are discussed to prepare the way for better and more responsible use of AI in the future.

Machine Learning-Based Real-Time Survival Prediction for Gastric Neuroendocrine Carcinoma.

This study aimed to develop a dynamic survival prediction model utilizing conditional survival (CS) analysis and machine learning techniques for gastric neuroendocrine carcinomas (GNECs). Data from the Surveillance, Epidemiology, and End Results (SEER) database (2004-2015) were analyzed and split into training and validation groups (7:3 ratio). CS profiles for patients with GNEC were examined in the full cohort. We utilized random survival forests (RSFs) and least absolute shrinkage and selection operator (LASSO) regression, alongside stepwise Cox regression, for variable selection. A CS-based nomogram was developed on the basis of key prognostic factors, followed by risk stratification and model validation. We included 654 patients with GNEC in our study, with 457 assigned to the training set and 197 to the validation set. The CS analysis demonstrated that the probability of achieving 5-year CS improved from 48% immediately after diagnosis to 68%, 81%, 88%, and 94% after surviving an additional year (i.e., at 1, 2, 3, and 4 years, respectively). Through the use of RSFs and LASSO regression, combined with multivariable regression analysis, we identified the optimal combination of prognostic factors, which included age, tumor grade, tumor stage, surgery, and chemotherapy. Utilizing these prognostic indicators, we successfully developed a nomogram model that incorporated CS and effectively stratified these patients by risk. Subsequent performance analyses further validated the superior efficacy of the nomogram. Our study highlights the value of CS in GNEC prognosis. The nomogram offers dynamic, individualized survival predictions, supporting personalized treatment strategies.

Development and validation of soil test crop response model for beetroot (Beta vulgaris) grown in ultisols of India

Development and validation of soil test crop response model for beetroot (Beta vulgaris) grown in ultisols of India

Development and validation of a risk prediction model for early postoperative high intraocular pressure after implantation of posterior chamber intraocular lens

Objective: To develop and validate a predictive model for assessing the risk of early postoperative high intraocular pressure (HIOP) following posterior chamber intraocular lens implantation. Methods: The clinical data of patients who underwent posterior chamber intraocular lens implantation at the Second Affiliated Hospital of Zhejiang University School of Medicine between May 2023 and April 2024 were retrospectively reviewed. Patients were divided into a modeling group and a validation group with a 7∶3 ratio using computerized random allocation. The modeling group was further categorized into normal intraocular pressure and HIOP groups based on the occurrence of early postoperative HIOP. Multivariate logistic regression analysis was employed to establish a risk prediction model. Receiver operating characteristic (ROC) curve analysis was conducted using R software to assess the predictive ability and accuracy of the model, along with calibration and decision curve analysis (DCA) curve plotting. Results: A total of 239 patients (478 eyes) were enrolled. The modeling group consisted of 168 patients (82 males and 86 females) (336 eyes), with an average age of (25.5±5.1) years. The validation group consisted of 71 patients (36 males and 35 females) (142 eyes), with an average age of (24.7±5.7) years. The early postoperative HIOP incidence rate after posterior chamber intraocular lens implantation was 32.4% (109/336), while it was 26.8% (38/142) in the validation group. The multivariate logistic regression analysis indicated that preoperative refractive diopter (OR=0.873, 95%CI: 0.782-0.976, P=0.017), corneal endothelial cell count (OR=0.996, 95%CI: 0.995-0.997, P<0.001), corneal thickness (OR=1.015, 95%CI: 1.006-1.024, P=0.001), and preoperative intraocular pressure (OR=1.470, 95%CI: 1.283-1.686, P<0.001) were associated factors for early HIOP. The Hosmer-Lemeshow test showed that the χ2 value was 8.444 (P=0.391), and the area under the ROC curve (AUC) was 0.888 (95%CI: 0.852-0.925). The Youden index reached its maximum value of 0.646, with a sensitivity of 0.862 and specificity of 0.784. AUC of validation group was 0.921 (95%CI: 0.870-0.972), with a maximum Youden index of 0.765, sensitivity of 0.842 and specificity of 0.923. Conclusions: The risk prediction model constructed based on preoperative refractive diopter, corneal endothelial cell count, corneal thickness, and preoperative intraocular pressure demonstrates good predictive performance for early postoperative HIOP following posterior chamber intraocular lens implantation.

Construction and validation of a nomogram prediction model for the occurrence of complications in patients following robotic radical surgery for gastric cancer.

In the last two decades, robotic-assisted gastrectomy has become a widely adopted surgical option for gastric cancer (GC) treatment. Despite its popularity, postoperative complications can significantly deteriorate patient quality of life and prognosis. Therefore, identifying risk factors for these complications is crucial for early detection and intervention. This research is designed to construct and validate a predictive model for assessing the risk of postoperative complications in patients undergoing robotic-assisted radical gastrectomy. A retrospective analysis was conducted on 500 GC patients from Gansu Provincial People's Hospital between December 2016 and October 2023. These patients formed the training cohort. An additional 136 patients from the 940th Hospital of Joint Logistic Support Force, the Chinese People's Liberation Army as the external validation cohort. Patients were categorized into groups with and without complications. Data collected included demographic details, laboratory results, CT quantitative body composition analysis, and clinical information. Variable selection was conducted through Lasso regression, succeeded by multivariable logistic regression to pinpoint independent risk factors. These elements facilitated the construction of a nomogram for prediction. The model's performance underwent internal validation via bootstrap techniques and external validation through a validation cohort. The efficacy of the model was quantified by the area under the receiver operating characteristic (ROC) curve (AUC), evaluated for calibration using calibration curves and the Hosmer-Lemeshow test, and assessed for clinical utility through decision curve analysis (DCA). Of the 500 patients in the training cohort, 65 experienced complications, a rate of 13%. The validation cohort had a similar complication rate of 13.24% (18 out of 136 patients). Independent risk factors identified included tumor diameter (OR = 1.99, 95% CI = 1.07-3.73), TNM stage III (OR = 2.12, 95% CI = 1.03-4.36), ASA class I (OR = 0.26, 95% CI = 0.13-0.53), ASA class III (OR = 4.75, 95% CI = 2.12-10.62), and visceral fat area (VFA) (OR = 2.52, 95% CI = 1.10-5.79). The nomogram demonstrated good discrimination (AUC = 0.81, 95% CI: 0.76-0.87) in internal validation and (AUC = 0.79, 95% CI: 0.67-0.90) in external validation. Both validations confirmed the model's accurate calibration and significant clinical utility, with net benefits observed at probability thresholds ranging from 2 to 79% and 2-71%. The developed nomogram, based on five independent risk factors-tumor diameter, TNM stage III, ASA class I, ASA class III, and VFA-effectively predicts the risk of complications in patients undergoing robotic-assisted radical gastrectomy, offering a valuable tool for clinical decision-making.

Commentary on “Validation of a Pretransplant Risk Prediction Model for Early Allograft Dysfunction After Living-donor Liver Transplantation”

Commentary on “Validation of a Pretransplant Risk Prediction Model for Early Allograft Dysfunction After Living-donor Liver Transplantation”

Development and validation of a novel risk-predicted model for early sepsis-associated acute kidney injury in critically ill patients: a retrospective cohort study

ObjectivesThis study aimed to develop a prediction model for the detection of early sepsis-associated acute kidney injury (SA-AKI), which is defined as AKI diagnosed within 48 hours of a sepsis...

From errors to excellence: the pre-analytical journey to improved quality in diagnostics. A scoping review.

This scoping review focuses on the evolution of pre-analytical errors (PAEs) in medical laboratories, a critical area with significant implications for patient care, healthcare costs, hospital length of stay, and operational efficiency. The Covidence Review tool was used to formulate the keywords, and then a comprehensive literature search was performed using several databases, importing the search results directly into Covidence (n=379). Title, abstract screening, duplicate removal, and full-text screening were done. The retrieved studies (n=232) were scanned for eligibility (n=228) and included in the review (n=83), and the results were summarised in a PRISMA flow chart. The review highlights the role of healthcare professionals in preventing PAEs in specimen collection and processing, as well as analyses. The review also discusses the use and advancements of artificial intelligence (AI) and machine learning in reducing PAEs and identifies inadequacies in standard definitions, measurement units, and education strategies. It demonstrates the need for further research to ensure model validation, address the regulatory validation of Risk Probability Indexation (RPI) models and consider regulatory, safety, and privacy concerns. The review suggests that comprehensive studies on the effectiveness of AI and software platforms in real-world settings and their implementation in healthcare are lacking, presenting opportunities for further research to advance patient care and improve the management of PAEs.

Development and internal validation of a metabolism-related model for predicting 30-day mortality in neonatal sepsis

ObjectiveNeonatal sepsis, a severe infectious disease associated with high mortality rates, is characterized by metabolic disturbances that play a crucial role in its progression. The aim of this study is to develop a metabolism-related model for assessing 30-day mortality in neonatal sepsis.MethodsThe clinical data of neonatal sepsis at Ganzhou Women and Children’s Health Care Hospital from January 2019 to December 2022 were retrospectively analyzed. Neonatal sepsis cases were divided into survival and non-survival groups. Multivariate logistic regression analysis was used to identify the independent risk factors for 30-day mortality. A nomogram model was developed based on these risk factors. Internal validation of the model was performed using 10-fold cross-validation. The predictive performance was evaluated through receiver operating characteristic (ROC) curves and calibration curve analyses. Decision curve analysis (DCA) was conducted to evaluate the clinical applicability of the developed model.ResultsThe study included a total of 156 cases of neonatal sepsis. Multivariate logistic regression analysis revealed that alanine(ALA), citrulline(CIT)), octadecanoylcarnitine(C18) and methionine(MET) were identified as independent risk factors for 30-day mortality of neonatal sepsis. The ROC curve showed an area under the curve of AUC = 0.866 (95% CI 0.796–0.936, P < 0.05). The calibration curve and DCA indicated excellent performance of the model.ConclusionThis study establishes a predictive model for neonatal sepsis-associated 30-day mortality, effectively capturing the perturbations in amino acid metabolism and fatty acid oxidation, thereby demonstrating robust predictive capabilities.

Development and validation of a prediction model of hospital mortality for patients with cardiac arrest survived 24 hours after cardiopulmonary resuscitation

ObjectiveResearch on predictive models for hospital mortality in patients who have survived 24 h following cardiopulmonary resuscitation (CPR) is limited. We aim to explore the factors associated with hospital mortality in these patients and develop a predictive model to aid clinical decision-making and enhance the survival rates of patients post-resuscitation.MethodsWe sourced the data from a retrospective study within the Dryad dataset, dividing patients who suffered cardiac arrest following CPR into a training set and a validation set at a 7:3 ratio. We identified variables linked to hospital mortality in the training set using Least Absolute Shrinkage and Selection Operator (LASSO) regression, as well as univariate and multivariate logistic analyses. Utilizing these variables, we developed a prognostic nomogram for predicting mortality post-CPR. Calibration curves, the area under receiver operating curves (ROC), decision curve analysis (DCA), and clinical impact curve were used to assess the discriminability, accuracy, and clinical utility of the nomogram.ResultsThe study population comprised 374 patients, with 262 allocated to the training group and 112 to the validation group. Of these, 213 patients were dead in the hospital. Multivariate logistic analysis revealed age (OR 1.05, 95% CI: 1.03–1.08), witnessed arrest (OR 0.28, 95% CI: 0.11–0.73), time to return of spontaneous circulation (ROSC) (OR 1.05, 95% CI: 1.02–1.08), non-shockable rhythm (OR 3.41, 95% CI: 1.61–7.18), alkaline phosphatase (OR 1.01, 95% CI: 1–1.01), and sequential organ failure assessment (SOFA) (OR 1.27, 95% CI: 1.15–1.4) were independent risk factors for hospital mortality for patients who survived 24 h after CPR. ROC of the nomogram showed the AUC in the training and validation group was 0.827 and 0.817, respectively. Calibration curves, DCA, and clinical impact curve demonstrated the nomogram with good accuracy and clinical utility.ConclusionOur prediction model had accurate predictive value for hospital mortality in patients who survived 24 h after CPR, which will be beneficial for assisting in identifying high-risk patients and intervention. Further confirmation of the model's accuracy required external validation data.

The Development and Validation of an Artificial Intelligence Model for Estimating Thumb Range of Motion Using Angle Sensors and Machine Learning: Targeting Radial Abduction, Palmar Abduction, and Pronation Angles

An accurate assessment of thumb range of motion is crucial for diagnosing musculoskeletal conditions, evaluating functional impairments, and planning effective rehabilitation strategies. In this study, we aimed to enhance the accuracy of estimating thumb range of motion using a combination of MediaPipe, which is an AI-based posture estimation library, and machine learning methods, taking the values obtained using angle sensors to be the true values. Radial abduction, palmar abduction, and pronation angles were estimated using MediaPipe based on coordinates detected from videos of 18 healthy participants (nine males and nine females with an age range of 30–49 years) selected to reflect a balanced distribution of height and other physical characteristics. A conical thumb movement model was constructed, and parameters were generated based on the coordinate data. Five machine learning models were evaluated, with LightGBM achieving the highest accuracy across all metrics. Specifically, for radial abduction, palmar abduction, and supination, the root mean square error (RMSE), mean absolute error (MAE), coefficient of determination (R2), and correlation coefficient were 4.67°, 3.41°, 0.94, and 0.97; 4.63°, 3.41°, 0.95, and 0.98; and 5.69°, 4.17°, 0.88, and 0.94, respectively. These results demonstrate that when estimating thumb range of motion, the AI model trained using angle sensor data and LightGBM achieved accuracy that was high and comparable to that of prior methods involving the use of MediaPipe and a protractor.

Development and validation of an EHR-based risk prediction model for geriatric patients undergoing urgent and emergency surgery

BackgroundClinical determination of patients at high risk of poor surgical outcomes is complex and may be supported by clinical tools to summarize the patient’s own personalized electronic health record (EHR) history and vitals data through predictive risk models. Since prior models were not readily available for EHR-integration, our objective was to develop and validate a risk stratification tool, named the Assessment of Geriatric Emergency Surgery (AGES) score, predicting risk of 30-day major postoperative complications in geriatric patients under consideration for urgent and emergency surgery using pre-surgical existing electronic health record (EHR) data.MethodsPatients 65-years and older undergoing urgent or emergency non-cardiac surgery within 21 hospitals 2017–2021 were used to develop the model (randomly split: 80% training, 20% test). The primary outcome was a 30-day composite outcome including several postoperative major complications and mortality; secondary outcomes included common individual complications included in the primary composite outcome (sepsis, progressive renal insufficiency or renal failure, and mortality). Patients’ EHR-based clinical history, vital signs, labs, and demographics were included in logistic regression, LASSO, decision tree, Random Forest, and XGBoost models. Area under the receiver operating characteristics curve (AUCROC) was used to compare model performance.ResultsOverall, 66,262 patients (median [IQR] age 78 [(70.9–84.0], female 53.9%, White race 68.5%) received urgent or emergency non-cardiac surgery (25.7% orthopedic cases, 21.9% general surgery cases). AUCROC ranged from 0.655 (Decision Tree) – 0.804 (XGBoost) for the primary composite outcome. XGBoost AUCROC was 0.823, 0.781, and 0.839 in predicting outcomes of sepsis, progressive renal insufficiency or renal failure, and mortality, respectively.ConclusionsWe developed a model to accurately predict major postoperative complications in geriatric patients undergoing urgent or emergency surgery using the patient’s own existing EHR data. EHR implementation of this model could efficiently support clinicians’ surgical risk assessment and perioperative decision-making discussions in this vulnerable patient population.

Development and Validation of a Machine Learning Model for Early Prediction of Sepsis Onset in Hospital Inpatients from All Departments

Background: With 11 million sepsis-related deaths worldwide, the development of tools for early prediction of sepsis onset in hospitalized patients is a global health priority. We developed a machine learning algorithm, capable of detecting the early onset of sepsis in all hospital departments. Methods: Predictors of sepsis from 45,127 patients from all departments of Valenciennes Hospital (France) were retrospectively collected for training. The binary classifier SEPSI Score for sepsis prediction was constructed using a gradient boosted trees approach, and assessed on the study dataset of 5270 patient stays, including 121 sepsis cases (2.3%). Finally, the performance of the model and its ability to detect early sepsis onset were evaluated and compared with existing sepsis scoring systems. Results: The mean positive predictive value of the SEPSI Score was 0.610 compared to 0.174 for the SOFA (Sepsis-related Organ Failure Assessment) score. The mean area under the precision–recall curve was 0.738 for SEPSI Score versus 0.174 for the most efficient score (SOFA). High sensitivity (0.845) and specificity (0.987) were also reported for SEPSI Score. The model was more accurate than all tested scores, up to 3 hours before sepsis onset. Half of sepsis cases were detected by the model at least 48 hours before their medically confirmed onset. Conclusions: The SEPSI Score model accurately predicted the early onset of sepsis, with performance exceeding existing scoring systems. It could be a valuable predictive tool in all hospital departments, allowing early management of sepsis patients. Its impact on associated morbidity-mortality needs to be further assessed.

External validation of a multivariable prediction model for positive resection margins in breast-conserving surgery

ObjectivesPositive resection margins after breast-conserving surgery (BCS) most often demands a repeat surgery. To preoperatively identify patients at risk of positive margins, a multivariable model has been developed that predicts positive margins after BCS with a high accuracy. This study aimed to externally validate this prediction model to explore its generalizability and assess if additional preoperatively available variables can further improve its predictive accuracy. The validation cohort included 225 patients with invasive breast cancer who underwent BCS at Aarhus University Hospital, Aarhus, Denmark during 2020–2022. Receiver operating characteristic (ROC) and calibration analysis were used to validate the prediction model. Univariable logistic regression was used to evaluate if additional variables available in the validation cohort were associated with positive margins and backward elimination to explore if these variables could further improve the model´s predictive accuracy.ResultsThe AUC of the model was 0.60 (95% CI: 0.50–0.70) indicating a lower discriminative capacity in the external cohort. We found weak evidence for an association between increased preoperative breast density on mammography and positive resection margins after BCS (p = 0.027), but the AUC of the model did not improve, when mammographic breast density was included as an additional variable in the model.

The alongshore tilt of mean dynamic topography and its implications for model validation and ocean monitoring

Abstract. Mean dynamic topography (MDT) plays an important role in the dynamics of shelf circulation. Coastal tide gauge observations in combination with the latest generation of geoid models are providing estimates of the alongshore tilt of MDT with unprecedented accuracy. Additionally, high-resolution ocean models are providing better representations of nearshore circulation and the associated tilt of MDT along their coastal boundaries. It has been shown that the newly available geodetic estimates can be used to validate model predictions of coastal MDT variability on global and basin scales. On smaller scales, however, there are significant variations in alongshore MDT that are of the same order of magnitude as the accuracy of the geoid models. In this study, we use a regional ocean model of the Gulf of Maine and Scotian Shelf (GoMSS) to demonstrate that the new observations of geodetically referenced coastal sea level can also provide valuable information for the validation of such high-resolution models. The predicted coastal MDT is in good agreement with coastal tide gauge observations referenced to the Canadian Gravimetric Geoid 2013 – Version A (CGG2013a), including a significant tilt of alongshore MDT along the coast of Nova Scotia. Using the validated GoMSS model and two idealized models, we show that this alongshore tilt of MDT can be interpreted in two complementary, and dynamically consistent, ways: in the coastal view, the tilt of MDT along the coast can provide a direct estimate of the average alongshore current. In the regional view, the tilt provides a measure of area-integrated nearshore circulation. This highlights the value of using geodetic MDT estimates for model validation and ocean monitoring.

Following intravenous thrombolysis, the outcome of diabetes mellitus associated with acute ischemic stroke was predicted via machine learning

This cohort study aimed to evaluate the prognostic outcomes of patients with acute ischemic stroke (AIS) and diabetes mellitus following intravenous thrombolysis, utilizing machine learning techniques. The analysis was conducted using data from Shenyang First People’s Hospital, involving 3,478 AIS patients with diabetes who received thrombolytic therapy from January 2018 to December 2023, ultimately focusing on 1,314 patients after screening. The primary outcome measured was the 90-day Modified Rankin Scale (MRS). An 80/20 train-test split was implemented for model development and validation, employing various machine learning classifiers, including artificial neural networks (ANN), random forest (RF), XGBoost (XGB), and LASSO regression. Results indicated that the average accuracy of the XGB model was 0.7355 (±0.0307), outperforming the other models. Key predictors for prognosis post-thrombolysis included the National Institutes of Health Stroke Scale (NIHSS) and blood platelet count. The findings underscore the effectiveness of machine learning algorithms, particularly XGB, in predicting functional outcomes in diabetic AIS patients, providing clinicians with a valuable tool for treatment planning and improving patient outcome predictions based on receiver operating characteristic (ROC) analysis and accuracy assessments.

Determination of The Level of Paracetamol and Tramadol Hcl Mixture in Tablets by Chemometric Method by Fourier Transform Infrared Spectrophotometry

Fixed-dose combinations of paracetamol and tramadol HCl are effective in the treatment of moderate to severe pain. Therefore, the determination of the levels of the mixture of paracetamol and tramadol HCl is very important for the pharmaceutical industry. The purpose of this study was to determine the levels of paracetamol and tramadol HCl mixture using Fourier Transform Infrared combined with partial least square (PLS) chemometric method. This study was conducted by making a calibration model and chemometric validation of each of the five concentrations measured absorbance at specific wave numbers of paracetamol and tramadol HCl. The results of the study on PLS multivariate calibration of paracetamol and tramadol HCl with RMSECV values of 0.06279 and 0.03785, PRESS values of 0.02366 and 0.00860, and R2 values of 0.9995 0.9998, respectively. All validation parameters are within the acceptable range. This indicates that the method can be used to accurately determine drug levels without separation

Toward Multi-Dimensional Separation of Nanoparticles in Tubular Centrifuges

The processing and preparation of particulate products is an important process in modern industry and science. The enormous potential for innovation in research and development is due to the complex interactions of solids with their environment. The aim of advanced particle production is to achieve high yields of narrowly distributed particle sizes, shapes or material compositions that provide advantageous product specifications. The integration of solid–liquid separation into these processes expands the process engineering scope in terms of product quality and efficiency. Designing these processes to accommodate a wide range of separation characteristics at small-particle-size scales is a major challenge. Taking these aspects into account, the present work aims to improve a dynamic simulation tool for tubular centrifuges that models the time- and space-dependent mass transport and thus, for the first time, can predict separation outcomes when processing both single- and multi-component systems. Utilizing an optical measurement technique, nanosuspension properties can be measured in real time during separation to support model validation. The simulation results align closely with experimental findings and offer plausible insights when addressing multi-dimensional property distributions of non-spherical particles. This study contributes to advanced modeling of separation experiments in tubular centrifuges in real time, taking into account multiple particle properties such as material density and particle form.

Experimental-Based Simulation of EV Drive Mechanism

A robust experimental-based vehicle drive system model and parameters tuned to match the real-world performance of the electro-mechanical systems is presented. The model is based on battery electric vehicle systems, focusing on the individual components of the drive system. The model includes motors, a battery pack, a driveline, and a DC-DC converter. Based on the collected empirical data using an on-board data acquisition system, and fine-tuning the critical vehicle parameters to align the simulation with the empirical data, stationary and drive cycle tests were designed and performed to replicate standard day-to-day driving conditions. Performance metrics such as the power consumption, the state of charge and the range were employed for the model validation. The validated model was then used for the performance study at different operating temperatures. The results of the study demonstrate that the model successfully mirrored the performance of the powertrain under various conditions and lined up with existing research work, providing a valuable tool for further research and development in battery electric vehicle technology.

Physics design of current drive and strategy of heating system for EHL-2 spherical torus

Abstract ENN He Long-2 (EHL-2) is the next-generation large Mega-Ampere (MA) spherical torus (ST) proposed and funded by ENN company. The design parameters are: Ti0 &gt; 30 keV, ne0~1×1020 m-3, Ip ~ 3 MA, Bt ~ 3 T. How to achieve several MA current flat-top with limited voltage-second (Vs) of center solenoid (CS) coils is one of the most challenges issue of EHL-2. In order to minimize the consumption of Vs, fully non-inductive start-up by ECRH will be applied in EHL-2. The ramp-up phase will be accomplished with the synergetic mode between CS and non-inductive methods. The strategy of non-inductive start-up and ramp-up with synergetic mode has been verified on EXL-50U’s experiments. Based on this strategy, numerical simulations give the feasibility of EHL-2 to achieve 3 MA plasma current. A high-performance steady-state scenario with Ip ~ 1.5 MA is also designed. In this scenario, the bootstrap current fraction fBS &gt; 70%, the safety factor q at the magnetic axis q0 &gt; 2, the minimum safety factor qmin &gt; 1, the poloidal beta β p &gt; 3 and normalized beta β N &gt; 2.3. Each design iteration integrates the validation of physical models with the constraints of engineering implementation, gradually optimizing the performance of the Heating and Current Drive (H&amp;CD) systems. Numerical simulation results for general auxiliary H&amp;CD systems such as neutral beam injection, electron cyclotron wave, ion cyclotron wave, and lower hybrid wave are presented. These simulation results ensure that the 31 MW H&amp;CD systems comprehensively cover all scenarios while maintaining engineering feasibility