Deep learning model based on primary tumor to predict lymph node status in clinical stage IA lung adenocarcinoma:a multicenter study

Objective:To develop a deep learning model to predict lymph node(LN)status in clinical stage IA lung adeno-carcinoma patients.Methods:This diagnostic study included 1,009 patients with pathologically confirmed clinical stage T1N0M0 lung adenocarcinoma from two independent datasets(699 from Cancer Hospital of Chinese Academy of Medical Sciences and 310 from PLA General Hospital)between January 2005 and December 2019.The Cancer Hospital dataset was randomly split into a training cohort(559 patients)and a validation cohort(140 patients)to train and tune a deep learning model based on a deep residual network(ResNet).The PLA Hospital dataset was used as a testing cohort to evaluate the generalization ability of the model.Thoracic radiologists manually segmented tumors and interpreted high-resolution computed tomography(HRCT)features for the model.The predictive performance was assessed by area under the curves(AUCs),accuracy,precision,recall,and F1 score.Subgroup analysis was performed to evaluate the potential bias of the study population.Results:A total of 1,009 patients were included in this study;409(40.5%)were male and 600(59.5%)were female.The median age was 57.0 years(inter-quartile range,IQR:50.0-64.0).The deep learning model achieved AUCs of 0.906(95%CI:0.873-0.938)and 0.893(95%CI:0.857-0.930)for predicting pN0 disease in the testing cohort and a non-pure ground glass nodule(non-pGGN)testing cohort,respectively.No significant difference was detected between the testing cohort and the non-pGGN testing cohort(P=0.622).The precisions of this model for predicting pN0 disease were 0.979(95%CI:0.963-0.995)and 0.983(95%CI:0.967-0.998)in the testing cohort and the non-pGGN testing cohort,respectively.The deep learning model achieved AUCs of 0.848(95%CI:0.798-0.898)and 0.831(95%CI:0.776-0.887)for predicting pN2 disease in the testing cohort and the non-pGGN testing cohort,respectively.No significant difference was detected between the testing cohort and the non-pGGN testing cohort(P=0.657).The recalls of this model for predicting pN2 disease were 0.903(95%CI:0.870-0.936)and 0.931(95%CI:0.901-0.961)in the testing cohort and the non-pGGN testing cohort,respectively.Conclusions:The superior performance of the deep learning model will help to target the extension of lymph node dissection and reduce the ineffective lymph node dissection in early-stage lung adenocarcinoma patients.

Present and future of new systemic therapies for early and intermediate stages of hepatocellular carcinoma

Hepatocellular carcinoma(HCC)is a high mortality neoplasm which usually appears on a cirrhotic liver.The therapeutic arsenal and subsequent prognostic outlook are intrinsically linked to the HCC stage at diagnosis.Notwithstanding the current deployment of treatments with curative intent(liver resection/local ablation and liver transplantation)in early and intermediate stages,a high rate of HCC recurrence persists,underscoring a pivotal clinical challenge.Emergent systemic therapies(ST),particularly immunotherapy,have demonstrate promising outcomes in terms of increase overall survival,but they are currently bound to the advanced stage of HCC.This review provides a comprehensive analysis of the literature,encompassing studies up to March 10,2024,evaluating the impact of novel ST in the early and intermediate HCC stages,specially focusing on the findings of neoadjuvant and adjuvant regimens,aimed at increasing significantly overall survival and recurrence-free survival after a treatment with curative intent.We also investigate the potential role of ST in enhancing the downstaging rate for the intermediate-stage HCC initially deemed ineligible for treatment with curative intent.Finally,we critically discuss about the current relevance of the results of these studies and the encouraging future implications of ST in the treatment schedules of early and intermediate HCC stages.

基于Multi-Agent的无人机集群体系自主作战系统设计

针对无人集群自主作战体系设计中的关键问题,提出基于Multi-Agent的无人集群自主作战系统设计方法。建立无人集群各节点的Agent模型及其推演规则;对于仿真系统模块化和通用化的需求,设计系统互操作式接口和无人集群自主作战的交互关系;开展无人集群系统仿真推演验证。仿真结果表明,所提设计方案不仅能够有效开展并完成自主作战网络生成-集群演化-效能评估的全过程动态演示验证,而且能够通过重复随机试验进一步评估无人集群的协同作战效能,最后总结了集群协同作战的策略和经验。

Stage at diagnosis of colorectal cancer through diagnostic route:Who should be screened?

BACKGROUND Colorectal cancer(CRC)is a global health concern,with advanced-stage diagnoses contributing to poor prognoses.The efficacy of CRC screening has been well-established;nevertheless,a significant proportion of patients remain unscreened,with>70%of cases diagnosed outside screening.Although identifying specific subgroups for whom CRC screening should be particularly recommended is crucial owing to limited resources,the association between the diagnostic routes and identification of these subgroups has been less appreciated.In the Japanese cancer registry,the diagnostic routes for groups discovered outside of screening are primarily categorized into those with comorbidities found during hospital visits and those with CRC-related symptoms.AIM To clarify the stage at CRC diagnosis based on diagnostic routes.METHODS We conducted a retrospective observational study using a cancer registry of patients with CRC between January 2016 and December 2019 at two hospitals.The diagnostic routes were primarily classified into three groups:Cancer screening,follow-up,and symptomatic.The early-stage was defined as Stages 0 or I.Multivariate and univariate logistic regressions were exploited to determine the odds of early-stage diagnosis in the symptomatic and cancer screening groups,referencing the follow-up group.The adjusted covariates were age,sex,and tumor location.RESULTS Of the 2083 patients,715(34.4%),1064(51.1%),and 304(14.6%)belonged to the follow-up,symptomatic,and cancer screening groups,respectively.Among the 2083 patients,CRCs diagnosed at an early stage were 57.3%(410 of 715),23.9%(254 of 1064),and 59.5%(181 of 304)in the follow-up,symptomatic,and cancer screening groups,respectively.The symptomatic group exhibited a lower likelihood of early-stage diagnosis than the follow-up group[P<0.001,adjusted odds ratio(aOR),0.23;95%confidence interval(95%CI):0.19-0.29].The likelihood of diagnosis at an early stage was similar between the follow-up and cancer screening groups(P=0.493,aOR for early-stage diagnosis in the cancer screening group vs follow-up group=1.11;95%CI=0.82-1.49).CONCLUSION CRCs detected during hospital visits for comorbidities were diagnosed earlier,similar to cancer screening.CRC screening should be recommended,particularly for patients without periodical hospital visits for comorbidities.

联合物理层与MAC层的multi-TRP上行重叠传输处理机制

针对非理想回程下现有协议难以有效处理多传输接收节点(multi-TRP)场景中多定时提前(multi-TA)导致的严重上行链路(UL)重叠传输问题,联合改进物理层复用技术和介质访问控制(MAC)令牌桶技术,提出了一种新型的UL重叠传输处理机制。该新型机制通过改进物理层重叠信道识别流程、复用要求及复用规则,将物理层复用信息与重叠信息反馈至MAC层,并对MAC层令牌桶技术进行优化。通过仿真实验对所提机制与现有协议机制进行对比,结果表明,在逻辑时隙不可重叠和可重叠2种情形下,物理上行控制信道(PUCCH)实际复用数量性能平均提升了57.58%和49.40%,物理上行共享信道(PUSCH)实际可用资源数量性能平均提升了12.09%和26.03%;优先级最高逻辑信道实际占用资源数量性能平均提升了33.33%和45.48%。

基于Multi-Agent的水电站变压器故障诊断系统

为了精准、快速完成水电站变压器的故障诊断,设计基于Multi-Agent的水电站变压器故障诊断系统。变压器状态监控agent将检测到的变压器故障信息发送给系统管理agent,系统管理agent通过通信agent将变压器故障信息发送给变压器故障诊断agent,变压器故障诊断agent利用小波变换方法提取变压器故障特征,并将其作为IFOA-SVM模型输入,完成变压器故障分类后,获取变压器故障诊断结果,该结果通过通信agent显示给用户。实验表明,该系统可有效诊断变压器故障诊断,诊断成功率受系统故障信息丢失率的影响较小,诊断耗时、耗能小,并具有较高故障诊断成功率。

Double contrast-enhanced ultrasonography improves diagnostic accuracy of T staging compared with multi-detector computed tomography in gastric cancer patients

BACKGROUND Gastric cancer(GC)is the most common malignant tumor and ranks third for cancer-related deaths among the worldwide.The disease poses a serious public health problem in China,ranking fifth for incidence and third for mortality.Knowledge of the invasive depth of the tumor is vital to treatment decisions.AIM To evaluate the diagnostic performance of double contrast-enhanced ultrasonography(DCEUS)for preoperative T staging in patients with GC by comparing with multi-detector computed tomography(MDCT).METHODS This single prospective study enrolled patients with GC confirmed by preoperative gastroscopy from July 2021 to March 2023.Patients underwent DCEUS,including ultrasonography(US)and intravenous contrast-enhanced ultrasonography(CEUS),and MDCT examinations for the assessment of preoperative T staging.Features of GC were identified on DCEUS and criteria developed to evaluate T staging according to the 8th edition of AJCC cancer staging manual.The diagnostic performance of DCEUS was evaluated by comparing it with that of MDCT and surgical-pathological findings were considered as the gold standard.RESULTS A total of 229 patients with GC(80 T1,33 T2,59 T3 and 57 T4)were included.Overall accuracies were 86.9%for DCEUS and 61.1%for MDCT(P<0.001).DCEUS was superior to MDCT for T1(92.5%vs 70.0%,P<0.001),T2(72.7%vs 51.5%,P=0.041),T3(86.4%vs 45.8%,P<0.001)and T4(87.7%vs 70.2%,P=0.022)staging of GC.CONCLUSION DCEUS improved the diagnostic accuracy of preoperative T staging in patients with GC compared with MDCT,and constitutes a promising imaging modality for preoperative evaluation of GC to aid individualized treatment decision-making.

Competitive Multi-Verse Optimization with Deep Learning Based Sleep Stage Classification

Sleep plays a vital role in optimum working of the brain and the body.Numerous people suffer from sleep-oriented illnesses like apnea,insomnia,etc.Sleep stage classification is a primary process in the quantitative examination of polysomnographic recording.Sleep stage scoring is mainly based on experts’knowledge which is laborious and time consuming.Hence,it can be essential to design automated sleep stage classification model using machine learning(ML)and deep learning(DL)approaches.In this view,this study focuses on the design of Competitive Multi-verse Optimization with Deep Learning Based Sleep Stage Classification(CMVODL-SSC)model using Electroencephalogram(EEG)signals.The proposed CMVODL-SSC model intends to effectively categorize different sleep stages on EEG signals.Primarily,data pre-processing is performed to convert the actual data into useful format.Besides,a cascaded long short term memory(CLSTM)model is employed to perform classification process.At last,the CMVO algorithm is utilized for optimally tuning the hyperparameters involved in the CLSTM model.In order to report the enhancements of the CMVODL-SSC model,a wide range of simulations was carried out and the results ensured the better performance of the CMVODL-SSC model with average accuracy of 96.90%.

Multi Location Field Evaluation of BC1F2 Sorghum Populations for Striga Resistance in Niger

In Niger, a landlocked country, sorghum is the second staple food cultivated over the country by smallholder farmer. The crop is important for human and animal consumption. Despite its importance, the crop is affected by biotic and abiotic constraints. Among those constraints, striga has a high impact on yield. In fact, to survive, farmers are growing their local preferred sorghum varieties wish is highly sensible to the weed. Striga management is a challenge that requires a permanent solution. In addition, the development of high-yielding Striga resistant genotypes will be appreciated by farmers. The development of striga resistance will be based on the breeding population performances under farmer’s diverse environmental conditions adaptation. The main objective of this study is to evaluate two breeding populations for striga resistance in two different environments at Boulke and Dibissou in Tahoua region, to identify the early and high-yielding striga tolerant genotypes under natural infestation.

A Robust Approach for Multi Classification-Based Intrusion Detection through Stacking Deep Learning Models

Intrusion detection is a predominant task that monitors and protects the network infrastructure.Therefore,many datasets have been published and investigated by researchers to analyze and understand the problem of intrusion prediction and detection.In particular,the Network Security Laboratory-Knowledge Discovery in Databases(NSL-KDD)is an extensively used benchmark dataset for evaluating intrusion detection systems(IDSs)as it incorporates various network traffic attacks.It is worth mentioning that a large number of studies have tackled the problem of intrusion detection using machine learning models,but the performance of these models often decreases when evaluated on new attacks.This has led to the utilization of deep learning techniques,which have showcased significant potential for processing large datasets and therefore improving detection accuracy.For that reason,this paper focuses on the role of stacking deep learning models,including convolution neural network(CNN)and deep neural network(DNN)for improving the intrusion detection rate of the NSL-KDD dataset.Each base model is trained on the NSL-KDD dataset to extract significant features.Once the base models have been trained,the stacking process proceeds to the second stage,where a simple meta-model has been trained on the predictions generated from the proposed base models.The combination of the predictions allows the meta-model to distinguish different classes of attacks and increase the detection rate.Our experimental evaluations using the NSL-KDD dataset have shown the efficacy of stacking deep learning models for intrusion detection.The performance of the ensemble of base models,combined with the meta-model,exceeds the performance of individual models.Our stacking model has attained an accuracy of 99%and an average F1-score of 93%for the multi-classification scenario.Besides,the training time of the proposed ensemble model is lower than the training time of benchmark techniques,demonstrating its efficiency and robustness.

Ecological network analysis reveals complex responses of tree species life stage interactions to stand variables

Tree interactions are essential for the structure,dynamics,and function of forest ecosystems,but variations in the architecture of life-stage interaction networks(LSINs)across forests is unclear.Here,we constructed 16 LSINs in the mountainous forests of northwest Hebei,China based on crown overlap from four mixed forests with two dominant tree species.Our results show that LSINs decrease the complexity of stand densities and basal areas due to the interaction cluster differentiation.In addition,we found that mature trees and saplings play different roles,the first acting as“hub”life stages with high connectivity and the second,as“bridges”controlling information flow with high centrality.Across the forests,life stages with higher importance showed better parameter stability within LSINs.These results reveal that the structure of tree interactions among life stages is highly related to stand variables.Our efforts contribute to the understanding of LSIN complexity and provide a basis for further research on tree interactions in complex forest communities.

Feedback linearization and equivalent-disturbance compensation control strategy for piezoelectric stage

Piezoelectric stages use piezoelectric actuators and flexure hinges as driving and amplifying mechanisms,respectively.These systems have high positioning accuracy and high-frequency responses,and they are widely used in various precision/ultra-precision positioning fields.However,the main challenge with these devices is the inherent hysteresis nonlinearity of piezoelectric actuators,which seriously affects the tracking accuracy of a piezoelectric stage.Inspired by this challenge,in this work,we developed a Hammerstein model to describe the hysteresis nonlinearity of a piezoelectric stage.In particular,in our proposed scheme,a feedback-linearization algorithm is used to eliminate the static hysteresis nonlinearity.In addition,a composite controller based on equivalent-disturbance compensation was designed to counteract model uncertainties and external disturbances.An analysis of the stability of a closed-loop system based on this feedback-linearization algorithm and composite controller was performed,and this was followed by extensive comparative experiments using a piezoelectric stage developed in the laboratory.The experimental results confirmed that the feedback-linearization algorithm and the composite controller offer improved linearization and trajectory-tracking performance.

An Elite-Class Teaching-Learning-Based Optimization for Reentrant Hybrid Flow Shop Scheduling with Bottleneck Stage

Bottleneck stage and reentrance often exist in real-life manufacturing processes;however,the previous research rarely addresses these two processing conditions in a scheduling problem.In this study,a reentrant hybrid flow shop scheduling problem(RHFSP)with a bottleneck stage is considered,and an elite-class teaching-learning-based optimization(ETLBO)algorithm is proposed to minimize maximum completion time.To produce high-quality solutions,teachers are divided into formal ones and substitute ones,and multiple classes are formed.The teacher phase is composed of teacher competition and teacher teaching.The learner phase is replaced with a reinforcement search of the elite class.Adaptive adjustment on teachers and classes is established based on class quality,which is determined by the number of elite solutions in class.Numerous experimental results demonstrate the effectiveness of new strategies,and ETLBO has a significant advantage in solving the considered RHFSP.

Mutual information oriented deep skill chaining for multi‐agent reinforcement learning

Multi‐agent reinforcement learning relies on reward signals to guide the policy networks of individual agents.However,in high‐dimensional continuous spaces,the non‐stationary environment can provide outdated experiences that hinder convergence,resulting in ineffective training performance for multi‐agent systems.To tackle this issue,a novel reinforcement learning scheme,Mutual Information Oriented Deep Skill Chaining(MioDSC),is proposed that generates an optimised cooperative policy by incorporating intrinsic rewards based on mutual information to improve exploration efficiency.These rewards encourage agents to diversify their learning process by engaging in actions that increase the mutual information between their actions and the environment state.In addition,MioDSC can generate cooperative policies using the options framework,allowing agents to learn and reuse complex action sequences and accelerating the convergence speed of multi‐agent learning.MioDSC was evaluated in the multi‐agent particle environment and the StarCraft multi‐agent challenge at varying difficulty levels.The experimental results demonstrate that MioDSC outperforms state‐of‐the‐art methods and is robust across various multi‐agent system tasks with high stability.

Dry eye rate and its relationship with disease stage in patients with primary hypertension:a cross-sectional study in Vietnam

AIM:To determine the dry eye(DE)rate and its relationship with disease stage in patients with primary hypertension.METHODS:A cross-sectional study included 432 patients with primary hypertension(with an equal number of patients in each group:144 in stage Ⅰ,Ⅱ,and Ⅲ hypertension)and 144 healthy subjects as a control group.The Ocular Surface Disease Index(OSDI)and Schirmer Ⅰ test without anesthetics were conducted on all 576 subjects.Subjects with OSDI scores<13 and Schirmer Ⅰ values equal to or under 10 mm were diagnosed with DE.RESULTS:The ratio of DE in hypertension patients was higher than in the control group(41.7%versus 18.8%;P<0.001).The proportion of patients with DE increased gradually according to the hypertension stage:27.1% in stage Ⅰ,40.3% in stage Ⅱ,and 57.6% in stage Ⅲ,P<0.001.Age,duration of hypertension,plasma urea,creatinine,and high-sensitivity C-reactive protein(CRP-hs)levels in hypertension patients with DE were higher than those without DE,P<0.001.Advanced age,a long duration of hypertension,diabetes mellitus,elevated plasma creatinine,and CRP-hs levels were independent factors associated with DE in primary hypertension patients,P<0.001.CONCLUSION:DE is a common disorder associated with advanced age,a long duration of hypertension,diabetes mellitus,elevated plasma CRP-hs,and creatinine levels in patients with primary hypertension.

A strength model for cement-stabilized mud subjected to various curing temperatures and its early-stage quality control

The influence of curing temperature on the strength development of cement-stabilized mud has been well documented in terms of strength-increase rate and ultimate strength.However,the strength development model is not mature for the extremely early stages.In addition,there is a lack of studies on quality control methods based on early-stage strength development.This paper presents a strength model for cement-stabilized mud to address these gaps,considering various curing temperatures and early-stage behaviors.In this study,a series of laboratory experiments was conducted on two types of muds treated with Portland blast furnace cement and ordinary Portland cement under four different temperatures.The results indicate that elevated temperatures expedite strength development and lead to higher long-term strength.The proposed model,which combines a three-step conversion process and a hyperbolic model at the reference temperature,enables accurate estimate of the strength development for cement-treated mud with any proportions cured under various temperatures.With this model,a practical early quality control method is introduced for applying cement-stabilized mud in field projects.The back-analysis parameters obtained from a 36-h investigation at temperature of 60C demonstrated a sufficient accuracy in predicting strength levels in practical applications.

CmMYB3-like negatively regulates anthocyanin biosynthesis and flower color formation during the post-flowering stage in Chrysanthemum morifolium

Color fading caused by a decrease in anthocyanin accumulation during the post-flowering stage significantly affects postharvest quality of chrysanthemum.However,the underlying mechanism by which anthocyanin accumulation decreases during the post-flowering stage still unclear,which greatly restricts design of molecular breeding in chrysanthemum.Here,a chrysanthemum SG7 R2R3 MYB transcription factor(TF),CmMYB3-like,was identified to have a function in regulating anthocyanin biosynthesis during the post-flowering stage.Quantitative real time PCR(qRT-PCR)assays showed that the expression of CmMYB3-like was gradually downregulated when anthocyanin content increased during the flowering stage and was significantly upregulated during the post-flowering stage.Genetic transformation of chrysanthemum and dual-luciferase assays in N.benthamiana leaves showed that CmMYB3-like suppressed anthocyanin accumulation by inhibiting the transcription of CmCHS and CmANS directly and that of CmF3H indirectly.However,overexpression or suppression of CmMYB3-like did not affect the biosynthesis of flavones or flavonols.Genetic transformation of chrysanthemum revealed that the overexpression of CmMYB3-like inhibited anthocyanin accumulation,but its suppression prevented the decrease in anthocyanin accumulation during the post-flowering stage.Our results revealed a crucial role of CmMYB3-like in regulating the color of petals during the post-flowering stage and provided a target gene for molecular design breeding to improve the postharvest quality of chrysanthemum.

Single-cell profiling of the pig cecum at various developmental stages

The gastrointestinal tract is essential for food digestion,nutrient absorption,waste elimination,and microbial defense.Single-cell transcriptome profiling of the intestinal tract has greatly enriched our understanding of cellular diversity,functional heterogeneity,and their importance in intestinal tract development and disease.Although such profiling has been extensively conducted in humans and mice,the single-cell gene expression landscape of the pig cecum remains unexplored.Here,single-cell RNA sequencing was performed on 45572 cells obtained from seven cecal samples in pigs at four different developmental stages(days(D)30,42,150,and 730).Analysis revealed 12 major cell types and 38 subtypes,as well as their distinctive genes,transcription factors,and regulons,many of which were conserved in humans.An increase in the relative proportions of CD8^(+)T and Granzyme A(low expression)natural killer T cells(GZMA^(low)NKT)cells and a decrease in the relative proportions of epithelial stem cells,Tregs,RHEX^(+)T cells,and plasmacytoid dendritic cells(pDCs)were noted across the developmental stages.Moreover,the post-weaning period exhibited an up-regulation in mitochondrial genes,COX2 and ND2,as well as genes involved in immune activation in multiple cell types.Cell-cell crosstalk analysis indicated that IBP6^(+)fibroblasts were the main signal senders at D30,whereas IBP6^(−)fibroblasts assumed this role at the other stages.NKT cells established interactions with epithelial cells and IBP6^(+)fibroblasts in the D730 cecum through mediation of GZMA-F2RL1/F2RL2 pairs.This study provides valuable insights into cellular heterogeneity and function in the pig cecum at different development stages.

Multi-scale data joint inversion of minerals and porosity in altered igneous reservoirs—A case study in the South China Sea

There are abundant igneous gas reservoirs in the South China Sea with significant value of research,and lithology classification,mineral analysis and porosity inversion are important links in reservoir evaluation.However,affected by the diverse lithology,complicated mineral and widespread alteration,conventional logging lithology classification and mineral inversion become considerably difficult.At the same time,owing to the limitation of the wireline log response equation,the quantity and accuracy of minerals can hardly meet the exploration requirements of igneous formations.To overcome those issues,this study takes the South China Sea as an example,and combines multi-scale data such as micro rock slices,petrophysical experiments,wireline log and element cutting log to establish a set of joint inversion methods for minerals and porosity of altered igneous rocks.Specifically,we define the lithology and mineral characteristics through core slices and mineral data,and establish an igneous multi-mineral volumetric model.Then we determine element cutting log correction method based on core element data,and combine wireline log and corrected element cutting log to perform the lithology classification and joint inversion of minerals and porosity.However,it is always difficult to determine the elemental eigenvalues of different minerals in inversion.This paper uses multiple linear regression methods to solve this problem.Finally,an integrated inversion technique for altered igneous formations was developed.The results show that the corrected element cutting log are in good agreement with the core element data,and the mineral and porosity results obtained from the joint inversion based on the wireline log and corrected element cutting log are also in good agreement with the core data from X-ray diffraction.The results demonstrate that the inversion technique is applicable and this study provides a new direction for the mineral inversion research of altered igneous formations.

Novel method for identifying the stages of discharge underwater based on impedance change characteristic

It is difficult to determine the discharge stages in a fixed time of repetitive discharge underwater due to the arc formation process being susceptible to external environmental influences. This paper proposes a novel underwater discharge stage identification method based on the Strong Tracking Filter(STF) and impedance change characteristics. The time-varying equivalent circuit model of the discharge underwater is established based on the plasma theory analysis of the impedance change characteristics and mechanism of the discharge process. The STF is used to reduce the randomness of the impedance of repeated discharges underwater, and then the universal identification resistance data is obtained. Based on the resistance variation characteristics of the discriminating resistance of the pre-breakdown, main, and oscillatory discharge stages, the threshold values for determining the discharge stage are obtained. These include the threshold values for the resistance variation rate(K) and the moment(t).Experimental and error analysis results demonstrate the efficacy of this innovative method in discharge stage determination, with a maximum mean square deviation of Scrless than 1.761.