Development and validation of a random forest model to predict functional outcome in patients with intracerebral hemorrhage.

OBJECTIVE: To develop and validate a machine learning (ML)-based model to predict functional outcome in Chinese patients with intracerebral hemorrhage (ICH). METHODS: This retrospective cohort study enrolled patients with ICH between November 2017 and November 2020. The follow-up period ended in February 2021. The study population was divided into training and testing sets with a ratio of 7:3. All variables were included in the least absolute shrinkage and selection operator (LASSO) regression for feature selection. The selected variables were incorporated into the random forest algorithm to construct the prediction model. The predictive performance of the model was evaluated via the area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and calibration curve. RESULTS: A total of 412 ICH patients were included, with 288 in the training set, and 124 in the testing set. Twelve attributes were selected: neurological deterioration, Glasgow Coma Scale (GCS) score at 24 h, baseline GCS score, time from onset to the emergency room, blood glucose, diastolic blood pressure (DBP) change in 24 h, hematoma volume change in 24 h, systemic immune-inflammatory index (SII), systolic blood pressure (SBP) change in 24 h, serum creatinine, serum sodium, and age. In the testing set, the accuracy, AUC, sensitivity, specificity, PPV, and NPV of the model were 0.895, 0.964, 0.872, 0.906, 0.810, and 0.939, respectively. The calibration curves showed a good calibration capability of the model. CONCLUSION: This developed random forest model performed well in predicting 3-month poor functional outcome for Chinese ICH patients.

Diagnosis of an Acinetobacter pittii from a patient in China with a multiplex PCR-based targeted gene sequencing platform of the cerebrospinal fluid: A case report with literature review.

BACKGROUND: The traditional diagnosis model has great challenges for the etiological diagnosis of the central nervous system (CNS) diseases with similar clinical manifestations, especially for the diagnosis of rare pathogens. It is very important to make rapid and accurate identification of pathogens for guiding clinical choices in administering countermeasures. CASE SUMMARY: On August 22, 2020, a 49 years old Chinese male patient had a headache for two days, and then the computed tomography (CT) scan of the brain showed subarachnoid hemorrhage. Subsequently, he underwent twice craniotomy and about 3 weeks of hospitalization. Since September 20, the patient was in the local rehabilitation hospital for hyperbaric oxygen therapy for about three weeks. Then the patient developed acute purulent meningoencephalitis. In the absence of diagnosis of specific pathogenic bacteria, vancomycin (1 g every 12 hours), ceftazidime (2 g every 8 hours), mannitol dehydration (125 mL, every 8 hours), and sodium valproate (0.4 g tid) was used timely according to cerebrospinal fluid (CSF) examination and clinical manifestations. CSF smear and routine culture test were negative during hospitalization. We used the metagenomic next-generation sequencing (mNGS) analysis of CSF for quick and accurate diagnosis, which identified human herpesvirus type 4 (EBV), Corynebacterium corynebacterium, Achromobacter xylose oxidation, and Acinetobacter baumannii, But the mapping degree was not high. Then, we used the modified method-multiplex PCR-based targeted gene sequencing platform (ptNGS) to detect CSF samples and found that the sequences detected were Acinetobacter pittii (A. pittii) and Staphylococcus epidermidis. S. epidermidis might come from skin colonization during lumbar puncture, so it was excluded from the etiological diagnosis. Therefore, we highly suspected that A. pittii was the pathogen in this case. After about three weeks of hospitalization treatment, the patient's symptoms were relieved. CONCLUSION: In conclusion, empirical medication before the identification of pathogens is very important. The ptNGS may be an effective method for the diagnosis of pathogens.

Realization of High Energy Density Sodium-Ion Hybrid Capacitors through Interface Engineering of Pseudocapacitive 3D-CoO-NrGO Hybrid Anodes.

Sodium-ion hybrid capacitors (SHCs) have attracted great attention owing to the improved power density and cycling stability in comparison with sodium-ion batteries. Nevertheless, the energy density (<100 Wh·kg-1) is usually limited by low specific capacity anodes (<150 mAh·g-1) and "kinetics mismatch" between the electrodes. Hence, we report a high energy density (153 Wh·kg-1) SHC based on a highly pseudocapacitive interface-engineered 3D-CoO-NrGO anode. This high-performance anode (445 mAh·g-1 @0.025 A·g-1, 135 mAh·g-1 @5.0 A·g-1) consists of CoO (∼6 nm) nanoparticles chemically bonded to the NrGO network through Co-O-C bonds. Exceptional pseudocapacitive charge storage (up to ∼81%) and capacity retention (∼80% after 5000 cycles) are also identified for this SHC. Excellent performance of the 3D-CoO-NrGO anode and SHC is owing to the synergistic effect of the CoO conversion reaction and pseudocapacitive sodium-ion storage induced by numerous Na2O/Co/NrGO nanointerfaces. Co-O-C bonds and the 3D microstructure facilitating efficient strain relaxation and charge-transfer correspondingly are also identified as vital factors accountable for the excellent electrochemical performance. The interface-engineering strategy demonstrated provides opportunities to design high-performance transition metal oxide-based anodes for advanced SHCs.

High-Energy-Density Sodium-Ion Hybrid Capacitors Enabled by Interface-Engineered Hierarchical TiO2 Nanosheet Anodes.

Sodium-ion hybrid capacitors are known for their high power densities and superior cycle life compared to Na-ion batteries. However, low energy densities (<100 Wh kg-1) due to the lack of high-capacity (>150 mAh g-1) anodes capable of fast charging are delaying their practical implementation. Herein, we report a high-performance Na-ion hybrid capacitor based on an interface-engineered hierarchical TiO2 nanosheet anode consisting of bronze (∼15%) and anatase (∼85%) crystallites (∼10 nm). This pseudocapacitive dual-phase anode demonstrated exceptional specific capacity of 289 mAh g-1 at 0.025 A g-1 and excellent rate capability (110 mAh g-1 at 1.0 A g-1). The Na-ion hybrid capacitor integrating a dual-phase hierarchical TiO2 nanosheet anode and an activated carbon cathode exhibited a high energy density of 200 Wh kg-1 (based on the total mass of active materials in both electrodes) and power density of 6191 W kg-1. These values are in the energy and power density range of Li-ion batteries (100-300 Wh kg-1) and supercapacitors (5000-15 000 W kg-1), respectively. Furthermore, exceptional capacity retention of 80% is observed after 5000 charge-discharge cycles. Outstanding electrochemical performance of the demonstrated Na-ion hybrid capacitor is credited to the enhanced pseudocapacitive Na-ion intercalation of the two-dimensional TiO2 anode resulting from nanointerfaces between bronze and anatase crystallites. Mechanistic investigations evidenced Na-ion storage through intercalation pseudocapacitance with minimal structural changes. This approach of nanointerface-induced pseudocapacitance presents great opportunities toward developing advanced electrode materials for next-generation Na-ion hybrid capacitors.

Cosensitized Quantum Dot Solar Cells with Conversion Efficiency over 12.

The improvement of sunlight utilization is a fundamental approach for the construction of high-efficiency quantum-dot-based solar cells (QDSCs). To boost light harvesting, cosensitized photoanodes are fabricated in this work by a sequential deposition of presynthesized Zn-Cu-In-Se (ZCISe) and CdSe quantum dots (QDs) on mesoporous TiO2 films via the control of the interactions between QDs and TiO2 films using 3-mercaptopropionic acid bifunctional linkers. By the synergistic effect of ZCISe-alloyed QDs with a wide light absorption range and CdSe QDs with a high extinction coefficient, the incident photon-to-electron conversion efficiency is significantly improved over single QD-based QDSCs. It is found that the performance of cosensitized photoanodes can be optimized by adjusting the size of CdSe QDs introduced. In combination with titanium mesh supported mesoporous carbon as a counterelectrode and a modified polysulfide solution as an electrolyte, a champion power conversion efficiency up to 12.75% (Voc = 0.752 V, Jsc = 27.39 mA cm-2 , FF = 0.619) is achieved, which is, as far as it is known, the highest efficiency for liquid-junction QD-based solar cells reported.