Identifying delayed neurological sequelae during the acute phase of carbon monoxide poisoning based on diffusion-weighted imaging and clinical features

Identifying delayed neurological sequelae during the acute phase of carbon monoxide poisoning based on diffusion-weighted imaging and clinical features | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identifying delayed neurological sequelae during the acute phase of carbon monoxide poisoning based on diffusion-weighted imaging and clinical features Siying Chen, Shijun Yang, Minghui Tan, Heying Lu, Huan Li, Jinlan Li, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5443111/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Background Delayed neurological sequelae (DNS) represents a critical and potentially fatal complication. Therefore, the timely recognition of individuals at risk of developing DNS in early phase holds significant clinical value. This study aims to identify diffusion-weighted imaging (DWI) characteristics related to developing DNS, and construct a predictive model encompassing DWI characteristics and clinical variables to early prediction of DNS during the acute phase of carbon monoxide (CO) poisoning. Methods We retrospectively include 120 poisoned patients with newly diagnosed with CO poisoning. The subjects were divided into non-DNS group (n = 75) and DNS group (n = 45) after at least 60-day follow-up. The fractional anisotropy value and mean diffusivity value were measured in the regions of interest placed on the amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle. A support vector machine (SVM) model integrated both DWI and clinical features was developed and evaluated. And mean impact value was used to rank the features that had impacts on classification. Results A total of 24 clinical features and 28 DWI features were included. 8 clinical features and 23 DWIs were included in the SVM model. Three SVM models were established based solely on clinical features or DWI features, and combined clinical and DWI features, with prediction accuracy of 0.76, 0.94, and 0.97, respectively. The precision, sensitivity, F1 score, macroscopic mean, weighted mean, and AUC of the combined model is 1.00, 0.92, 0.96, 0.98, 0.97, 0.97, respectively, and the result of 10-flod cross validation was 0.98. The mean fractional anisotropy of left caudate had the highest impact on the SVM model. Conclusions The fractional anisotropy and mean diffusivity of DWI may be a potential biomarker in identifying patients at risk of developing DNS. Our comprehensive SVM model with multimodal features had excellent accuracy and clinical practicability in identifying DNS. CO poisoning Delayed encephalopathy sequelae Diffusion-weighted imaging Fractional anisotropy Mean diffusivity Machine learning Figures Figure 1 Figure 2 1. Introduction Carbon monoxide (CO) is a kind of colorless and odorless poisonous gas, which exists widely in environment [ 1 ]. Acute CO poisoning stands as a significant contributor to both accidental and intentional injuries on a global scale, and often instigates multi-systemic afflictions encompassing neurological and cardiovascular impairments [ 2 ]. For the nervous system, acute CO poisoning in patients presents with symptoms such as dizziness, headache, nausea, vomiting, varying degrees of consciousness, and difficulty breathing. In the late stage, although most of the patients recovered after standard treatment, there are also some patients who still may experience neurological sequelae, namely delayed neurological sequelae (DNS) [ 3 – 5 ]. DNS emerges as an abrupt onset neurological disorder characterized by dementia, mental symptoms and extrapyramidal symptoms, typically appearing 2 to 40 days post regaining consciousness following acute CO exposure (average 22 days). In severe cases, DNS can prove fatal, with up to 50% of affected individuals encountering cognitive, neurological, or neurobehavioral sequelae [ 5 ]. Therefore, the early predictive capacity for DNS holds paramount clinical outcome, aiding in early intervention, clinical decision-making, and enhancing doctor-patient interactions. Previous studies showed that a range of traditional clinical indicators such as age onset, duration of exposure to CO, initial Glasgow coma scale, arterial HCO3-, white blood cell count, C-reactive protein, blood urea nitrogen, creatinine among others may be the risk factors of DNS [ 6 ]. In addition, imaging examinations play a crucial role in objectively and quantitatively assessing neurological damage and prognosis following acute CO poisoning, thereby aiding in early predicting the occurrence of DNS [ 5 ]. Magnetic Resonance Imaging (MRI) stands out as a widely utilized tool in diagnosing the acute phase of CO poisoning and DNS due to its superior brain tissue resolution. For example, during the acute phase of global cerebral ischemia and hypoxia, severe brain damage can occur, potentially leading to profound neurological dysfunction. Diffusion-weighted imaging (DWI) serves to characterize cytotoxic edema in damaged white matter with heightened sensitivity and early detection capabilities [ 5 ]. A retrospective study of 387 patients with acute CO poisoning revealed that 90% of elevated DWI signals correlated with reduced apparent diffusion coefficient (ADC), suggesting the cytotoxic edema's pivotal role in the pathogenesis of DNS and also demonstrated that individuals developing acute brain lesions on DWI, termed Acute Brain Lesions on DWI (ABLDs), faced a 14 times higher risk of subsequent DNS compared to those without ABLDs [ 6 ]. Therefore, DWI may be a valuable predictor of prognosis for acute CO poisoning. At present, a number of studies also showed that the combination of advanced age, low early GCS score, prolonged CO exposure duration, and abnormal DWI signal in acute stage serve as predictive factors for DNS, indicating a poorer long-term neurological prognosis [ 9 – 11 ]. Machine learning has been widely used in the early detection and prognosis prediction of various systemic diseases, such as the screening and early detection of breast cancer in order to control the development of disease as early as possible [ 12 ], as well as the individualized treatment of osteosarcoma patients and prognosis prediction through the detection of tumor microenvironment and its biomarkers in osteosarcoma metastasis [ 13 ]. At the same time, machine learning is also widely used in the diagnosis and treatment of neurological diseases. It not only helps in the diagnosis of neurological diseases such as stroke [ 14 ], but also plays an important role in the evaluation of disease treatment effect and prognosis prediction [ 15 , 16 ]. A previous study has shown the effectiveness of machine learning algorithms utilizing Random Forest Classifier (RFC) methods in predicting globus pallidus necrosis in patients with CO poisoning [ 17 ]. Building upon this, there is a keen interest in integrating multiple parameters to enhance the predictive capability of machine learning and evaluate its feasibility in forecasting DNS in CO poisoning patients. Another clinical toxicology article establishes a clinical-based Heart-Brain 346-7 Scoring system to help identify CO poisoning patients at higher risk of death by predicting hospitalization and long-term mortality [ 18 ]. The development of machine learning in medicine opens a new vision and adds a new mean for human beings to fight against diseases. Hence, the purposes of this study are twofold: (1) to identify certain DWI features that can predict DNS within DWI sequences during the acute phase of CO poisoning, and (2) to early predict the occurrence of DNS based on a combination of clinical and DWI features. Through machine learning algorithms, these biomarkers and indicators can improve prediction accuracy and contribute to the early recognition and management of DNS. 2. Materials and Methods 2.1 Study population This retrospective study analyzed the clinical data and DWI data of 120 patients with CO poisoning from January 8, 2018 to February 17, 2023 (Fig. 1 ). Inclusion criteria: 1) Ages ≥ 12 years old; 2) History of CO exposure; 3) Central nervous system injury occurs after poisoning; 4) The blood carboxyhemoglobin (COHb) level meets the diagnostic criteria. Exclusion criteria: 1) Without DWI test results or DWI examination more than 72 hours after CO exposure; 2) Patients with dementia, epilepsy, stroke, parkinson's disease, or other neurological diseases; 3) Patients who treated irregularly; 4) No follow-up data. 2.2 Clinical measures Age, gender, blood pressure, laboratory test data, Glasgow Coma Scale (GCS) scores, the initial vital signs, neuropsychiatric symptoms, positive signs, duration of exposure to CO were collected from medical records. Blood samples were collected within 24 hours after the onset of acute CO poisoning. The GCS score at presentation and subsequent clinical feature extraction were both assessed by at least two experienced independent clinicians. 2.3 Definitions DNS was defined as neurological symptoms that occurred within 60 days after discharge, including cognitive impairment, dysarthria, dyspraxia, motor deficits, parkinsonism, memory impairment, seizures, psychosis, and neuropsychological disorders [ 8 , 19 , 20 ]. The Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) can be used to assess general cognitive functioning and memory impairment. According to the 60-minute protocol proposed in the vascular cognitive impairment coordination standard, patients undergo neuropsychological assessment within 60 days of acute CO poisoning, including attention, processing speed, working memory, learning ability, and frontal lobe execution ability [ 21 ]. 2.4 DWI acquisition and DWI features The patients underwent MRI examination within 3 days of being exposed to high concentrations of CO. All MRI scans were acquired through a 3 Tesla MRI scanner (Achieva 3.0, Philips Medical Systems, Best, The Netherlands). DWI sequence parameters were as follows: 5mm slices; FOV (24×24); repetition time/echo time, (7891/87 ms); diffusion directions: all (b value, 1000 s/mm 2 , and b = 0 s/mm 2 ). All DWI images are independently analyzed by at least two radiologists. DWI images in the DICOM format were collected, and then the DICOM format was converted into BIDS format. The python version 3.6.5 was used for processing DWI data. First, the “BIDS Layout” package was utilized to load data from the specified BIDS format dataset into python. Second, the DWI data for each session was processed, including read b-values and b-vectors, as well as DWI images. Third, the fractional anisotropy (FA) and mean diffusivity (MD) were calculated by fitting the diffusion tensor model with the Tensor Model of “DIPY” packages. Finally, the regions of interests (ROI) analysis were carried out on the selected brain regions using the specific brain region map. The ROI were placed on the amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle. Then the average FA and MD of 14 ROIs including bilateral amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle were automatically calculated for each subject. The raters were blinded to the subject's details. 2.5 Machine learning model To early identify the DNS in acute CO poisoning, three support vector machine (SVM) models were established. The model was made up features selection, SVM classification, and model evaluation. 120 cases were divided into training (n = 96) and testing (n = 24) sets. 36 patients from the DNS group and 60 patients from non-DNS group were randomly selected as the training set to establish three SVM models, and the remaining 24 patients were selected as a test set to evaluate the model. A total of 24 clinical features and 28 features of the DWI sequence were compared between DNS and non-DNS groups. Features with p < 0.05 after the Monte Carlo correction for data were selected to establish the models in training set. Evaluate the predictive performance of the model using accuracy, precision, sensitivity, F1 score, macroscopic mean, weighted mean, area under the receiver operating characteristic curve (AUC), and 10-flod cross validation. The formulas were follows: $$\:Accuracy\left(A\right)=\left(\frac{TP+TN}{TP+FN+FP+TN}\right)$$ $$\:Precision\left(P\right)=\left(\frac{TP}{TP+FP}\right)$$ $$\:Sensitivity\left(S\right)=\left(\frac{TP}{TP+FN}\right)$$ $$\:F1-score=\left(\frac{2*P*S}{P+S}\right)$$ 2.6 Feature importance evaluation To identify the features the most significant features for the model, the mean impact value (MIV) was used to calculate and categorize the contribution of all MRI features [ 22 ]. After training the SVM model, each time the independent features increased or decreased by 10%, two new training sets was be obtained for fitting the model. Then, calculate the average feature difference between two simulation results based on MIV. Finally, the sequence of the features was sorted according to absolute MIVs, and identified potential lead features. 2.7 Statistical analysis Statistical analysis of the data was conducted using SPSS 25.0 software. Measurement data in mean ± standard deviation ( \(\:\stackrel{-}{\varvec{x}}\pm\:\mathbf{s}\) ) represents. Non normal distribution data is represented as median and quartile, and evaluated using Wilcoxon test. t -test is used for comparison between two groups, and count data is compared using t -test χ2-test or Fisher's exact probability method, using multiple repeated ANOVA at multiple time points between groups. P<0.05 was considered to indicate statistical significance. 3. Results 3.1 Clinical characteristics This study included a total of 605 newly diagnosed patients with acute CO poisoning. After further screening, a total of 120 patients were included in the retrospective study, of which 75 were assigned to the non-DNS cohort and 45 were assigned to the DNS cohort after at least 60-day follow-up based on the definitions of DNS (Fig. 1 ) [ 8 , 19 , 20 ]. The statistical analysis of clinical predictive factors in the multivariate model showed that 8 out of 24 variables met the threshold of p < 0.01, including age, carbon monoxide exposure time, initial GCS score, monocytes, LDH, arterial HCO3-, COHb, and lactate (Table 1 ). Table 1 Clinical and demographic characteristics in our study. Variable Non-DNS group (n = 75) DNS group (n = 45) P value Gender 0.22 Male 42(56%) 20(44.44%) Female 33(64%) 25(55.56%) Age, years 47.75 ± 20.13 56.47 ± 17.10 0.017 CO exposure time, hours 1.98 ± 1.42 3.65 ± 3.66 0.001 Initial GCS score 14.08 ± 1.89 12.49 ± 3.34 0.001 Diabetes 0.925 Yes 2(2.67%) 1(2.22%) No 73(97.33%) 44(97.78%) Hypertension 0.394 Yes 12(16%) 10(22.22%) No 63(84%) 35(77.78%) Systolic blood pressure, mmHg 130.73 ± 20.81 128.44 ± 18.37 0.544 Diastolic blood pressure, mmHg 89.55 ± 95.57 79.00 ± 11.95 0.463 Leukocyte, 10⁹/L 8.31 ± 3.65 9.54 ± 5.44 0.138 Neutrophile granulocyte, x10⁹/L 6.42 ± 3.78 7.73 ± 5.51 0.126 Leukomonocyte, 10⁹/L 1.38 ± 0.65 1.23 ± 0.61 0.211 Monocyte, 10⁹/L 0.40 ± 0.19 0.49 ± 0.24 0.032 Blood platelet, 10⁹/L 205 ± 79 209 ± 66 0.794 BUN, mmol/L 6.79 ± 2.87 6.62 ± 2.83 0.725 Cr, µmol/L 65.9 ± 31.8 75.6 ± 30.5 0.103 CK, U/L 742.5 ± 2736.8 2280.1 ± 5895.1 0.055 CK-MB, U/L 41.4 ± 110.6 103.5 ± 267.5 0.078 LDH, U/L 233.7 ± 153.1 309.1 ± 252.3 0.043 HBD, U/L 163.7 ± 109.5 211.3 ± 154.9 0.052 CRP, mg/L 14.78 ± 39.45 33.13 ± 62.93 0.052 Arterial pH 7.41 ± 0.68 7.40 ± 0.61 0.158 Arterial HCO3-, mmol/L 22.44 ± 2.67 20.47 ± 3.65 0.001 COHb, % 6.41 ± 7.57 10.21 ± 11.00 0.027 Lactic acid, mmol/L 2.18 ± 1.31 3.62 ± 2.43 < 0.001 Categorical variables are expressed as numbers (%) and continuous variables are expressed as means ± standard deviations. P -value < 0.05 was considered statistically significant. Abbreviations: DNS: Delayed neurological sequelae; CO: Carbon monoxide; GCS score: Glasgow Coma Scale score; BUN: Blood urea nitrogen; Cr: Creatinine; CK: Creatine kinase; CK-MB: Creatine kinase-MB; LDH: Lactate dehydrogenase; HBD: Hydroxy butyrate dehydrogenase; CRP: C-reactive protein; COHb: Carboxyhemoglobin. 3.2 DWI characteristics After analyzing DWI features, we found there were very high predictive value in 23 of 28 radiographic predictors, including the mean FA of left amygdala, right amygdala, left caudate, right caudate, left hippocampus, right hippocampus, left pallidus, right pallidus, left putamen, right putamen, left thalamus and right thalamus, and the mean MD of left amygdala, right amygdala, left caudate, left hippocampus, right hippocampus, left pallidus, right pallidus, left putamen, right putamen, left thalamus and right thalamus, which were all significantly correlated with the prediction efficiency (Table 2 ). Finally, we jointly established SVM model based on clinical characteristics and DWI characteristics as research objects. This model holds paramount importance in its capabilities for predicting the onset of DNS. Table 2 DWI characteristics in our study. Features Non-DNS group (n = 75) DNS group (n = 45) P value Mean FA of left amygdala 0.43(0.45 ± 0.30) 0.78(0.75 ± 0.31) <0.001 Mean MD of left amygdala 0.58(0.55 ± 0.28) 0.66(0.69 ± 0.29) 0.012 Mean FA of right amygdala 0.45(0.47 ± 0.25) 0.67(0.70 ± 0.32) <0.001 Mean MD of right amygdala 0.51(0.53 ± 0.29) 0.75(0.71 ± 0.31) 0.002 Mean FA of left caudate 0.43(0.49 ± 0.29) 0.88(0.85 ± 0.25) <0.001 Mean MD of left caudate 0.53(0.53 ± 0.29) 0.70(0.65 ± 0.25) 0.026 Mean FA of right caudate 0.48(0.50 ± 0.29) 0.81(0.76 ± 0.30) <0.001 Mean MD of right caudate 0.65(0.64 ± 0.30) 0.66(0.63 ± 0.30) 0.226 Mean FA of left hippocampus 0.52(0.52 ± 0.29) 0.76(0.76 ± 0.33) <0.001 Mean MD of left hippocampus 0.41(0.46 ± 0.30) 0.74(0.67 ± 0.33) <0.001 Mean FA of right hippocampus 0.42(0.44 ± 0.26) 0.66(0.71 ± 0.31) <0.001 Mean MD of right hippocampus 0.57(0.57 ± 0.27) 0.74(0.73 ± 0.26) 0.002 Mean FA of left pallidus 0.52(0.51 ± 0.29) 0.65(0.69 ± 0.32) 0.002 Mean MD of left pallidus 0.45(0.46 ± 0.29) 0.60(0.66 ± 0.30) <0.001 Mean FA of right pallidus 0.43(0.48 ± 0.30) 0.82(0.79 ± 0.32) <0.001 Mean MD of right pallidus 0.55(0.52 ± 0.31) 0.76(0.74 ± 0.29) <0.001 Mean FA of left putamen 0.40(0.48 ± 0.29) 0.58(0.68 ± 0.31) <0.001 Mean MD of left putamen 0.47(0.48 ± 0.27) 0.74(0.70 ± 0.27) <0.001 Mean FA of right putamen 0.51(0.51 ± 0.30) 0.79(0.78 ± 0.31) <0.001 Mean MD of right putamen 0.46(0.46 ± 0.27) 0.56(0.65 ± 0.29) <0.001 Mean FA of left thalamus 0.44(0.46 ± 0.30) 0.86(0.76 ± 0.34) <0.001 Mean MD of left thalamus 0.40(0.41 ± 0.26) 0.70(0.71 ± 0.25) <0.001 Mean FA of right thalamus 0.55(0.50 ± 0.30) 0.67(0.71 ± 0.29) <0.001 Mean MD of right thalamus 0.47(0.48 ± 0.28) 0.65(0.66 ± 0.30) 0.001 Mean FA of left ventricle 0.52(0.52 ± 0.24) 0.61(0.56 ± 0.29) 0.076 Mean MD of left ventricle 0.41(0.46 ± 0.29) 0.55(0.54 ± 0.28) 0.155 Mean FA of right ventricle 0.47(0.51 ± 0.30) 0.46(0.49 ± 0.28) 0.675 Mean MD of right ventricle 0.36(0.45 ± 0.29) 0.37(0.43 ± 0.29) 0.692 Continuous variables are expressed as means ± standard deviations. P -value < 0.05 was considered statistically significant. Abbreviations: DNS: Delayed neurological sequelae. FA: Fractional Anisotropy; MD: Mean Diffusivity. 3.3 SVM model We established three SVM models that based on only clinical features (clinical model), only DWI features (DWI model), combined clinical and DWI features (clinical-DWI model), whose predictive accuracies of DNS were different. The prediction accuracy of the SVM model established solely based on clinical features is 0.76 [95% CI 0.67–0.82]. The accuracy of the SVM model based solely on DWI features is 0.94 [95% CI 0.88–0.98]. The accuracy of the SVM model based on combined clinical and DWI features is 0.97 [95% CI 0.94–1.00]. In addition, the performance of the combined clinical and DWI features model is higher than that of clinical or DWI feature models, with a precision of 1.00 [95% CI 0.95–1.00], sensitivity of 0.92 [95% CI 0.86–0.98], F1 score of 0.96 [95% CI 0.92–1.00], macroscopic mean of 0.98 [95% CI 0.0.83–1.00], weighted mean of 0.97 [95% CI 0.0.93–1.00], and AUC of 0.97 [95% CI 0.94–1.00], and the result of 10-flod cross validation was 0.98 [95% CI 0.92–1.00] (Table 3 and Fig. 2 ). The mean FA of left caudate had the highest impact on the SVM model based on MIV analysis (Table 4 ). Table 3 The performance of three models for identifying delayed encephalopathy after acute carbon monoxide poisoning. AC PR SE F1-score MA WA AUC 10-flod cross validation Clinical model Training 0.78 0.60 0.67 0.65 0.70 0.74 0.76 0.72 Test 0.76 0.62 0.67 0.64 0.72 0.75 0.76 DWI model Training 0.92 0.92 0.95 0.90 0.89 0.93 0.92 0.93 Test 0.94 0.92 0.92 0.92 0.94 0.94 0.94 Clinical-DWI model Training 0.94 0.94 0.91 0.95 0.99 0.97 0.98 0.98 Test 0.97 1.00 0.92 0.96 0.98 0.97 0.97 DWI: Diffusion-weighted imaging; AC: Accuracy; PR: Precision; SE: Sensitivity; MA: Macro-average; WA: Weighted-average; AUC: Area under the curve. Table 4 The feature importance sequence of MIV. Ranking Features MIV 1 Mean FA of left caudate 0.208 2 Mean FA of left thalamus 0.192 3 Mean MD of left thalamus 0.192 4 Mean FA of right pallidus 0.175 5 Mean FA of left amygdala 0.163 6 Lactic acid 0.158 7 Mean FA of right caudate 0.158 8 Mean FA of right hippocampus 0.158 9 Mean MD of right putamen 0.158 10 Arterial HCO3- 0.146 11 Mean FA of left pallidus 0.142 12 Mean MD of left putamen 0.142 13 Mean FA of right putamen 0.142 14 Mean FA of left hippocampus 0.129 15 COHb 0.128 16 Mean FA of right amygdala 0.125 17 Mean FA of left putamen 0.125 18 CO exposure time 0.115 19 LDH 0.113 20 Initial GCS score 0.108 21 Mean MD of right amygdala 0.108 22 Mean MD of left hippocampus 0.108 23 Mean MD of right hippocampus 0.108 24 Mean MD of right pallidus 0.108 25 Mean MD of right thalamus 0.108 26 Mean FA of right thalamus 0.104 27 Mean MD of left amygdala 0.092 28 Mean MD of left caudate 0.092 29 Mean MD of left pallidus 0.092 30 Monocyte 0.086 31 Age 0.074 MIV: Mean impact value; FA: Fractional Anisotropy; MD: Mean Diffusivity; COHb: carboxyhemoglobin; LDH: lactate dehydrogenase; GCS score: Glasgow Coma Scale score 4. Discussion Numerous predictors based on clinical data and statistical analyses have been identified in previous studies regarding CO poisoning. Lactic acid level was emerged as a significant predictor, as evidenced by a p < 0.001 in our study, indicating its high predictive value. Elevated lactate levels, indicative of tissue hypoxia, are commonly utilized in intensive care units as a reliable prognostic factor for critically ill patients [ 23 , 24 ]. Studies have suggested that initial blood lactate levels may be associated with patient prognosis [ 25 ] and the severity of CO poisoning [ 26 , 27 ]. Zhang et al. has highlighted that a longer duration of CO exposure and lower GCS scores upon arrival are independent predictors of DNS following CO poisoning [ 28 ]. Pepe et al. also demonstrated that a CO exposure duration exceeding 6 hours may elevate the risk of DNS development [ 29 ]. These studies collectively underscore the critical role of CO exposure duration and initial GCS score in the progression of delayed neurological complications. At the same time, some prospective studies have consistently identified a GCS score of less than 9 as a crucial predictor of DNS [ 29 , 30 ]. The GCS score serves as an objective reflection of the patient's level of consciousness, with lower scores indicating a poorer prognosis. In our study, we get the same result, as evidenced by the significant correlation ( p < 0.001) between the duration of CO exposure, initial GCS score, and prognosis in CO poisoning cases. DWI conducted during the acute phase of CO poisoning can offer valuable insights into predicting long-term neurological outcomes post-discharge [ 9 ]. Previous studies have demonstrated that MRI-DWI scans revealing acute ischemic brain lesions are closely linked to the development of DNS. In particular, the pallidum emerges as the most commonly affected region in DWI scans, with other frequently impacted areas including the cerebellum, hippocampus, putamen, amygdala and corpus callosum—all of which are associated with the occurrence of DNS and often involved in both sides simultaneously [ 7 , 31 ]. Our study delved into the extraction of FA and MD, two commonly used research parameters, revealing significant associations with the occurrence of DNS in the later stages. These findings underscore the potential of DWI in identifying individuals at risk of DNS and enabling targeted interventions to mitigate adverse neurological sequelae. Although many predecessors have explored clinical features in the context of acute carbon monoxide poisoning, however, there remains a paucity of research focusing on the concurrent extraction of DWI features for the development of predictive models. Existing investigations that integrate clinical and imaging data have primarily extracted features such as MRI, and diffusion tensor imaging, while many analyses of DWI data have been limited to generic descriptions of high signal lesions or the presence of abnormalities in DWI-MRI scans, with minimal emphasis on delineating specific brain regions and parameter alterations. The lack of comprehensive feature extraction poses a challenge in achieving a nuanced understanding of the disease pathophysiology [ 8 ]. In contrast, our study presents a novel approach by integrating clinical and DWI features to formulate a SVM model for prognostic evaluation in acute CO poisoning. There are some advantages of our method. First, we combined the clinical features and DWI features to construct a SVM model with good performance for the prognosis assessment of acute CO poisoning and guide the subsequent prevention and treatment, which can be easily obtained from clinical records and admission examinations. Second, our study counted 605 cases of acute CO poisoning from 2018 to 2023, and the remaining 120 cases met the inclusion criteria after layers of screening, which is a unique study with a large sample size for this regional disease. Third, the MIV of clinically significant variables can reflect the correlation between clinical and imaging features and prognosis of our study clearly and intuitively, enhancing the interpretability of our findings. However, there are several limitations in our study. First, this study is a single-center design based on retrospective data, so the results may not be generalized to other centers, and further multi-center studies are needed to verify and expand the practical application value for DNS prediction. Second, the variability in CO exposure concentrations, variations in pre-hospital interventions, and discrepancies in the duration of oxygen therapy among individual patients poses potential confounders that could impact prognostic outcomes. Finally, when we took DWI features, we tried to extract and analyze the average diffusion coefficient (ADC), signal intensity ratio (SIR) and other factors commonly extracted and analyzed in the central system. However, we are only satisfied with the parameters of FA and MD in certain ROIs indicating a limited scope in parameter selection within the DWI domain. This restricted parameter set highlights the potential for further scalability in DWI analysis to encompass a broader array of relevant factors for a more comprehensive evaluation of DNS prognosis. 5. Conclusions We demonstrated the changes between DNS group and non-DNS group in FA and MD in DWI during the acute phase of CO poisoning, and the utilization of DWI can be instrumental in identifying individuals at risk of developing DNS. A SVM model that combines multi-modal features including DWI features and clinical parameters, has been proven to effectively predict DNS episodes in patients with CO poisoning. Furthermore, the model's practicality in a clinical setting enhances its value as a tool for early intervention and tailored management strategies for patients at risk of DNS following CO poisoning. Abbreviations CO Carbon monoxide DNS Delayed neurological sequelae MRI Magnetic Resonance Imaging DWI Diffusion-weighted imaging ADC Apparent diffusion coefficient ABLDs Brain Lesions on DWI RFC Random Forest Classifier COHb Carboxyhemoglobin GCS Glasgow Coma Scale MMSE Mini-Mental State Examination MoCA Montreal Cognitive Assessment FA Fractional anisotropy MD Mean diffusivity SVM Support vector machine AUC Area under the receiver operating characteristic curve MIV Mean impact value Declarations Acknowledgment We would also like to thank our staff, who assisted in the data collection and analysis. Author Contributions CSY, YSJ, THM and LHY were joint first authors; LQH was correspondence author; CSY, YSJ and LQH designed the study; TMH and LHY reviewed the literature; LJL and LH collected the data; TMH and LHY performed the follow-up activity; YSJ performed the statistical analysis; CSY and YSJ wrote the manuscript; LJL acquired fund; YXQ, HYZ, TSB and LQH revised the manuscript. All authors have read and approved the final manuscript. Funding This study was sponsored by the Enshi Tujia and Miao Autonomous Prefecture Science and Technology Bureau Project in 2023 (Authorization number: D20230077). Data availability The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study. Ethics Approval and Consent to Participate This study was reviewed and approved by the Ethics Committee of the Central Hospital of the Enshi Prefecture, with ethics approval reference (2023-011-02). All patients and their families have signed informed consent forms. Consent for publication Not applicable. Clinical trial number Not applicable. Conflict of Interest The authors declare that they have no conflict of interest. References Inagaki T, Ishino H, Seno H, et al. A long-term follow-up study of serial magnetic resonance images in patients with delayed encephalopathy after acute carbon monoxide poisoning. J Neuropsychiatry Clin Neurosci. 1997;51(6):421–3. Choi IS. Carbon monoxide poisoning: systemic manifestations and complications. J Korean Med Sci. 2001;16:253–61. Weaver LK. Clinical practice. Carbon monoxide poisoning. N Engl J Med. 2009;360(12):1217–25. Ernst A, Zibrak JD. Carbon monoxide poisoning. N Engl J Med. 1998;339(22):1603–8. Beppu T. The role of MR imaging in assessment of brain damage from carbon monoxide poisoning: a review of the literature. AJNR Am J Neuroradiol. 2014;35(4):625–31. Moon JM, Chun BJ, Cho YS. The predictive value of scores based on peripheral complete blood cell count for long-term neurological outcome in acute carbon monoxide intoxication. Volume 124. Basic & clinical pharmacology & toxicology; 2019. pp. 500–10. 4. Sönksen SE, Kühn S, Basner M, et al. Brain structure and neurocognitive function in two professional mountaineers during 35 days of severe normobaric hypoxia. Eur J Neurol. 2022;29(10):3112–6. Jeon SB, Sohn CH, Seo DW, et al. Acute Brain Lesions on Magnetic Resonance Imaging and Delayed Neurological Sequelae in Carbon Monoxide Poisoning. JAMA Neurol. 2018;75(4):436–43. Sert ET, Kokulu K, Mutlu H. Clinical predictors of delayed neurological sequelae in charcoal-burning carbon monoxide poisoning. Am J Emerg Med. 2021;48:12–7. Moon JM, Chun BJ, Baek BH, et al. Initial diffusion-weighted MRI and long-term neurologic outcomes in charcoal-burning carbon monoxide poisoning. Clin Toxicol (Phila). 2018;56(3):161–9. Kim YS, Cha YS, Kim MS, et al. The usefulness of diffusion-weighted magnetic resonance imaging performed in the acute phase as an early predictor of delayed neuropsychiatric sequelae in acute carbon monoxide poisoning. Hum Exp Toxicol. 2018;37(6):587–95. Ghorbian M, Ghorbian S. Usefulness of machine learning and deep learning approaches in screening and early detection of breast cancer. Heliyon. 2023;9(12):e22427. Liu G, Wang S, Liu J, et al. Using machine learning methods to study the tumour microenvironment and its biomarkers in osteosarcoma metastasis. Heliyon. 2024;10(7):e29322. Miyamoto N, Ueno Y, Yamashiro K, et al. Stroke classification and treatment support system artificial intelligence for usefulness of stroke diagnosis. Front Neurol. 2023;14:1295642. Methods In Medicine CAM. Retracted: Artificial Intelligence Algorithm-Based MRI in Evaluating the Treatment Effect of Acute Cerebral Infarction. Computational and mathematical methods in medicine. 2023, 9794357. Zhang L, Li Y, Bian L, et al. Cognitive Impairment of Patient With Neurological Cerebrovascular Disease Using the Artificial Intelligence Technology Guided by MRI. Front public health. 2021;9:813641. Chan MJ, Hu CC, Huang WH, et al. An artificial intelligence algorithm for analyzing globus pallidus necrosis after carbon monoxide intoxication. Hum Exp Toxicol. 2023;42:9603271231190906. Rose JJ, Zhang MS, Pan J, et al. Heart-Brain 346-7 Score: the development and validation of a simple mortality prediction score for carbon monoxide poisoning utilizing deep learning. Clin Toxicol (Phila). 2023;61(7):492–9. Kokulu K, Mutlu H, Sert ET. Serum netrin-1 levels at presentation and delayed neurological sequelae in unintentional carbon monoxide poisoning. Clin Toxicol (Phila). 2020;58(12):1313–9. Kim YJ, Sohn CH, Seo DW, et al. Clinical Predictors of Acute Brain Injury in Carbon Monoxide Poisoning Patients With Altered Mental Status at Admission to Emergency Department. Acad Emerg medicine: official J Soc Acad Emerg Med. 2019;26(1):60–7. Hachinski V, Iadecola C, Petersen RC, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke. 2006;37(9):2220–41. Jiang JL, Su X, Zhang H, et al. A novel approach to active compounds identification based on support vector regression model and mean impact value. Volume 81. Chemical biology & drug design; 2013. pp. 650–7. 5. Bernardin G, Pradier C, Tiger F, et al. Blood pressure and arterial lactate level are early indicators of short-term survival in human septic shock. Intensive Care Med. 1996;22(1):17–25. Abramson D, Scalea TM, Hitchcock R et al. Lactate clearance and survival following injury. The Journal of trauma, 1993, 35(4): 584-8; discussion 8–9. Inoue S, Saito T, Tsuji T, et al. Lactate as a prognostic factor in carbon monoxide poisoning: a case report. Am J Emerg Med. 2008;26(8):e9661–3. Benaissa ML, Mégarbane B, Borron SW, et al. Is elevated plasma lactate a useful marker in the evaluation of pure carbon monoxide poisoning? Intensive Care Med. 2003;29(8):1372–5. Sokal JA, Kralkowska E. The relationship between exposure duration, carboxyhemoglobin, blood glucose, pyruvate and lactate and the severity of intoxication in 39 cases of acute carbon monoxide poisoning in man. Arch Toxicol. 1985;57(3):196–9. Zhang Y, Lu Q, Jia J, et al. Multicenter retrospective analysis of the risk factors for delayed neurological sequelae after acute carbon monoxide poisoning. Am J Emerg Med. 2021;46:165–9. Pepe G, Castelli M, Nazerian P, et al. Delayed neuropsychological sequelae after carbon monoxide poisoning: predictive risk factors in the Emergency Department. A retrospective study. Scand J Trauma Resusc Emerg Med. 2011;19:16. Han S, Choi S, Nah S, et al. Cox regression model of prognostic factors for delayed neuropsychiatric sequelae in patients with acute carbon monoxide poisoning: A prospective observational study. Neurotoxicology. 2021;82:63–8. Kim DM, Lee IH, Park JY, et al. Acute carbon monoxide poisoning: MR imaging findings with clinical correlation. Diagn Interv Imaging. 2017;98(4):299–306. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 18 Nov, 2024 Editor assigned by journal 15 Nov, 2024 Submission checks completed at journal 15 Nov, 2024 First submitted to journal 12 Nov, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5443111","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":379240705,"identity":"eaac3c23-03fd-465f-94f0-729185b29a80","order_by":0,"name":"Siying Chen","email":"","orcid":"","institution":"Hubei Minzu University, The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Siying","middleName":"","lastName":"Chen","suffix":""},{"id":379240706,"identity":"d8f5da87-9afd-49be-8da5-7e01dd2a32c1","order_by":1,"name":"Shijun Yang","email":"","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Shijun","middleName":"","lastName":"Yang","suffix":""},{"id":379240708,"identity":"201c8e70-d379-432a-a12a-b372a8fbbbf6","order_by":2,"name":"Minghui Tan","email":"","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Minghui","middleName":"","lastName":"Tan","suffix":""},{"id":379240709,"identity":"bce2f430-0499-4104-8d34-405f8cfdbfd8","order_by":3,"name":"Heying Lu","email":"","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Heying","middleName":"","lastName":"Lu","suffix":""},{"id":379240710,"identity":"c5663e51-1768-45f1-bc3c-c9a71b75da48","order_by":4,"name":"Huan Li","email":"","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Huan","middleName":"","lastName":"Li","suffix":""},{"id":379240711,"identity":"2eda2c32-5145-4ac9-8fa5-6748239667e0","order_by":5,"name":"Jinlan Li","email":"","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":false,"prefix":"","firstName":"Jinlan","middleName":"","lastName":"Li","suffix":""},{"id":379240712,"identity":"56d9c2be-2018-40fa-ae71-e278b3997fdf","order_by":6,"name":"Xiuqiong Yang","email":"","orcid":"","institution":"The Xian Feng County People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xiuqiong","middleName":"","lastName":"Yang","suffix":""},{"id":379240713,"identity":"1ae6d5c8-5fe7-4018-a829-3fcec0ffb35e","order_by":7,"name":"Yanzhi Huang","email":"","orcid":"","institution":"The Xuan En County People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yanzhi","middleName":"","lastName":"Huang","suffix":""},{"id":379240714,"identity":"3dcb3181-ce51-4e3f-9dd4-7c6c11d4a8a9","order_by":8,"name":"Senbiao Tian","email":"","orcid":"","institution":"The Xian Feng Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Senbiao","middleName":"","lastName":"Tian","suffix":""},{"id":379240715,"identity":"678529af-aa35-47d4-a5eb-2bf666589bc1","order_by":9,"name":"Qunhui Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3ElEQVRIiWNgGAWjYBACPmYwdYCBjb358IMPFRJy/IS0sMG18BxLM5xxxsJYsoGQFgaoFgYJHwNp3raKxA0EtbDzGN7m3XEnsU+Cx8Bw5jwJxg0MzA8f3cDrMB5ja94zzxLbpNsKHnzcJsFszsBmbJyDX4sZ0D2HE9tkDm8wnLlNgs2ygYdNmjgtEglAv8wBOu8A8VpSgFoaJCSI0MJWbDm37bBxGziQj0kYSDYT8As//+GNN962HZad3w6Kypq6+n725oeP8WkBAQlULjMB5Vi0jIJRMApGwShAAwA5uUZ/cPDVNwAAAABJRU5ErkJggg==","orcid":"","institution":"The Central Hospital of Enshi Tujia and Miao Autonomous Prefecture","correspondingAuthor":true,"prefix":"","firstName":"Qunhui","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2024-11-13 02:53:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5443111/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5443111/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":70481007,"identity":"d965b2aa-3227-4729-8189-0defd2a5b5aa","added_by":"auto","created_at":"2024-12-03 14:54:05","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":182397,"visible":true,"origin":"","legend":"\u003cp\u003eFlow chart.\u003c/p\u003e\n\u003cp\u003eDWI: diffusion-weighted imaging; ACOP: acute carbon monoxide poisoning; DNS: delayed neurological sequelae; SVM: support vector machine.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5443111/v1/811316da2d40edc3322e3d6b.jpeg"},{"id":70481006,"identity":"4be3497b-dce0-4f02-a73b-823635b69f77","added_by":"auto","created_at":"2024-12-03 14:54:05","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":83311,"visible":true,"origin":"","legend":"\u003cp\u003eROC of the three SVM models in validation cohort.\u003c/p\u003e\n\u003cp\u003eROC: Receiver Operating Characteristic; SVM: Support Vector Machine; DWI: diffusion-weighted imaging.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-5443111/v1/1ed10e4f5305472e2a996120.png"},{"id":70481454,"identity":"6f107cb8-8b50-47a7-ba06-b7f5937fdb70","added_by":"auto","created_at":"2024-12-03 15:02:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1032400,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5443111/v1/0d62e391-2bd7-4b54-b88b-5e1f6c3851b7.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identifying delayed neurological sequelae during the acute phase of carbon monoxide poisoning based on diffusion-weighted imaging and clinical features","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eCarbon monoxide (CO) is a kind of colorless and odorless poisonous gas, which exists widely in environment [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Acute CO poisoning stands as a significant contributor to both accidental and intentional injuries on a global scale, and often instigates multi-systemic afflictions encompassing neurological and cardiovascular impairments [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. For the nervous system, acute CO poisoning in patients presents with symptoms such as dizziness, headache, nausea, vomiting, varying degrees of consciousness, and difficulty breathing. In the late stage, although most of the patients recovered after standard treatment, there are also some patients who still may experience neurological sequelae, namely delayed neurological sequelae (DNS) [\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. DNS emerges as an abrupt onset neurological disorder characterized by dementia, mental symptoms and extrapyramidal symptoms, typically appearing 2 to 40 days post regaining consciousness following acute CO exposure (average 22 days). In severe cases, DNS can prove fatal, with up to 50% of affected individuals encountering cognitive, neurological, or neurobehavioral sequelae [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Therefore, the early predictive capacity for DNS holds paramount clinical outcome, aiding in early intervention, clinical decision-making, and enhancing doctor-patient interactions.\u003c/p\u003e \u003cp\u003ePrevious studies showed that a range of traditional clinical indicators such as age onset, duration of exposure to CO, initial Glasgow coma scale, arterial HCO3-, white blood cell count, C-reactive protein, blood urea nitrogen, creatinine among others may be the risk factors of DNS [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. In addition, imaging examinations play a crucial role in objectively and quantitatively assessing neurological damage and prognosis following acute CO poisoning, thereby aiding in early predicting the occurrence of DNS [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMagnetic Resonance Imaging (MRI) stands out as a widely utilized tool in diagnosing the acute phase of CO poisoning and DNS due to its superior brain tissue resolution. For example, during the acute phase of global cerebral ischemia and hypoxia, severe brain damage can occur, potentially leading to profound neurological dysfunction. Diffusion-weighted imaging (DWI) serves to characterize cytotoxic edema in damaged white matter with heightened sensitivity and early detection capabilities [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. A retrospective study of 387 patients with acute CO poisoning revealed that 90% of elevated DWI signals correlated with reduced apparent diffusion coefficient (ADC), suggesting the cytotoxic edema's pivotal role in the pathogenesis of DNS and also demonstrated that individuals developing acute brain lesions on DWI, termed Acute Brain Lesions on DWI (ABLDs), faced a 14 times higher risk of subsequent DNS compared to those without ABLDs [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Therefore, DWI may be a valuable predictor of prognosis for acute CO poisoning. At present, a number of studies also showed that the combination of advanced age, low early GCS score, prolonged CO exposure duration, and abnormal DWI signal in acute stage serve as predictive factors for DNS, indicating a poorer long-term neurological prognosis [\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMachine learning has been widely used in the early detection and prognosis prediction of various systemic diseases, such as the screening and early detection of breast cancer in order to control the development of disease as early as possible [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], as well as the individualized treatment of osteosarcoma patients and prognosis prediction through the detection of tumor microenvironment and its biomarkers in osteosarcoma metastasis [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. At the same time, machine learning is also widely used in the diagnosis and treatment of neurological diseases. It not only helps in the diagnosis of neurological diseases such as stroke [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], but also plays an important role in the evaluation of disease treatment effect and prognosis prediction [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. A previous study has shown the effectiveness of machine learning algorithms utilizing Random Forest Classifier (RFC) methods in predicting globus pallidus necrosis in patients with CO poisoning [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Building upon this, there is a keen interest in integrating multiple parameters to enhance the predictive capability of machine learning and evaluate its feasibility in forecasting DNS in CO poisoning patients. Another clinical toxicology article establishes a clinical-based Heart-Brain 346-7 Scoring system to help identify CO poisoning patients at higher risk of death by predicting hospitalization and long-term mortality [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The development of machine learning in medicine opens a new vision and adds a new mean for human beings to fight against diseases.\u003c/p\u003e \u003cp\u003eHence, the purposes of this study are twofold: (1) to identify certain DWI features that can predict DNS within DWI sequences during the acute phase of CO poisoning, and (2) to early predict the occurrence of DNS based on a combination of clinical and DWI features. Through machine learning algorithms, these biomarkers and indicators can improve prediction accuracy and contribute to the early recognition and management of DNS.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study population\u003c/h2\u003e \u003cp\u003eThis retrospective study analyzed the clinical data and DWI data of 120 patients with CO poisoning from January 8, 2018 to February 17, 2023 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Inclusion criteria: 1) Ages\u0026thinsp;\u0026ge;\u0026thinsp;12 years old; 2) History of CO exposure; 3) Central nervous system injury occurs after poisoning; 4) The blood carboxyhemoglobin (COHb) level meets the diagnostic criteria. Exclusion criteria: 1) Without DWI test results or DWI examination more than 72 hours after CO exposure; 2) Patients with dementia, epilepsy, stroke, parkinson's disease, or other neurological diseases; 3) Patients who treated irregularly; 4) No follow-up data.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Clinical measures\u003c/h2\u003e \u003cp\u003eAge, gender, blood pressure, laboratory test data, Glasgow Coma Scale (GCS) scores, the initial vital signs, neuropsychiatric symptoms, positive signs, duration of exposure to CO were collected from medical records. Blood samples were collected within 24 hours after the onset of acute CO poisoning. The GCS score at presentation and subsequent clinical feature extraction were both assessed by at least two experienced independent clinicians.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Definitions\u003c/h2\u003e \u003cp\u003eDNS was defined as neurological symptoms that occurred within 60 days after discharge, including cognitive impairment, dysarthria, dyspraxia, motor deficits, parkinsonism, memory impairment, seizures, psychosis, and neuropsychological disorders [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) can be used to assess general cognitive functioning and memory impairment. According to the 60-minute protocol proposed in the vascular cognitive impairment coordination standard, patients undergo neuropsychological assessment within 60 days of acute CO poisoning, including attention, processing speed, working memory, learning ability, and frontal lobe execution ability [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 DWI acquisition and DWI features\u003c/h2\u003e \u003cp\u003eThe patients underwent MRI examination within 3 days of being exposed to high concentrations of CO. All MRI scans were acquired through a 3 Tesla MRI scanner (Achieva 3.0, Philips Medical Systems, Best, The Netherlands). DWI sequence parameters were as follows: 5mm slices; FOV (24\u0026times;24); repetition time/echo time, (7891/87 ms); diffusion directions: all (b value, 1000 s/mm\u003csup\u003e2\u003c/sup\u003e, and b\u0026thinsp;=\u0026thinsp;0 s/mm\u003csup\u003e2\u003c/sup\u003e). All DWI images are independently analyzed by at least two radiologists.\u003c/p\u003e \u003cp\u003eDWI images in the DICOM format were collected, and then the DICOM format was converted into BIDS format. The python version 3.6.5 was used for processing DWI data. First, the \u0026ldquo;BIDS Layout\u0026rdquo; package was utilized to load data from the specified BIDS format dataset into python. Second, the DWI data for each session was processed, including read b-values and b-vectors, as well as DWI images. Third, the fractional anisotropy (FA) and mean diffusivity (MD) were calculated by fitting the diffusion tensor model with the Tensor Model of \u0026ldquo;DIPY\u0026rdquo; packages. Finally, the regions of interests (ROI) analysis were carried out on the selected brain regions using the specific brain region map. The ROI were placed on the amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle. Then the average FA and MD of 14 ROIs including bilateral amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle were automatically calculated for each subject. The raters were blinded to the subject's details.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Machine learning model\u003c/h2\u003e \u003cp\u003eTo early identify the DNS in acute CO poisoning, three support vector machine (SVM) models were established. The model was made up features selection, SVM classification, and model evaluation.\u003c/p\u003e \u003cp\u003e120 cases were divided into training (n\u0026thinsp;=\u0026thinsp;96) and testing (n\u0026thinsp;=\u0026thinsp;24) sets. 36 patients from the DNS group and 60 patients from non-DNS group were randomly selected as the training set to establish three SVM models, and the remaining 24 patients were selected as a test set to evaluate the model. A total of 24 clinical features and 28 features of the DWI sequence were compared between DNS and non-DNS groups. Features with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 after the Monte Carlo correction for data were selected to establish the models in training set.\u003c/p\u003e \u003cp\u003eEvaluate the predictive performance of the model using accuracy, precision, sensitivity, F1 score, macroscopic mean, weighted mean, area under the receiver operating characteristic curve (AUC), and 10-flod cross validation. The formulas were follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Accuracy\\left(A\\right)=\\left(\\frac{TP+TN}{TP+FN+FP+TN}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:Precision\\left(P\\right)=\\left(\\frac{TP}{TP+FP}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:Sensitivity\\left(S\\right)=\\left(\\frac{TP}{TP+FN}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:F1-score=\\left(\\frac{2*P*S}{P+S}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Feature importance evaluation\u003c/h2\u003e \u003cp\u003eTo identify the features the most significant features for the model, the mean impact value (MIV) was used to calculate and categorize the contribution of all MRI features [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. After training the SVM model, each time the independent features increased or decreased by 10%, two new training sets was be obtained for fitting the model. Then, calculate the average feature difference between two simulation results based on MIV. Finally, the sequence of the features was sorted according to absolute MIVs, and identified potential lead features.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Statistical analysis\u003c/h2\u003e \u003cp\u003eStatistical analysis of the data was conducted using SPSS 25.0 software. Measurement data in mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{\\varvec{x}}\\pm\\:\\mathbf{s}\\)\u003c/span\u003e\u003c/span\u003e) represents. Non normal distribution data is represented as median and quartile, and evaluated using Wilcoxon test. \u003cem\u003et\u003c/em\u003e-test is used for comparison between two groups, and count data is compared using \u003cem\u003et\u003c/em\u003e-test χ2-test or Fisher's exact probability method, using multiple repeated ANOVA at multiple time points between groups. \u003cem\u003eP\u0026lt;0.05\u003c/em\u003e was considered to indicate statistical significance.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Clinical characteristics\u003c/h2\u003e \u003cp\u003eThis study included a total of 605 newly diagnosed patients with acute CO poisoning. After further screening, a total of 120 patients were included in the retrospective study, of which 75 were assigned to the non-DNS cohort and 45 were assigned to the DNS cohort after at least 60-day follow-up based on the definitions of DNS (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The statistical analysis of clinical predictive factors in the multivariate model showed that 8 out of 24 variables met the threshold of \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01, including age, carbon monoxide exposure time, initial GCS score, monocytes, LDH, arterial HCO3-, COHb, and lactate (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eClinical and demographic characteristics in our study.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNon-DNS group\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;75)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDNS group\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;45)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e42(56%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20(44.44%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33(64%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25(55.56%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47.75\u0026thinsp;\u0026plusmn;\u0026thinsp;20.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56.47\u0026thinsp;\u0026plusmn;\u0026thinsp;17.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.017\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCO exposure time, hours\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.98\u0026thinsp;\u0026plusmn;\u0026thinsp;1.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.65\u0026thinsp;\u0026plusmn;\u0026thinsp;3.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInitial GCS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14.08\u0026thinsp;\u0026plusmn;\u0026thinsp;1.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12.49\u0026thinsp;\u0026plusmn;\u0026thinsp;3.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiabetes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.925\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2(2.67%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1(2.22%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e73(97.33%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e44(97.78%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHypertension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.394\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12(16%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10(22.22%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e63(84%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35(77.78%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSystolic blood pressure, mmHg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e130.73\u0026thinsp;\u0026plusmn;\u0026thinsp;20.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e128.44\u0026thinsp;\u0026plusmn;\u0026thinsp;18.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.544\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiastolic blood pressure, mmHg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e89.55\u0026thinsp;\u0026plusmn;\u0026thinsp;95.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79.00\u0026thinsp;\u0026plusmn;\u0026thinsp;11.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.463\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeukocyte, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.31\u0026thinsp;\u0026plusmn;\u0026thinsp;3.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9.54\u0026thinsp;\u0026plusmn;\u0026thinsp;5.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.138\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeutrophile granulocyte, x10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.42\u0026thinsp;\u0026plusmn;\u0026thinsp;3.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.73\u0026thinsp;\u0026plusmn;\u0026thinsp;5.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.126\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeukomonocyte, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.38\u0026thinsp;\u0026plusmn;\u0026thinsp;0.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.23\u0026thinsp;\u0026plusmn;\u0026thinsp;0.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.211\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMonocyte, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.40\u0026thinsp;\u0026plusmn;\u0026thinsp;0.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.49\u0026thinsp;\u0026plusmn;\u0026thinsp;0.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlood platelet, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u0026thinsp;\u0026plusmn;\u0026thinsp;79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e209\u0026thinsp;\u0026plusmn;\u0026thinsp;66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.794\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBUN, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.79\u0026thinsp;\u0026plusmn;\u0026thinsp;2.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.62\u0026thinsp;\u0026plusmn;\u0026thinsp;2.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.725\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCr, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e65.9\u0026thinsp;\u0026plusmn;\u0026thinsp;31.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75.6\u0026thinsp;\u0026plusmn;\u0026thinsp;30.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCK, U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e742.5\u0026thinsp;\u0026plusmn;\u0026thinsp;2736.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2280.1\u0026thinsp;\u0026plusmn;\u0026thinsp;5895.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.055\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCK-MB, U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41.4\u0026thinsp;\u0026plusmn;\u0026thinsp;110.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e103.5\u0026thinsp;\u0026plusmn;\u0026thinsp;267.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.078\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLDH, U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e233.7\u0026thinsp;\u0026plusmn;\u0026thinsp;153.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e309.1\u0026thinsp;\u0026plusmn;\u0026thinsp;252.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.043\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHBD, U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e163.7\u0026thinsp;\u0026plusmn;\u0026thinsp;109.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e211.3\u0026thinsp;\u0026plusmn;\u0026thinsp;154.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.052\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCRP, mg/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14.78\u0026thinsp;\u0026plusmn;\u0026thinsp;39.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33.13\u0026thinsp;\u0026plusmn;\u0026thinsp;62.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.052\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArterial pH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.41\u0026thinsp;\u0026plusmn;\u0026thinsp;0.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.40\u0026thinsp;\u0026plusmn;\u0026thinsp;0.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArterial HCO3-, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.44\u0026thinsp;\u0026plusmn;\u0026thinsp;2.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20.47\u0026thinsp;\u0026plusmn;\u0026thinsp;3.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCOHb, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.41\u0026thinsp;\u0026plusmn;\u0026thinsp;7.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10.21\u0026thinsp;\u0026plusmn;\u0026thinsp;11.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLactic acid, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.18\u0026thinsp;\u0026plusmn;\u0026thinsp;1.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.62\u0026thinsp;\u0026plusmn;\u0026thinsp;2.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eCategorical variables are expressed as numbers (%) and continuous variables are expressed as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eAbbreviations: DNS: Delayed neurological sequelae; CO: Carbon monoxide; GCS score: Glasgow Coma Scale score; BUN: Blood urea nitrogen; Cr: Creatinine; CK: Creatine kinase; CK-MB: Creatine kinase-MB; LDH: Lactate dehydrogenase; HBD: Hydroxy butyrate dehydrogenase; CRP: C-reactive protein; COHb: Carboxyhemoglobin.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.2 DWI characteristics\u003c/h2\u003e \u003cp\u003eAfter analyzing DWI features, we found there were very high predictive value in 23 of 28 radiographic predictors, including the mean FA of left amygdala, right amygdala, left caudate, right caudate, left hippocampus, right hippocampus, left pallidus, right pallidus, left putamen, right putamen, left thalamus and right thalamus, and the mean MD of left amygdala, right amygdala, left caudate, left hippocampus, right hippocampus, left pallidus, right pallidus, left putamen, right putamen, left thalamus and right thalamus, which were all significantly correlated with the prediction efficiency (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Finally, we jointly established SVM model based on clinical characteristics and DWI characteristics as research objects. This model holds paramount importance in its capabilities for predicting the onset of DNS.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDWI characteristics in our study.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeatures\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNon-DNS group\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;75)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDNS group\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;45)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.43(0.45\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.78(0.75\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.58(0.55\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.66(0.69\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.45(0.47\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.67(0.70\u0026thinsp;\u0026plusmn;\u0026thinsp;0.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.51(0.53\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.75(0.71\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.43(0.49\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.88(0.85\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.53(0.53\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.70(0.65\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.48(0.50\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.81(0.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.65(0.64\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.66(0.63\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.226\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.52(0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.76(0.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.41(0.46\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.74(0.67\u0026thinsp;\u0026plusmn;\u0026thinsp;0.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.42(0.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.66(0.71\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.57(0.57\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.74(0.73\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.52(0.51\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.65(0.69\u0026thinsp;\u0026plusmn;\u0026thinsp;0.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.45(0.46\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.60(0.66\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.43(0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.82(0.79\u0026thinsp;\u0026plusmn;\u0026thinsp;0.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.55(0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.76(0.74\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.40(0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.58(0.68\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.47(0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.74(0.70\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.51(0.51\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.79(0.78\u0026thinsp;\u0026plusmn;\u0026thinsp;0.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.46(0.46\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.56(0.65\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.44(0.46\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.86(0.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.40(0.41\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.70(0.71\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.55(0.50\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.67(0.71\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.47(0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.65(0.66\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of left ventricle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.52(0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.61(0.56\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.076\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of left ventricle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.41(0.46\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.55(0.54\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.155\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean FA of right ventricle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.47(0.51\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.46(0.49\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.675\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean MD of right ventricle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.36(0.45\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.37(0.43\u0026thinsp;\u0026plusmn;\u0026thinsp;0.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.692\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eContinuous variables are expressed as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eAbbreviations: DNS: Delayed neurological sequelae. FA: Fractional Anisotropy; MD: Mean Diffusivity.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.3 SVM model\u003c/h2\u003e \u003cp\u003eWe established three SVM models that based on only clinical features (clinical model), only DWI features (DWI model), combined clinical and DWI features (clinical-DWI model), whose predictive accuracies of DNS were different. The prediction accuracy of the SVM model established solely based on clinical features is 0.76 [95% CI 0.67\u0026ndash;0.82]. The accuracy of the SVM model based solely on DWI features is 0.94 [95% CI 0.88\u0026ndash;0.98]. The accuracy of the SVM model based on combined clinical and DWI features is 0.97 [95% CI 0.94\u0026ndash;1.00]. In addition, the performance of the combined clinical and DWI features model is higher than that of clinical or DWI feature models, with a precision of 1.00 [95% CI 0.95\u0026ndash;1.00], sensitivity of 0.92 [95% CI 0.86\u0026ndash;0.98], F1 score of 0.96 [95% CI 0.92\u0026ndash;1.00], macroscopic mean of 0.98 [95% CI 0.0.83\u0026ndash;1.00], weighted mean of 0.97 [95% CI 0.0.93\u0026ndash;1.00], and AUC of 0.97 [95% CI 0.94\u0026ndash;1.00], and the result of 10-flod cross validation was 0.98 [95% CI 0.92\u0026ndash;1.00] (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The mean FA of left caudate had the highest impact on the SVM model based on MIV analysis (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eThe performance of three models for identifying delayed encephalopathy after acute carbon monoxide poisoning.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eF1-score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eMA\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eWA\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e10-flod cross\u0026nbsp;validation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.72\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDWI model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eClinical-DWI model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"10\"\u003eDWI: Diffusion-weighted imaging; AC: Accuracy; PR: Precision; SE: Sensitivity; MA: Macro-average; WA: Weighted-average; AUC: Area under the curve.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eThe feature importance sequence of MIV.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRanking\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFeatures\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMIV\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.208\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.192\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.192\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.175\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.163\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLactic acid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of right putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArterial HCO3-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.146\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.142\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.142\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.142\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCOHb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.128\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of left putamen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCO exposure time\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.115\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLDH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.113\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInitial GCS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of right amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of right hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of right pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of right thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean FA of right thalamus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.104\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left amygdala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.092\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left caudate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.092\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean MD of left pallidus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.092\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMonocyte\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.086\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.074\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003eMIV: Mean impact value; FA: Fractional Anisotropy; MD: Mean Diffusivity; COHb: carboxyhemoglobin; LDH: lactate dehydrogenase; GCS score: Glasgow Coma Scale score\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eNumerous predictors based on clinical data and statistical analyses have been identified in previous studies regarding CO poisoning. Lactic acid level was emerged as a significant predictor, as evidenced by a \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001 in our study, indicating its high predictive value. Elevated lactate levels, indicative of tissue hypoxia, are commonly utilized in intensive care units as a reliable prognostic factor for critically ill patients [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Studies have suggested that initial blood lactate levels may be associated with patient prognosis [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] and the severity of CO poisoning [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Zhang et al. has highlighted that a longer duration of CO exposure and lower GCS scores upon arrival are independent predictors of DNS following CO poisoning [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Pepe et al. also demonstrated that a CO exposure duration exceeding 6 hours may elevate the risk of DNS development [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. These studies collectively underscore the critical role of CO exposure duration and initial GCS score in the progression of delayed neurological complications. At the same time, some prospective studies have consistently identified a GCS score of less than 9 as a crucial predictor of DNS [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The GCS score serves as an objective reflection of the patient's level of consciousness, with lower scores indicating a poorer prognosis. In our study, we get the same result, as evidenced by the significant correlation (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) between the duration of CO exposure, initial GCS score, and prognosis in CO poisoning cases.\u003c/p\u003e \u003cp\u003eDWI conducted during the acute phase of CO poisoning can offer valuable insights into predicting long-term neurological outcomes post-discharge [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Previous studies have demonstrated that MRI-DWI scans revealing acute ischemic brain lesions are closely linked to the development of DNS. In particular, the pallidum emerges as the most commonly affected region in DWI scans, with other frequently impacted areas including the cerebellum, hippocampus, putamen, amygdala and corpus callosum\u0026mdash;all of which are associated with the occurrence of DNS and often involved in both sides simultaneously [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Our study delved into the extraction of FA and MD, two commonly used research parameters, revealing significant associations with the occurrence of DNS in the later stages. These findings underscore the potential of DWI in identifying individuals at risk of DNS and enabling targeted interventions to mitigate adverse neurological sequelae.\u003c/p\u003e \u003cp\u003eAlthough many predecessors have explored clinical features in the context of acute carbon monoxide poisoning, however, there remains a paucity of research focusing on the concurrent extraction of DWI features for the development of predictive models. Existing investigations that integrate clinical and imaging data have primarily extracted features such as MRI, and diffusion tensor imaging, while many analyses of DWI data have been limited to generic descriptions of high signal lesions or the presence of abnormalities in DWI-MRI scans, with minimal emphasis on delineating specific brain regions and parameter alterations. The lack of comprehensive feature extraction poses a challenge in achieving a nuanced understanding of the disease pathophysiology [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In contrast, our study presents a novel approach by integrating clinical and DWI features to formulate a SVM model for prognostic evaluation in acute CO poisoning.\u003c/p\u003e \u003cp\u003eThere are some advantages of our method. First, we combined the clinical features and DWI features to construct a SVM model with good performance for the prognosis assessment of acute CO poisoning and guide the subsequent prevention and treatment, which can be easily obtained from clinical records and admission examinations. Second, our study counted 605 cases of acute CO poisoning from 2018 to 2023, and the remaining 120 cases met the inclusion criteria after layers of screening, which is a unique study with a large sample size for this regional disease. Third, the MIV of clinically significant variables can reflect the correlation between clinical and imaging features and prognosis of our study clearly and intuitively, enhancing the interpretability of our findings.\u003c/p\u003e \u003cp\u003eHowever, there are several limitations in our study. First, this study is a single-center design based on retrospective data, so the results may not be generalized to other centers, and further multi-center studies are needed to verify and expand the practical application value for DNS prediction. Second, the variability in CO exposure concentrations, variations in pre-hospital interventions, and discrepancies in the duration of oxygen therapy among individual patients poses potential confounders that could impact prognostic outcomes. Finally, when we took DWI features, we tried to extract and analyze the average diffusion coefficient (ADC), signal intensity ratio (SIR) and other factors commonly extracted and analyzed in the central system. However, we are only satisfied with the parameters of FA and MD in certain ROIs indicating a limited scope in parameter selection within the DWI domain. This restricted parameter set highlights the potential for further scalability in DWI analysis to encompass a broader array of relevant factors for a more comprehensive evaluation of DNS prognosis.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eWe demonstrated the changes between DNS group and non-DNS group in FA and MD in DWI during the acute phase of CO poisoning, and the utilization of DWI can be instrumental in identifying individuals at risk of developing DNS. A SVM model that combines multi-modal features including DWI features and clinical parameters, has been proven to effectively predict DNS episodes in patients with CO poisoning. Furthermore, the model's practicality in a clinical setting enhances its value as a tool for early intervention and tailored management strategies for patients at risk of DNS following CO poisoning.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCO Carbon monoxide\u003c/p\u003e\n\u003cp\u003eDNS Delayed neurological sequelae\u003c/p\u003e\n\u003cp\u003eMRI Magnetic Resonance Imaging\u003c/p\u003e\n\u003cp\u003eDWI Diffusion-weighted imaging\u003c/p\u003e\n\u003cp\u003eADC Apparent diffusion coefficient\u003c/p\u003e\n\u003cp\u003eABLDs Brain Lesions on DWI\u003c/p\u003e\n\u003cp\u003eRFC Random Forest Classifier\u003c/p\u003e\n\u003cp\u003eCOHb Carboxyhemoglobin \u003c/p\u003e\n\u003cp\u003eGCS Glasgow Coma Scale\u003c/p\u003e\n\u003cp\u003eMMSE Mini-Mental State Examination\u003c/p\u003e\n\u003cp\u003eMoCA Montreal Cognitive Assessment\u003c/p\u003e\n\u003cp\u003eFA Fractional anisotropy\u003c/p\u003e\n\u003cp\u003eMD Mean diffusivity\u003c/p\u003e\n\u003cp\u003eSVM Support vector machine\u003c/p\u003e\n\u003cp\u003eAUC Area under the receiver operating characteristic curve \u003c/p\u003e\n\u003cp\u003eMIV Mean impact value \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would also like to thank our staff, who assisted in the data collection and analysis.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCSY, YSJ, THM and LHY were joint first authors; LQH was correspondence author; CSY, YSJ and LQH designed the study; TMH and LHY reviewed the literature; LJL and LH collected the data; TMH and LHY performed the follow-up activity; YSJ performed the statistical analysis; CSY and YSJ wrote the manuscript; LJL acquired fund; YXQ, HYZ, TSB and LQH revised the manuscript. All authors have read and approved the final manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was sponsored by the Enshi Tujia and Miao Autonomous Prefecture Science and Technology Bureau Project in 2023 (Authorization number: D20230077).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was reviewed and approved by the Ethics Committee of the Central Hospital of the Enshi Prefecture, with ethics approval reference (2023-011-02). All patients and their families have signed informed consent forms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003c/strong\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no conflict of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eInagaki T, Ishino H, Seno H, et al. A long-term follow-up study of serial magnetic resonance images in patients with delayed encephalopathy after acute carbon monoxide poisoning. J Neuropsychiatry Clin Neurosci. 1997;51(6):421\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChoi IS. Carbon monoxide poisoning: systemic manifestations and complications. J Korean Med Sci. 2001;16:253\u0026ndash;61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeaver LK. Clinical practice. Carbon monoxide poisoning. N Engl J Med. 2009;360(12):1217\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eErnst A, Zibrak JD. Carbon monoxide poisoning. N Engl J Med. 1998;339(22):1603\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeppu T. The role of MR imaging in assessment of brain damage from carbon monoxide poisoning: a review of the literature. AJNR Am J Neuroradiol. 2014;35(4):625\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoon JM, Chun BJ, Cho YS. The predictive value of scores based on peripheral complete blood cell count for long-term neurological outcome in acute carbon monoxide intoxication. Volume 124. Basic \u0026amp; clinical pharmacology \u0026amp; toxicology; 2019. pp. 500\u0026ndash;10. 4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eS\u0026ouml;nksen SE, K\u0026uuml;hn S, Basner M, et al. Brain structure and neurocognitive function in two professional mountaineers during 35 days of severe normobaric hypoxia. Eur J Neurol. 2022;29(10):3112\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeon SB, Sohn CH, Seo DW, et al. Acute Brain Lesions on Magnetic Resonance Imaging and Delayed Neurological Sequelae in Carbon Monoxide Poisoning. JAMA Neurol. 2018;75(4):436\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSert ET, Kokulu K, Mutlu H. Clinical predictors of delayed neurological sequelae in charcoal-burning carbon monoxide poisoning. Am J Emerg Med. 2021;48:12\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoon JM, Chun BJ, Baek BH, et al. Initial diffusion-weighted MRI and long-term neurologic outcomes in charcoal-burning carbon monoxide poisoning. Clin Toxicol (Phila). 2018;56(3):161\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim YS, Cha YS, Kim MS, et al. The usefulness of diffusion-weighted magnetic resonance imaging performed in the acute phase as an early predictor of delayed neuropsychiatric sequelae in acute carbon monoxide poisoning. Hum Exp Toxicol. 2018;37(6):587\u0026ndash;95.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhorbian M, Ghorbian S. Usefulness of machine learning and deep learning approaches in screening and early detection of breast cancer. Heliyon. 2023;9(12):e22427.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu G, Wang S, Liu J, et al. Using machine learning methods to study the tumour microenvironment and its biomarkers in osteosarcoma metastasis. Heliyon. 2024;10(7):e29322.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiyamoto N, Ueno Y, Yamashiro K, et al. Stroke classification and treatment support system artificial intelligence for usefulness of stroke diagnosis. Front Neurol. 2023;14:1295642.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMethods In Medicine CAM. Retracted: Artificial Intelligence Algorithm-Based MRI in Evaluating the Treatment Effect of Acute Cerebral Infarction. Computational and mathematical methods in medicine. 2023, 9794357.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang L, Li Y, Bian L, et al. Cognitive Impairment of Patient With Neurological Cerebrovascular Disease Using the Artificial Intelligence Technology Guided by MRI. Front public health. 2021;9:813641.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan MJ, Hu CC, Huang WH, et al. An artificial intelligence algorithm for analyzing globus pallidus necrosis after carbon monoxide intoxication. Hum Exp Toxicol. 2023;42:9603271231190906.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRose JJ, Zhang MS, Pan J, et al. Heart-Brain 346-7 Score: the development and validation of a simple mortality prediction score for carbon monoxide poisoning utilizing deep learning. Clin Toxicol (Phila). 2023;61(7):492\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKokulu K, Mutlu H, Sert ET. Serum netrin-1 levels at presentation and delayed neurological sequelae in unintentional carbon monoxide poisoning. Clin Toxicol (Phila). 2020;58(12):1313\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim YJ, Sohn CH, Seo DW, et al. Clinical Predictors of Acute Brain Injury in Carbon Monoxide Poisoning Patients With Altered Mental Status at Admission to Emergency Department. Acad Emerg medicine: official J Soc Acad Emerg Med. 2019;26(1):60\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHachinski V, Iadecola C, Petersen RC, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke. 2006;37(9):2220\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang JL, Su X, Zhang H, et al. A novel approach to active compounds identification based on support vector regression model and mean impact value. Volume 81. Chemical biology \u0026amp; drug design; 2013. pp. 650\u0026ndash;7. 5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBernardin G, Pradier C, Tiger F, et al. Blood pressure and arterial lactate level are early indicators of short-term survival in human septic shock. Intensive Care Med. 1996;22(1):17\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbramson D, Scalea TM, Hitchcock R et al. Lactate clearance and survival following injury. The Journal of trauma, 1993, 35(4): 584-8; discussion 8\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInoue S, Saito T, Tsuji T, et al. Lactate as a prognostic factor in carbon monoxide poisoning: a case report. Am J Emerg Med. 2008;26(8):e9661\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenaissa ML, M\u0026eacute;garbane B, Borron SW, et al. Is elevated plasma lactate a useful marker in the evaluation of pure carbon monoxide poisoning? Intensive Care Med. 2003;29(8):1372\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSokal JA, Kralkowska E. The relationship between exposure duration, carboxyhemoglobin, blood glucose, pyruvate and lactate and the severity of intoxication in 39 cases of acute carbon monoxide poisoning in man. Arch Toxicol. 1985;57(3):196\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y, Lu Q, Jia J, et al. Multicenter retrospective analysis of the risk factors for delayed neurological sequelae after acute carbon monoxide poisoning. Am J Emerg Med. 2021;46:165\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePepe G, Castelli M, Nazerian P, et al. Delayed neuropsychological sequelae after carbon monoxide poisoning: predictive risk factors in the Emergency Department. A retrospective study. Scand J Trauma Resusc Emerg Med. 2011;19:16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan S, Choi S, Nah S, et al. Cox regression model of prognostic factors for delayed neuropsychiatric sequelae in patients with acute carbon monoxide poisoning: A prospective observational study. Neurotoxicology. 2021;82:63\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim DM, Lee IH, Park JY, et al. Acute carbon monoxide poisoning: MR imaging findings with clinical correlation. Diagn Interv Imaging. 2017;98(4):299\u0026ndash;306.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-neurology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"nurl","sideBox":"Learn more about [BMC Neurology](http://bmcneurol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/nurl","title":"BMC Neurology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"CO poisoning, Delayed encephalopathy sequelae, Diffusion-weighted imaging, Fractional anisotropy, Mean diffusivity, Machine learning","lastPublishedDoi":"10.21203/rs.3.rs-5443111/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5443111/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eDelayed neurological sequelae (DNS) represents a critical and potentially fatal complication. Therefore, the timely recognition of individuals at risk of developing DNS in early phase holds significant clinical value. This study aims to identify diffusion-weighted imaging (DWI) characteristics related to developing DNS, and construct a predictive model encompassing DWI characteristics and clinical variables to early prediction of DNS during the acute phase of carbon monoxide (CO) poisoning.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe retrospectively include 120 poisoned patients with newly diagnosed with CO poisoning. The subjects were divided into non-DNS group (n\u0026thinsp;=\u0026thinsp;75) and DNS group (n\u0026thinsp;=\u0026thinsp;45) after at least 60-day follow-up. The fractional anisotropy value and mean diffusivity value were measured in the regions of interest placed on the amygdala, caudate, hippocampus, pallidus, putamen, thalamus and ventricle. A support vector machine (SVM) model integrated both DWI and clinical features was developed and evaluated. And mean impact value was used to rank the features that had impacts on classification.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eA total of 24 clinical features and 28 DWI features were included. 8 clinical features and 23 DWIs were included in the SVM model. Three SVM models were established based solely on clinical features or DWI features, and combined clinical and DWI features, with prediction accuracy of 0.76, 0.94, and 0.97, respectively. The precision, sensitivity, F1 score, macroscopic mean, weighted mean, and AUC of the combined model is 1.00, 0.92, 0.96, 0.98, 0.97, 0.97, respectively, and the result of 10-flod cross validation was 0.98. The mean fractional anisotropy of left caudate had the highest impact on the SVM model.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThe fractional anisotropy and mean diffusivity of DWI may be a potential biomarker in identifying patients at risk of developing DNS. Our comprehensive SVM model with multimodal features had excellent accuracy and clinical practicability in identifying DNS.\u003c/p\u003e","manuscriptTitle":"Identifying delayed neurological sequelae during the acute phase of carbon monoxide poisoning based on diffusion-weighted imaging and clinical features","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-03 14:54:00","doi":"10.21203/rs.3.rs-5443111/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-11-18T05:33:56+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-15T14:40:55+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-15T14:40:29+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Neurology","date":"2024-11-13T02:39:30+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-neurology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"nurl","sideBox":"Learn more about [BMC Neurology](http://bmcneurol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/nurl","title":"BMC Neurology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"43fdd4d0-de38-428c-acd0-542de78297e4","owner":[],"postedDate":"December 3rd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-23T11:53:29+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-03 14:54:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5443111","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5443111","identity":"rs-5443111","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}