Accurately predicting the risk of unfavorable outcomes after endovascular coil therapy in patients with aneurysmal subarachnoid hemorrhage: an interpretable machine learning model.

BACKGROUND: Despite endovascular coiling as a valid modality in treatment of aneurysmal subarachnoid hemorrhage (aSAH), there is a risk of poor prognosis. However, the clinical utility of previously proposed early prediction tools remains limited. We aimed to develop a clinically generalizable machine learning (ML) models for accurately predicting unfavorable outcomes in aSAH patients after endovascular coiling. METHODS: Functional outcomes at 6 months after endovascular coiling were assessed via the modified Rankin Scale (mRS) and unfavorable outcomes were defined as mRS 3-6. Five ML algorithms (logistic regression, random forest, support vector machine, deep neural network, and extreme gradient boosting) were used for model development. The area under precision-recall curve (AUPRC) and receiver operating characteristic curve (AUROC) was used as main indices of model evaluation. SHapley Additive exPlanations (SHAP) method was applied to interpret the best-performing ML model. RESULTS: A total of 371 patients were eventually included into this study, and 85.4% of them had favorable outcomes. Among the five models, the DNN model had a better performance with AUPRC of 0.645 (AUROC of 0.905). Postoperative GCS score, size of aneurysm, and age were the top three powerful predictors. The further analysis of five random cases presented the good interpretability of the DNN model. CONCLUSION: Interpretable clinical prediction models based on different ML algorithms have been successfully constructed and validated, which would serve as reliable tools in optimizing the treatment decision-making of aSAH. Our DNN model had better performance to predict the unfavorable outcomes at 6 months in aSAH patients compared with Yan's nomogram model.

Pre- and Post-Operative Online Prediction of Outcome in Patients Undergoing Endovascular Coiling after Aneurysmal Subarachnoid Hemorrhage: Visual and Dynamic Nomograms.

BACKGROUND: Aneurysmal subarachnoid hemorrhage (aSAH) causes long-term functional dependence and death. Early prediction of functional outcomes in aSAH patients with appropriate intervention strategies could lower the risk of poor prognosis. Therefore, we aimed to develop pre- and post-operative dynamic visualization nomograms to predict the 1-year functional outcomes of aSAH patients undergoing coil embolization. METHODS: Data were obtained from 400 aSAH patients undergoing endovascular coiling admitted to the People's Hospital of Hunan Province in China (2015-2019). The key indicator was the modified Rankin Score (mRS), with 3-6 representing poor functional outcomes. Multivariate logistic regression (MLR)-based visual nomograms were developed to analyze baseline characteristics and post-operative complications. The evaluation of nomogram performance included discrimination (measured by C statistic), calibration (measured by the Hosmer-Lemeshow test and calibration curves), and clinical usefulness (measured by decision curve analysis). RESULTS: Fifty-nine aSAH patients (14.8%) had poor outcomes. Both nomograms showed good discrimination, and the post-operative nomogram demonstrated superior discrimination to the pre-operative nomogram with a C statistic of 0.895 (95% CI: 0.844-0.945) vs. 0.801 (95% CI: 0.733-0.870). Each was well calibrated with a Hosmer-Lemeshow p-value of 0.498 vs. 0.276. Moreover, decision curve analysis showed that both nomograms were clinically useful, and the post-operative nomogram generated more net benefit than the pre-operative nomogram. Web-based online calculators have been developed to greatly improve the efficiency of clinical applications. CONCLUSIONS: Pre- and post-operative dynamic nomograms could support pre-operative treatment decisions and post-operative management in aSAH patients, respectively. Moreover, this study indicates that integrating post-operative variables into the nomogram enhanced prediction accuracy for the poor outcome of aSAH patients.

Associations of Lipoprotein(a) Level with Cerebral Small Vessel Disease in Patients with Alzheimer's Disease.

BACKGROUND: This study aimed to examine the association of lipoprotein(a) [Lp(a)] level with the burden of cerebral small vessel disease (CSVD) in patients with Alzheimer's disease (AD). METHODS: Data from 111 consecutive patients with AD admitted to Nanjing First Hospital from 2015 to 2022 were retrospectively analyzed in this study. Serum Lp(a) concentrations were grouped into tertiles (T1-T3). Brain magnetic resonance imaging (MRI) was rated for the presence of CSVD, including enlarged perivascular spaces (EPVS), lacunes, white-matter lesions, and cerebral microbleeds (CMBs). The CSVD burden was calculated by summing the scores of each MRI marker at baseline. A binary or ordinal logistic regression model was used to estimate the relationship of serum Lp(a) levels with CSVD burden and each MRI marker. RESULTS: Patients with higher tertiles of Lp(a) levels were less likely to have any CSVD (T1, 94.6%; T2, 78.4%; T3, 66.2%; p = 0.013). Multivariable analysis found that Lp(a) levels were inversely associated with the presence of CSVD (T2 vs. T1: adjusted odds ratio [aOR] 0.132, 95% confidence interval [CI] 0.018-0.946, p = 0.044; T3 vs. T1: aOR 0.109, 95% CI 0.016-0.737, p = 0.023) and CSVD burden (T3 vs. T1: aOR 0.576, 95% CI 0.362-0.915, p = 0.019). The independent relationship between Lp(a) levels and individual CSVD features was significant for moderate-to-severe EPVS in the centrum semiovale (T2 vs. T1: aOR 0.059, 95% CI 0.006-0.542, p = 0.012; T3 vs. T1: aOR 0.029, 95% CI 0.003-0.273, p = 0.002) and CMBs (T3 vs. T1: aOR 0.144, 95% CI 0.029-0.716, p = 0.018). CONCLUSIONS: In this study, serum Lp(a) level was inversely associated with CSVD in AD patients.

Dynamic Prediction of Mechanical Thrombectomy Outcome for Acute Ischemic Stroke Patients Using Machine Learning.

The unfavorable outcome of acute ischemic stroke (AIS) with large vessel occlusion (LVO) is related to clinical factors at multiple time points. However, predictive models used for dynamically predicting unfavorable outcomes using clinically relevant preoperative and postoperative time point variables have not been developed. Our goal was to develop a machine learning (ML) model for the dynamic prediction of unfavorable outcomes. We retrospectively reviewed patients with AIS who underwent a consecutive mechanical thrombectomy (MT) from three centers in China between January 2014 and December 2018. Based on the eXtreme gradient boosting (XGBoost) algorithm, we used clinical characteristics on admission ("Admission" Model) and additional variables regarding intraoperative management and the postoperative National Institute of Health stroke scale (NIHSS) score ("24-Hour" Model, "3-Day" Model and "Discharge" Model). The outcome was an unfavorable outcome at the three-month mark (modified Rankin scale, mRS 3-6: unfavorable). The area under the receiver operating characteristic curve and Brier scores were the main evaluating indexes. The unfavorable outcome at the three-month mark was observed in 156 (62.0%) of 238 patients. These four models had a high accuracy in the range of 75.0% to 87.5% and had a good discrimination with AUC in the range of 0.824 to 0.945 on the testing set. The Brier scores of the four models ranged from 0.122 to 0.083 and showed a good predictive ability on the testing set. This is the first dynamic, preoperative and postoperative predictive model constructed for AIS patients who underwent MT, which is more accurate than the previous prediction model. The preoperative model could be used to predict the clinical outcome before MT and support the decision to perform MT, and the postoperative models would further improve the predictive accuracy of the clinical outcome after MT and timely adjust therapeutic strategies.

A Dynamic Nomogram to Identify Patients at High Risk of Poor Outcome in Stroke Patients with Chronic Kidney Disease.

Background and Purpose: Predicting poor outcome for stroke patients with chronic kidney disease (CKD) in clinical practice is difficult. There are no tools available to use for predicting poor outcome in these patients. We aimed to construct and validate a dynamic nomogram to identify CKD-stroke patients at high risk of a 3-month poor outcome. Patients and Methods: We used data for 502 CKD patients who had an acute ischemic stroke, from Nanjing First Hospital, between September 2014 and September 2020, to train the nomogram. An additional 108 patients enrolled from October 2020 to May 2021 were used for temporal external validation. The performance of the nomogram was evaluated by the area under the receiver operating characteristics curve (AUC) and a calibration plot. The clinical utility of the nomogram was measured by decision curve analysis (DCA) and the clinical impact curve (CIC). Results: The median age of the cohort was 79 (70-84) years. Age, urea, premorbid modified Rankin Scale (mRS), National Institutes of Health Stroke Scale (NIHSS) on admission, hemiplegia, mechanical thrombectomy, early neurological deterioration, and respiratory infection were used as predictors of 3-month poor outcome to develop the nomogram. In the training set, the AUC of the dynamic nomogram was 0.873 and the calibration plot showed good predictive ability, and both DCA and CIC indicated the excellent clinical usefulness and applicability of the nomogram. In the external validation set, the AUC was 0.875 and the calibration plot also showed good agreement. Conclusion: This is the first dynamic nomogram constructed for CKD-stroke patients to precisely and expediently identify patients with a high risk of 3-month poor outcome. The outstanding performance and great clinical predictive utility demonstrated the ability of the dynamic nomogram to help clinicians to deploy preventive interventions.

FLAIR vascular hyperintensity predicts early neurological deterioration in patients with acute ischemic stroke receiving endovascular thrombectomy.

Fluid-attenuated inversion recovery vascular hyperintensity (FVH) is frequently observed in patients with acute ischemic stroke (AIS). FVH is associated with functional outcome at 3 months in AIS patients receiving endovascular thrombectomy. In the present study, we assessed whether FVH predicted early neurological deterioration (END) and hemorrhagic transformation (HT) within 72 h in AIS patients receiving endovascular thrombectomy. We retrospectively analyzed 104 patients with acute internal-carotid-artery or proximal middle-cerebral-artery occlusion within 16 h after symptom onset. Before thrombectomy, all patients underwent brain magnetic resonance imaging. END was defined as an increase of 4 points or more from baseline National Institutes of Health Stroke Scale (NIHSS) during 72 h following onset. HT was assessed by brain computed tomography. Statistical analyses were performed to predict END and HT. The proportion of high FVH score, high American Society of Intervention and Therapeutic Neuroradiology/Society of Interventional Radiology (ASITN/SIR) grade in non-END group was higher than that in END group (p < 0.001, p < 0.001, respectively). FVH score was positively correlated with ASITN/SIR grade (r = 0.461, p < 0.001). FVH score was a predictor factor for END (adjusted OR, 13.552; 95% CI, 2.408-76.260; p = 0.003), while FVH score was not a predictor factor for HT. Furthermore, NIHSS at admission (adjusted OR, 1.112; 95% CI, 1.006-1.228; p = 0.038) and high-density lipoprotein cholesterol (adjusted OR, 18.865; 95% CI, 2.998-118.683; p = 0.002) were predictor factors for HT. To assess FVH score before thrombectomy might be useful for predicting END in AIS patients receiving endovascular thrombectomy.

Using machine learning to predict atrial fibrillation diagnosed after ischemic stroke.

BACKGROUND: Selecting best candidates for prolonged poststroke cardiac monitoring in acute ischemic stroke (AIS) patients is still challenging. We aimed to develop a machine learning (ML) model to select AIS patients at high risk of poststroke atrial fibrillation (AF) for prolonged cardiac monitoring and then to compare ML model with traditional risk scores and classic statistical logistic regression (classic-LR) model. METHODS: AIS patients from July 2012 to September 2020 across Nanjing First Hospital were collected. We performed the LASSO regression for selecting the critical features and built five ML models to assess the risk of poststroke AF. The SHAP and partial dependence plot (PDP) method were introduced to interpret the optimal model. We also compared ML model with CHADS2 score, CHA2DS2-VASc score, AS5F score, HAVOC score, and classic-LR model. RESULTS: A total of 3929 AIS patients were included. Among the five ML models, deep neural network (DNN) was the model with best performance. It also exhibited superior performance compared with CHADS2 score, CHA2DS2-VASc score, AS5F score, HAVOC score and classic-LR model. The results of SHAP and PDP method revealed age, cardioembolic stroke, large-artery atherosclerosis stroke, and NIHSS score at admission were the top four important features and revealed the DNN model had good interpretability and reliability. CONCLUSION: The DNN model achieved best performance and improved prediction performance compared with traditional risk scores and classic-LR model. The DNN model can be applied to identify AIS patients at high risk of poststroke AF as best candidates for prolonged poststroke cardiac monitoring.

Prediction-Driven Decision Support for Patients With Mild Stroke: A Model Based on Machine Learning Algorithms.

Background and Purpose: Treatment for mild stroke remains an open question. We aim to develop a decision support tool based on machine learning (ML) algorithms, called DAMS (Disability After Mild Stroke), to identify mild stroke patients who would be at high risk of post-stroke disability (PSD) if they only received medical therapy and, more importantly, to aid neurologists in making individual clinical decisions in emergency contexts. Methods: Ischemic stroke patients were prospectively recorded in the National Advanced Stroke Center of Nanjing First Hospital (China) between July 2016 and September 2020. The exclusion criteria were patients who received thrombolytic therapy, age <18 years, lack of 3-month modified Rankin Scale (mRS), disabled before the index stroke, with an admission National Institute of Health stroke scale (NIHSS) > 5. The primary outcome was PSD, corresponding to 3-month mRS ≥ 2. We developed five ML models and assessed the area under curve (AUC) of receiver operating characteristic, calibration curve, and decision curve analysis. The optimal ML model was selected to be DAMS. In addition, SHapley Additive exPlanations (SHAP) approach was introduced to rank the feature importance. Finally, rapid-DAMS (R-DAMS) was constructed for a more urgent situation based on DAMS. Results: A total of 1,905 mild stroke patients were enrolled in this study, and patients with PSD accounted for 23.4% (447). There was no difference in AUCs between the five models (ranged from 0.691 to 0.823). Although there was similar discriminative performance between ML models, the support vector machine model exhibited higher net benefit and better calibration (Brier score, 0.159, calibration slope, 0.935, calibration intercept, 0.035). Therefore, this model was selected for DAMS. In addition, SHAP approach showed that the most crucial feature was NIHSS on admission. Finally, R-DAMS was constructed and there was similar discriminative performance between R-DAMS and DAMS, but the former performed worse on calibration. Conclusions: DAMS and R-DAMS, as prediction-driven decision support tools, were designed to aid clinical decision-making for mild stroke patients in emergency contexts. In addition, even within a narrow range of baseline scores, NIHSS on admission is the strongest feature that contributed to the prediction.

Fluid-Attenuated Inversion Recovery Vascular Hyperintensity as a Potential Predictor for the Prognosis of Acute Stroke Patients After Intravenous Thrombolysis.

BACKGROUND: Fluid-attenuated inversion recovery vascular hyperintensity (FVH) can reflect the collateral status, which may be a valuable indicator to predict the functional outcome of acute stroke (AS) patients. METHODS: A total of 190 AS patients with large vessel occlusion (LVO) were retrospectively investigated. All patients completed a 6-month follow-up and their modified Rankin Scale (mRS) scores were recorded at 1, 3, and 6 months after intravenous thrombolysis (IVT). Based on their mRS at 3 months, patients were divided into two groups: poor prognosis (131 patients; 68.9% of all subjects) and favorable prognosis (59 patients; 31.1% of all subjects). The death records of 28 patients were also analyzed in the poor prognosis group. RESULTS: (1) Univariate and multivariate analyses showed that the higher National Institutes of Health Stroke Scale (NIHSS) score at admission, higher fasting blood glucose, and lower FVH score were independent risk factors to predict the poor prognosis of IVT. (2) Survival analysis indicated that FVH score was the only baseline factor to predict the 6-month survival after IVT. (3) Baseline FVH score had great prediction performance for the prognosis of IVT (area under the curve = 0.853). (4) Baseline FVH score were negatively correlated with the NIHSS score at discharge and mRS score at 1, 3, and 6 months. CONCLUSION: Among various baseline clinical factors, only the FVH score might have implications for 3-month outcome and 6-month survival of AS patients after IVT. Baseline FVH score showed great potential to predict the prognosis of the AS patients.

Effect of metal oxide nanoparticles on amino acids in wheat grains (Triticum aestivum) in a life cycle study.

Engineered nanoparticles (NPs) are now used as additives in pesticides and fungicides and as novel fertilizers in agriculture so there is an urgent need to explore their effects on crop yield and quality in a full life cycle study. In the present study, three widely used NPs (TiO2, Fe2O3 and CuO NPs applied at doses of 50 and 500 mg/kg) were selected to investigate their long-term impact on wheat growth. TiO2 NPs did not affect the growth and development of wheat, but Fe2O3 NPs promoted wheat precocity and CuO NPs inhibited the growth and development of the wheat grains. The Cu content in grains treated with CuO NP increased by 18.84%-30.45% compared with the control. However, the contents of Fe and Zn were both significantly lower in the CuO NP treatments. Univariate and multivariate analyses were used to analyze the effect of different NPs on the composition of amino acids in wheat grains. Exposure to TiO2 NPs at dose of 500 mg/kg increased the overall amino acid nutrition in the edible portion of wheat. Fe2O3 NPs at both doses increased the contents of cysteine (Cys) and tyrosine (Tyr). The addition of CuO NPs reduced the level of some essential amino acids in wheat grains, isoleucine (Ile), leucine (Leu), threonine (Thr) and histidine (His). Overall, evaluation of the potential impacts of metal-based NPs on the nutritional quality of wheat grains could provide important information for their safe use when incorporated into agrichemicals in sustainable agriculture.

Alteration of Crop Yield and Quality of Wheat upon Exposure to Silver Nanoparticles in a Life Cycle Study.

As a result of the rapid development of nanotechnology, metal-based nanoparticles (NPs) are inadvertently released into the environment and may pose a potential threat to the ecosystem. However, information for food quality and safety in NP-treated crops is limited. In the present study, wheat ( Triticum aestivum L.) was grown in different concentrations of Ag-NP-amended soil (20, 200, and 2000 mg kg-1) for 4 months. At harvest, physiological parameters, Ag and micronutrient (Fe, Cu, and Zn) contents, and amino acid and total protein contents were measured. Results showed that, with increasing the exposure doses, Ag NPs exhibited severe phytotoxicity, including lower biomass, shorter plant height, and lower grain weight. Ag accumulation in roots was significantly higher than that in shoots and grains. Decreases in the content of micronutrients (Fe, Cu, and Zn) in Ag-NP-treated grains suggested low crop quality. The results of amino acid and protein contents in Ag-NP-treated wheat grains indicated that Ag NPs indeed altered the nutrient contents in the edible portion. In the amino acid profile, the presence of Ag NPs significantly decreased the contents of arginine and histidine by 13.0 and 11.8%, respectively. In summary, the effects of metal-based NPs on the edible portion of crops should be taken into account in the evaluation of nanotoxicity to terrestrial plants. Moreover, investigation of the potential impacts of NP-caused nutrient alterations on human health could further our understandings on NP-induced phytotoxicity.

Chaperonin-Nanocaged Hemin as an Artificial Metalloenzyme for Oxidation Catalysis.

Taking inspiration from biology's effectiveness in functionalizing protein-based nanocages for chemical processes, we describe here a rational design of an artificial metalloenzyme for oxidations with the bacterial chaperonin GroEL, a nanocage for protein folding in nature, by supramolecular anchoring of catalytically active hemin in its hydrophobic central cavity. The promiscuity of the chaperonin cavity is an essential element of this design, which can mimic the hydrophobic binding pocket in natural metalloenzymes to accept cofactor and substrate without requiring specific ligand-protein interactions. The success of this approach is manifested in the efficient loading of multiple monomeric hemin cofactors to the GroEL cavity by detergent dialysis and good catalytic oxidation properties of the resulting biohybrid in tandem with those of the clean oxidant of H2O2. Investigation of the mechanism of hemin-GroEL-catalyzed oxidation of two-model substrates reveals that the kinetic behavior of the complex follows a ping-pong mechanism in both cases. Through comparison with horseradish peroxidase, the oxidative activity and stability of hemin-GroEL were observed to be similar to those found in natural peroxidases. Adenosine 5'-triphosphate (ATP)-regulated partial dissociation of the biohybrid, as assessed by the reduction of its catalytic activity with the addition of the nucleotide, raises the prospect that ATP may be used to recycle the chaperonin scaffold. Moreover, hemin-GroEL can be applied to the chromogenic detection of H2O2, which (or peroxide in general) is commonly contained in industrial wastes. Considering the rich chemistry of free metalloporphyrins and the ease of production of GroEL and its supramolecular complex with hemin, this work should seed the creation of many new artificial metalloenzymes with diverse reactivities.

Effects of TiO2 nanoparticles on wheat (Triticum aestivum L.) seedlings cultivated under super-elevated and normal CO2 conditions.

Concerns over the potential risks of nanomaterials to ecosystem have been raised, as it is highly possible that nanomaterials could be released to the environment and result in adverse effects on living organisms. Carbon dioxide (CO2) is one of the main greenhouse gases. The level of CO2 keeps increasing and subsequently causes a series of environmental problems, especially for agricultural crops. In the present study, we investigated the effects of TiO2 NPs on wheat seedlings cultivated under super-elevated CO2 conditions (5000 mg/L CO2) and under normal CO2 conditions (400 mg/L CO2). Compared to the normal CO2 condition, wheat grown under the elevated CO2 condition showed increases of root biomass and large numbers of lateral roots. Under both CO2 cultivation conditions, the abscisic acid (ABA) content in wheat seedlings increased with increasing concentrations of TiO2 NPs. The indolepropioponic acid (IPA) and jasmonic acid (JA) content notably decreased in plants grown under super-elevated CO2 conditions, while the JA content increased with increasing concentrations of TiO2 NPs. Ti accumulation showed a dose-response manner in both wheat shoots and roots as TiO2 NPs concentrations increased. Additionally, the presence of elevated CO2 significantly promoted Ti accumulation and translocation in wheat treated with certain concentrations of TiO2 NPs. This study will be of benefit to the understanding of the joint effects and physiological mechanism of high-CO2 and nanoparticle to terrestrial plants.