Combining clinical and imaging data for predicting functional outcomes after acute ischemic stroke: an automated machine learning approach.

This study aimed to develop and validate an automated machine learning (ML) system that predicts 3-month functional outcomes in acute ischemic stroke (AIS) patients by combining clinical and neuroimaging features. Functional outcomes were categorized as unfavorable (modified Rankin Scale ≥ 3) or not. A clinical model employing optimal clinical features (Model_A), a convolutional neural network model incorporating imaging data (Model_B), and an integrated model combining both imaging and clinical features (Model_C) were developed and tested to predict unfavorable outcomes. The developed models were compared with each other and with traditional risk-scoring models. The dataset comprised 4147 patients from a multicenter stroke registry, with 1268 (30.6%) experiencing unfavorable outcomes. Age, initial NIHSS, and early neurologic deterioration were identified as the most important clinical features. The ML model prediction achieved an area under the curves of 0.757 (95% CI 0.726-0.789) for Model_A, 0.725 (95% CI 0.693-0.755) for Model_B, and 0.786 (95% CI 0.757-0.814) for Model_C in the test set. The integrated models outperformed traditional risk-scoring models by 0.21 (95% CI 0.16-0.25) for HIAT and 0.15 (95% CI 0.11-0.19) for THRIVE. In conclusion, the integrated ML system enhanced stroke outcome prediction by combining imaging data and clinical features, outperforming traditional risk-scoring models.

Interpretable machine learning for prediction of clinical outcomes in acute ischemic stroke.

Background and aims: Predicting the prognosis of acute ischemic stroke (AIS) is crucial in a clinical setting for establishing suitable treatment plans. This study aimed to develop and validate a machine learning (ML) model that predicts the functional outcome of AIS patients and provides interpretable insights. Methods: We included AIS patients from a multicenter stroke registry in this prognostic study. ML-based methods were utilized to predict 3-month functional outcomes, which were categorized as either favorable [modified Rankin Scale (mRS) ≤ 2] or unfavorable (mRS ≥ 3). The SHapley Additive exPlanations (SHAP) method was employed to identify significant features and interpret their contributions to the predictions of the model. Results: The dataset comprised a derivation set of 3,687 patients and two external validation sets totaling 250 and 110 patients each. Among them, the number of unfavorable outcomes was 1,123 (30.4%) in the derivation set, and 93 (37.2%) and 32 (29.1%) in external sets A and B, respectively. Among the ML models used, the eXtreme Gradient Boosting model demonstrated the best performance. It achieved an area under the receiver operating characteristic curve (AUC-ROC) of 0.790 (95% CI: 0.775-0.806) on the internal test set and 0.791 (95% CI: 0.733-0.848) and 0.873 (95% CI: 0.798-0.948) on the two external test sets, respectively. The key features for predicting functional outcomes were the initial NIHSS, early neurologic deterioration (END), age, and white blood cell count. The END displayed noticeable interactions with several other features. Conclusion: ML algorithms demonstrated proficient prediction for the 3-month functional outcome in AIS patients. With the aid of the SHAP method, we can attain an in-depth understanding of how critical features contribute to model predictions and how changes in these features influence such predictions.

Assessment of thrombus using susceptibility-weighted filtered-phase images in patients with acute ischemic stroke.

BACKGROUND AND PURPOSE: Recognizing the location and length of the thrombus responsible for large vessel occlusion in patients with acute ischemic stroke can facilitate effective endovascular recanalization therapy (ERT). We hypothesized that the aliasing or dipole effect produced by filtered-phase susceptibility-weighted imaging (SWI) would facilitate thrombus delineation. METHODS: Of the patients with middle cerebral artery occlusion who underwent ERT, we screened those who underwent noncontrast CT (NCCT), multiphase CT angiography (mCTA), and SWI before the endovascular procedure. We used an arbitrary index termed measurement of equivalence in thrombus assessed by digital subtraction angiography (METAD) defined as having the same location as the thrombus observed in the digital subtraction angiography (DSA) and length differing by less than 5 mm. For NCCT, mCTA, SWI_m (magnitude), and SWI_p (phase), the length of the thrombus and METAD were assessed. RESULTS: The mean lengths of the thrombi determined using NCCT, mCTA, SWI_m, SWI_p, and DSA were 14.03, 13.47, 13.89, 9.93, and 8.96 mm, respectively. The absolute agreement for thrombus length was excellent for SWI_p and DSA (intraclass correlation coefficient [ICC] = .96), moderate for SWI_m and DSA (ICC = .53), and poor for mCTA and DSA (ICC = .14). The METADs were 26.7%, 40.0%, 33.3%, and 73.3% for NCCT, mCTA, SWI_m, and SWI_p, respectively. The METADs for NCCT and SWI_p were significantly different (p = .045) and those for mCTA and SWI_m were not (p = .537 and .093, respectively). CONCLUSIONS: The SWI_p was best matched with the DSA for the measurement of the lengths and locations of thrombi. The use of pre-thrombectomy SWI_p imaging for acute ischemic stroke may facilitate a successful ERT strategy.

Glymphatic system impairment in patients with status epilepticus.

OBJECTIVE: The aim of this study was to compare the function of the glymphatic system in patients with status epilepticus (SE) with that in healthy controls by diffusion tensor image analysis along the perivascular space (DTI-ALPS) method. We also investigated the association between glymphatic system function and the clinical characteristics of SE. METHODS: We retrospectively enrolled 28 patients with SE and 31 healthy controls matched for age and sex. All study participants underwent diffusion tensor imaging using the same 3-T MRI scanner, and the DTI-ALPS index was calculated. We compared the DTI-ALPS index between the SE group and the control group. We also evaluated the associations of the DTI-ALPS index with etiology and type of SE, age, putative duration of seizure, time interval until MRI, seizure-related changes on diffusion-weighted imaging, and any previous structural lesions. RESULTS: The DTI-ALPS index was significantly lower in the SE group than in the control group (1.462 ± 0.297 vs. 1.632 ± 0.270, p = 0.026) and was negatively correlated with age (r = - 0.280, p = 0.032) in the SE group. However, there were no significant between-group differences in the DTI-ALPS index according to other clinical factors. SIGNIFICANCE: The finding of a significantly lower DTI-ALPS index in the SE group suggests that the glymphatic system is impaired in patients with SE. DTI-ALPS is a useful tool for evaluation of the function of the glymphatic system in these patients.

Alterations of limbic structure volumes and limbic covariance network in patients with cluster headache.

The aim of this study was to compare the limbic structures and covariance network in patients with cluster headache to those of healthy controls. We enrolled 23 patients with newly diagnosed cluster headache and 31 healthy controls. They underwent three-dimensional T1-weighted imaging utilizing a 3.0 Tesla MRI scanner. Volumetric analysis of the subcortical limbic structures, including the hippocampus, amygdala, thalamus, mammillary body, hypothalamus, basal forebrain, septal nuclei, fornix, and nucleus accumbens, was performed. We examined the limbic covariance network using a graph theory. The volumes of the limbic structures between patients with cluster headache and healthy controls were significantly different. The volume of the left hippocampus in patients with cluster headache was significantly lower than that in healthy controls (0.256 vs 0.291 %, p = 0.002). Patients with cluster headache showed significant alterations of the limbic covariance network. The average strength, global efficiency, local efficiency, mean clustering coefficient, and transitivity were lower (5.238 vs 10.322, p = 0.030; 0.355 vs 0.608, p = 0.020; 0.547 vs 1.553, p = 0.020; 0.424 vs 0.895, p = 0.016; respectively), whereas the characteristic path length was higher (3.314 vs 1.752, p = 0.040) in patients with cluster headache than in healthy controls. We detected alterations of limbic structure volumes in patients with cluster headache compared to healthy controls, especially in the hippocampus. We also found significant alterations in the limbic covariance network in patients with cluster headache who showed decreased segregation and integration. These abnormalities could be related to the pathophysiology of cluster headache.

Hippocampal injury in patients with status epilepticus: Quantitative analysis of hippocampal volume and structural co-variance network.

OBJECTIVE: This study aimed to evaluate the differences in hippocampal structural volumes and intra-hippocampal networks between patients with status epilepticus (SE) and healthy controls. METHODS: We enrolled 45 patients with SE and 35 age- and sex-matched healthy controls. We excluded patients with active structural lesions, which could be a direct cause of SE, but included patients with co-existing lesions. Co-existing lesions were defined as any lesions possibly related to the occurrence of SE, including encephalomalacia, cavernous malformation, dural arteriovenous fistula, and normal pressure hydrocephalus, etc. We divided 45 patients into those with co-existing lesions (n = 21) and those without co-existing lesions (n = 24). We conducted a volumetric analysis using FreeSurfer (version 7), and the intra-hippocampal structural co-variance network was analyzed with a graph theoretical analysis based on the structural volumes of the hippocampal subfields. RESULTS: The structural volumes and intra-hippocampal structural co-variance networks were not different between patients with and without co-existing lesions. However, both structural volumes and intra-hippocampal structural co-variance networks were significantly different in patients with SE compared to healthy controls, and the ratio of the volume difference: [(volume of controls-volume of patients)/volume of controls] was highest in the left hippocampus (0.195), left amygdala (0.143), left thalamus (0.126), and right cortex (0.084). In addition, the global connectivity measurements including radius, diameter, eccentricity, and assortativity were significantly increased, and the small-worldness index was significantly decreased in patients with SE. Notably, structural volumes were negatively related to age but not to the duration of SE. SIGNIFICANCE: Our study revealed significant alterations in structural volumes and intra-hippocampal structural co-variance networks in patients with SE compared to healthy controls, even though hippocampal atrophy was not evident on visual analysis; this is likely due to the direct effect of SE itself.

Compression injury of the circular stapler for gastrointestinal end-to-end anastomosis: preliminary in-vitro study.

PURPOSE: This preliminary in-vitro study was designed to evaluate the risk factors of compression injury from use of a circular stapler for end-to-end anastomosis. METHODS: Transparent collagen plates were prepared in dry and wet conditions. Physical properties of collagen plates and porcine colon tissue were examined using a rheometer. Adjustable and fixed-type circular staplers were applied on the collagen plates and the gap distance and compressive pressure were measured during anvil approximation. Tissue injury was evaluated using a compression injury scale. Compression properties were accessed to optimal or overcompression based on gap distance. RESULTS: Unacceptable injuries were rarely observed on the dry collagens, regardless of compression device. In the adjustable compression, the compressibility ratio was similar between dry and wet collagen. Overcompression and unacceptable injury increased on the wet collagens. In the fixed compression, the compressibility ratio increased significantly and unacceptable injuries were observed in more than 50% of wet collagens. Peak pressure was significantly higher in the fixed-compression types than those of adjustable type. On bivariate correlation analysis, fixed-compression type and wet collagens were respectively associated with overcompression. On multivariate analysis, edematous collagen condition was the most important risk factor and proximal anvil side, fixed compression type, and overcompression were also independent risk factors for unacceptable compression injury. CONCLUSION: In the edematous tissue condition, unintentional overcompression could be increased and result in tissue injury on the compression line of the circular stapler.

Proteomic Profiling of Emiliania huxleyi Using a Three-Dimensional Separation Method Combined with Tandem Mass Spectrometry.

Emiliania huxleyi is one of the most abundant marine planktons, and it has a crucial feature in the carbon cycle. However, proteomic analyses of Emiliania huxleyi have not been done extensively. In this study, a three-dimensional liquid chromatography (3D-LC) system consisting of strong cation exchange, high- and low-pH reversed-phase liquid chromatography was established for in-depth proteomic profiling of Emiliania huxleyi. From tryptic proteome digest, 70 fractions were generated and analyzed using liquid chromatography-tandem mass spectrometry. In total, more than 84,000 unique peptides and 10,000 proteins groups were identified with a false discovery rate of ≤0.01. The physicochemical properties of the identified peptides were evaluated. Using ClueGO, approximately 700 gene ontology terms and 15 pathways were defined from the identified protein groups with p-value ≤0.05, covering a wide range of biological processes, cellular components, and molecular functions. Many biological processes associated with CO2 fixation, photosynthesis, biosynthesis, and metabolic process were identified. Various molecular functions relating to protein binding and enzyme activities were also found. The 3D-LC strategy is a powerful approach for comparative proteomic studies on Emiliania huxleyi to reveal changes in its protein level and related mechanism.

Characteristics of seizure-induced signal changes on MRI in patients with first seizures.

PURPOSE: The aim of this study was to investigate the predictive factors and identify the characteristics of the seizure-induced signal changes on MRI (SCM) in patients with first seizures. METHODS: We conducted a retrospective study of patients with first seizures from March 2010 to August 2014. The inclusion criteria for this study were patients with 1) first seizures, and 2) MRI and EEG performed within 24h of the first seizures. The definition of SCM was hyper-intensities in the brain not applying to cerebral arterial territories. Multivariate logistic regression was performed with or without SCM as a dependent variable. RESULTS: Of 431 patients with seizures visiting the ER, 69 patients met the inclusion criteria. Of 69 patients, 11 patients (15.9%) had SCM. Epileptiform discharge on EEG (OR 29.7, 95% CI 1.79-493.37, p=0.018) was an independently significant variable predicting the presence of SCM in patients with first seizures. In addition, the topography of SCM was as follows; i) ipsilateral hippocampus, thalamus and cerebral cortex (5/11), ii) unilateral cortex (4/11), iii) ipsilateral thalamus and cerebral cortex (1/11), iv) bilateral hippocampus (1/11). Moreover, 6 out of 7 patients who underwent both perfusion CT and MRI exhibited unilateral cortical hyperperfusion with ipsilateral thalamic involvement reflecting unrestricted vascular territories. CONCLUSION: There is an association between epileptiform discharges and SCM. Additionally, the involvement of the unilateral cortex and ipsilateral thalamus in SCM and its hyperperfusion state could be helpful in differentiating the consequences of epileptic seizures from other pathologies.

Clinical Significance of Fluid-Attenuated Inversion Recovery Vascular Hyperintensities in Borderzone Infarcts.

BACKGROUND AND PURPOSE: Fluid-attenuated inversion recovery vascular hyperintensities (FVHs) are seen in some cases with cerebral hemodynamic impairment and collateral flow. Because the worst outcomes of patients with borderzone infarcts were mainly correlated with impaired hemodynamics, the presence of FVH might provide another clue for predicting the prognosis of patients with borderzone infarcts. METHODS: We reviewed 1377 consecutive patients with ischemic stroke. Cortical borderzone (CBZ) and internal borderzone infarcts were selected based on diffusion-weighted imaging. FVHs were defined as tubular- or serpentine-shaped hyperintensities in the subarachnoid space. We investigated the clinical significance of FVHs in borderzone-infarcted patients. RESULTS: Among 87 patients with borderzone infarcts, the presence of FVH was observed in 30 (34.5%). We identified 62 patients with CBZ infarcts and 25 patients with internal borderzone infarcts. In the cases with CBZ infarcts, the initial National Institutes of Health Stroke Scale scores and the portions of nonfavorable outcome at 3 months in the FVH(+) group were significantly higher than in the FVH(-) group (P<0.05 and P<0.001, respectively). Unlike the cases with CBZ infarcts, there were no significant differences of these clinical features between the FVH(+) group and the FVH(-) group in the patients with internal borderzone infarcts. CONCLUSIONS: The findings of FVH are associated with relatively severe clinical presentation and nonfavorable prognosis in patients with CBZ infarcts, but not in patients with internal borderzone infarcts. The presence of FVH may help to identify CBZ-infarcted patients who require close observation and hemodynamic control.

Two conserved cysteine residues are critical for the enzymic function of the human platelet-derived growth factor receptor-beta: evidence for different roles of Cys-822 and Cys-940 in the kinase activity.

The platelet-derived growth factor receptor-beta (PDGFR-beta) has a number of conserved cysteine residues on its cytoplasmic domain. We have examined whether the cysteine residues play a role in the enzymic function of PDGFR-beta. We found that N-ethylmaleimide, which selectively alkylates free thiol groups of cysteine residues, completely inhibited the kinase activity of PDGFR-beta. We then identified, through site-directed mutagenesis, two conserved cysteine residues critical for the enzymic function of PDGFR-beta. Cys to Ser mutations for either Cys-822, positioned in the catalytic loop, or Cys-940, located in the C-terminal kinase subdomain, significantly reduced the activities of autophosphorylation and phosphorylation towards exogenous substrates. The non-reducing gel analysis indicated that neither of these cysteine residues contributes to the kinase activity by disulphide-bond formation. In addition, the individual mutation of Cys-822 and Cys-940 had no effect on protein stability or the binding of substrates or ATP, implying that these cysteine residues are involved in enzyme catalysis. Finally, proteolytic cleavage assays showed that the mutation of Cys-940, but not Cys-822, induced a protein conformational change. Taken together, these results suggest that Cys-940 contributes to the catalytic activity of PDGFR-beta by playing a structural role, whereas Cys-822 contributes through a different mechanism.

Structural basis for the NAD-dependent deacetylase mechanism of Sir2.

The NAD-dependent histone/protein deacetylase activity of Sir2 (silent information regulator 2) accounts for its diverse biological roles including gene silencing, DNA damage repair, cell cycle regulation, and life span extension. We provide crystallographic evidence that 2'-O-acetyl ADP-ribose is the reaction product that is formed at the active site of Sir2 from the 2.6-A co-crystal structure of 2'-O-acetyl-ADP-ribose and Sir2 from Archaeoglobus fulgidus. In addition, we show that His-116 and Phe-159 play critical roles in the catalysis and substrate recognition. The conserved Ser-24 and Asp-101 contribute to the stability for NAD binding rather than being directly involved in the catalysis. The crystal structures of wild type and mutant derivatives of Sir2, in conjunction with biochemical analyses of the mutants, provide novel insights into the reaction mechanism of Sir2-mediated deacetylation.