Arterial Recanalization During Interhospital Transfer for Thrombectomy.

BACKGROUND: Patients with acute ischemic stroke harboring a large vessel occlusion admitted to nonendovascular-capable centers often require interhospital transfer for thrombectomy. We evaluated the incidence and predictors of arterial recanalization during transfer, as well as the relationship between interhospital recanalization and clinical outcomes. METHODS: We analyzed data from 2 cohorts of patients with an anterior circulation large vessel occlusion transferred for consideration of thrombectomy to a comprehensive center, with arterial imaging at the referring hospital and on comprehensive stroke center arrival. Interhospital recanalization was determined by comparison of the baseline and posttransfer arterial imaging and was defined as revised arterial occlusive lesion (rAOL) score 2b to 3. Pretransfer variables independently associated with interhospital recanalization were studied using multivariable logistic regression analysis. RESULTS: Of the 520 included patients (Montpellier, France, n=237; Stanford, United States, n=283), 111 (21%) experienced interhospital recanalization (partial [rAOL=2b] in 77% and complete [rAOL=3] in 23%). Pretransfer variables independently associated with recanalization were intravenous thrombolysis (adjusted odds ratio, 6.8 [95% CI, 4.0-11.6]), more distal occlusions (intracranial carotid occlusion as reference: adjusted odds ratio, 2.0 [95% CI, 0.9-4.5] for proximal first segment of the middle cerebral artery, 5.1 [95% CI, 2.3-11.5] for distal first segment of the middle cerebral artery, and 5.0 [95% CI, 2.1-11.8] for second segment of the middle cerebral artery), and smaller clot burden (clot burden score 0-4 as reference: adjusted odds ratio, 3.4 [95% CI, 1.5-7.6] for 5-7 and 5.6 [95% CI, 2.4-12.7] for 8-9). Recanalization on arrival at the comprehensive center was associated with less interhospital infarct growth (rAOL, 0-2a: 11.6 mL; rAOL, 2b: 2.2 mL; rAOL, 3: 0.6 mL; Ptrend<0.001) and greater interhospital National Institutes of Health Stroke Scale score improvement (0 versus -5 versus -6; Ptrend<0.001). Interhospital recanalization was associated with reduced 3-month disability (adjusted common odds ratio, 2.51 [95% CI, 1.68-3.77]) with greater benefit from complete than partial recanalization. CONCLUSIONS: Recanalization is frequently observed during interhospital transfer for thrombectomy and is strongly associated with favorable outcomes, even when partial. Broadening thrombolysis indications in primary centers, and developing therapies that increase recanalization during transfer, will likely improve clinical outcomes.

Clinical change during inter-hospital transfer for thrombectomy: Incidence, associated factors, and relationship with outcome.

BACKGROUND: Patients with acute ischemic stroke with a large vessel occlusion (LVO) admitted to non endovascular-capable centers often require inter-hospital transfer for thrombectomy. We aimed to describe the incidence of substantial clinical change during transfer, the factors associated with clinical change, and its relationship with 3-month outcome. METHODS: We analyzed data from two cohorts of acute stroke patients transferred for thrombectomy to a comprehensive center (Stanford, USA, November 2019 to January 2023; Montpellier, France, January 2015 to January 2017), regardless of whether thrombectomy was eventually attempted. Patients were included if they had evidence of an LVO at the referring hospital and had a National Institute of Health Stroke Scale (NIHSS) score documented before and immediately after transfer. Inter-hospital clinical change was categorized as improvement (⩾4 points and ⩾25% decrease between the NIHSS score in the referring hospital and upon comprehensive center arrival), deterioration (⩾4 points and ⩾25% increase), or stability (neither improvement nor deterioration). The stable group was considered as the reference and was compared to the improvement or deterioration groups separately. RESULTS: A total of 504 patients were included, of whom 22% experienced inter-hospital improvement, 14% deterioration, and 64% were stable. Pre-transfer variables independently associated with clinical improvement were intravenous thrombolysis use, more distal occlusions, and lower serum glucose; variables associated with deterioration included more proximal occlusions and higher serum glucose. On post-transfer imaging, clinical improvement was associated with arterial recanalization and smaller infarct growth and deterioration with larger infarct growth. As compared to stable patients, those with clinical improvement had better 3-month functional outcome (adjusted common odds ratio (cOR) = 2.43; 95% confidence interval (CI) = 1.59-3.71; p < 0.001), while those with deterioration had worse outcome (adjusted cOR = 0.60; 95% CI = 0.37-0.98; p = 0.044). CONCLUSION: Substantial inter-hospital clinical changes are frequently observed in LVO-related ischemic strokes, with significant impact on functional outcome. There is a need to develop treatments that improves the clinical status during transfer. DATA ACCESS STATEMENT: The data that support the findings of this study are available upon reasonable request.

Is Learning in Biological Neural Networks Based on Stochastic Gradient Descent? An Analysis Using Stochastic Processes.

In recent years, there has been an intense debate about how learning in biological neural networks (BNNs) differs from learning in artificial neural networks. It is often argued that the updating of connections in the brain relies only on local information, and therefore a stochastic gradient-descent type optimization method cannot be used. In this article, we study a stochastic model for supervised learning in BNNs. We show that a (continuous) gradient step occurs approximately when each learning opportunity is processed by many local updates. This result suggests that stochastic gradient descent may indeed play a role in optimizing BNNs.

Are Gamers Prone to eThrombosis during Long Gaming Sessions?

During the last two decades, several cases of venous thrombosis (VTE) after a prolonged period at a computer have been described, denominated as "eThrombosis". Video gaming on a computer has become very popular and can be a social activity where several players gather to play against each other or in a virtual environment for several days ("LAN (i.e., Local Area Network) parties") where the participants are sedentary and consuming calorie-rich food items. The aim of this study was to investigate potential coagulation activation during a 42 h LAN party. Nine male gamers volunteered for the LAN party. Citrated blood was sampled before and every 6 h, and plasma was analyzed for thrombin generation, thrombin-antithrombin complexes (TAT), prothrombin fragment 1 + 2 (F1 + 2), and D-dimer. Thrombin generation increased slightly but not significantly during the LAN party, whereas the coagulation activation markers were unchanged. These results do not indicate that the coagulation system is activated significantly during 42 h of gaming with minimal physical activity. Although increased activity cannot be excluded, it does not directly indicate a risk of VTE in general.

Survey of Danish Head and Neck Cancer Patients' Positions on Personalized Medicine, Gene Tests, and Personalized Follow-Up.

The field of personalized medicine (PM) has grown rapidly because of the "omics revolution", but PM may be difficult for patients to comprehend. This study sought to explore head and neck cancer (HNC) patients' positions and knowledge of PM, gene tests, and follow-up and to compare HNC patients' positions to a sample from a national Danish questionnaire. To do this, patients with prior HNC were invited to participate in a questionnaire. Initial interviews revealed a heterogenic understanding of PM between patients. A total of 226 patients were included in the survey and 177 patients with complete data were included for analysis. Most patients were more positive than negative towards gene tests and gene research (83% and 93%, respectively), but 72% had little or no knowledge of the subject. Almost all patients, 98%, were satisfied with their follow-up. Significantly more patients with HNC were positive towards gene research compared to a sample from a national Danish questionnaire (p < 0.001). Patients with HNC were positive towards gene tests and PM, but patients may not understand or comprehend the information given, and it is important to inform and educate patients and health professionals to establish common ground in PM.

Exploring D-Lactate as a Biomarker for Acute Intestinal Necrosis in 2958 Patients: A Prospective Cross-Sectional Study.

BACKGROUND: Timely diagnosis of acute intestinal necrosis (AIN) is lifesaving, but challenging due to unclear clinical presentation. D-lactate has been proposed as an AIN biomarker. OBJECTIVES: We aimed to test the diagnostic performance in a clinical setting. METHODS: We performed a cross-sectional prospective study, including all adult patients with acute referral to a single tertiary gastrointestinal surgical department during 2015-2016 and supplemented by enrollment of high-risk in-hospital patients suspected of having AIN during 2016-2019. AIN was verified intraoperatively, and D-lactate was analyzed using an automatic spectrophotometric set-up. A D-lactate cut-off for AIN was estimated using the receiver operating characteristic curve. The performance according to patient subgroups was estimated using the area under the receiver operating characteristic curve (AUC). Given the exploratory nature of this study, a formal power calculation was not feasible. RESULTS: Forty-four AIN patients and 2914 controls were enrolled. The D-lactate cut-off was found to be 0.0925 mM. Due to lipemic interference, D-lactate could not be quantified in half of the patients, leaving 23 AIN patients and 1456 controls for analysis. The AUC for the diagnosis of AIN by D-lactate was 0.588 (95% confidence interval 0.475-0.712), with a sensitivity of 0.261 and specificity of 0.892. Analysis of high-risk patients showed similar results (AUC 0.579; 95% confidence interval 0.422-0.736). CONCLUSION: D-lactate showed low sensitivity for AIN in both average-risk and high-risk patients. Moreover, lipemic interference precluded valid spectrophotometric assessment of D-lactate in half of the patients, further disqualifying the clinical utility of D-lactate as a diagnostic marker for AIN.

Global N2O emissions from our planet: Which fluxes are affected by man, and can we reduce these?

In some places, N2O emissions have doubled during the last 2-3 decades. Therefore, it is crucial to identify N2O emission hotspots from terrestrial and aquatic systems. Large variation in N2O emissions occur in managed as well as in natural areas. Natural unmanaged tropical and subtropical wet forests are important N2O sources globally. Emission hotspots, often coupled to human activities, vary across climate zones, whereas N2O emissions are most often a few kg N ha-1 year-1 from arable soils, drained organic soils in the boreal and temperate zones often release 20-30 kg N ha-1 year-1. Similar high N2O emissions occur from some tropical crops like tea, palm oil and bamboo. This strong link between increased N2O emissions and human activities highlight the potential to mitigate large emissions. In contrast, water where oxic and anoxic conditions meet are N2O emission hotspots as well, but not possible to reduce.

DLG1 functions upstream of SDCCAG3 and IFT20 to control ciliary targeting of polycystin-2.

Polarized vesicular trafficking directs specific receptors and ion channels to cilia, but the underlying mechanisms are poorly understood. Here we describe a role for DLG1, a core component of the Scribble polarity complex, in regulating ciliary protein trafficking in kidney epithelial cells. Conditional knockout of Dlg1 in mouse kidney caused ciliary elongation and cystogenesis, and cell-based proximity labelling proteomics and fluorescence microscopy showed alterations in the ciliary proteome upon loss of DLG1. Specifically, the retromer-associated protein SDCCAG3, IFT20 and polycystin-2 (PC2) were reduced in cilia of DLG1 deficient cells compared to control cells. This phenotype was recapitulated in vivo and rescuable by re-expression of wildtype DLG1, but not a Congenital Anomalies of the Kidney and Urinary Tract (CAKUT)-associated DLG1 variant, p.T489R. Finally, biochemical approaches and Alpha Fold modelling suggested that SDCCAG3 and IFT20 form a complex that associates, at least indirectly, with DLG1. Our work identifies a key role for DLG1 in regulating ciliary protein composition and suggests that ciliary dysfunction of the p.T489R DLG1 variant may contribute to CAKUT.

Expression of genes and enzymes involved in ovarian steroidogenesis in relation to human follicular development.

Introduction: Granulosa cells (GCs) and theca cells (TCs) play a pivotal role in human ovarian steroidogenesis, facilitating the conversion of cholesterol into sex steroids that regulate normal reproductive function. This study aims to explore the expression patterns of key enzymes that govern human ovarian steroidogenesis throughout follicle development, employing both genomic and immunological methodologies. Methods: Follicles and GCs obtained from women undergoing ovarian tissue cryopreservation (OTC) and in vitro fertilisation treatment were utilized. Gene expression data were obtained from a Chinese study using RNA sequencing and from microarray data generated in our laboratory to comprehensively analyse gene expression profiles across distinct stages of follicular development. To corroborate the localisation of key enzymes within GCs and TCs, immunohistochemistry analyses utilizing colourimetric and fluorescent techniques were conducted. Results: Steroidogenesis-related enzymes displayed low gene expression levels during early follicle development. However, a notable upregulation of HSD3B2 was observed in GCs as follicles progressed to the antral/preovulatory stage, confirmed consistently using both microarray and RNA sequencing methodologies. Furthermore, immunohistochemical analyses effectively demonstrated that HSD3B2 were not only expressed in GCs, but co-localised with CYP17A1 within a specific subset of TCs surrounding human small antral follicles. Contributing to an enhanced progesterone production during the second half of the follicular phase was a significant upregulation of CYB5A in both microarray and RNA-seq datasets as follicles transition from the antral stage to the pre-ovulatory stage. Moreover, an augmented expression of DHCR24 and LDLR in both types of data, along with HMGCR expression expression in the microarray data, indicates increased substrate availability for ovarian steroidogenesis. Discussion: This study confirms and extends that GCs gradually augment expression of HSD3B2 thereby enhancing their capacity for progesterone synthesis as follicles reach the size of selection at around 10 mm in diameter. This is supported by the expression CYB5A and possibly augmented availability of steroid precursors. A subset of TCs exhibit concurrent expression of CYP17A1 and HSD3B2, collectively contributing to the synthesis of 17-hydroxyprogesterone. These data significantly enhance our understanding of the dynamic regulation of progesterone throughout the process of follicular development.

Intake of marine n-3 polyunsaturated fatty acids and risk of hemorrhagic stroke and its subtypes: a Danish follow-up study.

BACKGROUND: A diet rich in marine n-3 polyunsaturated fatty acids (PUFAs) may lower the risk of coronary heart disease and ischemic stroke. However, the association between intake of marine n-3 PUFAs and risk of hemorrhagic stroke has only been sparsely explored. We aimed to investigate the associations between intake of the major marine n-3 PUFAs, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) and their sum in relation to incident hemorrhagic stroke and its subtypes intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH). METHODS: We analyzed data from the Danish Diet, Cancer and Health Cohort, which was established between 1993 and 1997. Information on dietary intake of marine n-3 PUFAs was obtained through a validated food frequency questionnaire. Potential hemorrhagic stroke cases were identified by linkage to the Danish National Patient Register and subsequently validated. Hazard ratios obtained by Cox proportional hazard regression were used as measures of association. RESULTS: A total of 394 subjects among 55,519 individuals developed hemorrhagic stroke during a median follow-up period of 13.5 years. In multivariable analyses including adjustment for established risk factors, we observed weak and statistically non-significant indications of inverse associations between intake of EPA, DHA, and EPA + DHA and the rate of incident hemorrhagic stroke. In analyses of hemorrhagic stroke subtypes, we found indications of lower rates of ICH among participants in the highest quartile of EPA, DHA, and EPA + DHA compared with those in the lowest quartile, and indications of lower rates of SAH in the highest quartile of EPA intake compared to the lowest quartile but the findings were statistically non-significant. CONCLUSIONS: Indications of inverse statistically non-significant associations were found between EPA, DHA, and EPA + DHA and hemorrhagic stroke.

Semiautomated Detection of Early Infarct Signs on Noncontrast CT Improves Interrater Agreement.

BACKGROUND: Acute ischemic infarct identification on noncontrast computed tomography (NCCT) is highly variable between raters. A semiautomated method for segmentation of acute ischemic lesions on NCCT may improve interrater reliability. METHODS: Patients with successful endovascular reperfusion from the DEFUSE 3 trial (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke) were included. We created relative NCCT (rNCCT) color-gradient overlays by comparing the density of a voxel on NCCT to the homologous region in the contralateral hemisphere. Regions with a relative hypodensity of at least 5% were visualized. We coregistered baseline and follow-up images. Two neuroradiologists and 6 nonradiologists segmented the acute ischemic lesion on the baseline scans with 2 methods: (1) manually outlining hypodense regions on the NCCT (unassisted segmentation) and (2) manually excluding areas deemed outside of the ischemic lesion on the rNCCT color map (rNCCT-assisted segmentation). Voxelwise interrater agreement was quantified using the Dice similarity coefficient and volumetric agreement between raters with the detection index (DI), defined as the true positive volume minus the false positive volume. RESULTS: From a total of 92, we included 61 patients. Median age was 59 (64-77), and 57% were female. Stroke onset was known in 39%. Onset to NCCT was median, 8.5 hours (7-11) and median 10 hours (8.4-11.5) in patients with known and unknown onset, respectively. Compared with unassisted NCCT segmentation, rNCCT-assisted segmentation increased the Dice similarity coefficient by >50% for neuroradiologists (Dice similarity coefficient, 0.38 versus 0.83; P<0.001) and nonradiologists (Dice similarity coefficient, 0.14 versus 0.84; P<0.001), and improved the DI among nonradiologists (mean improvement, 5.8 mL [95% CI, 3.1-8.5] mL, P<0.001) but not among neuroradiologists. CONCLUSIONS: The high variability of manual segmentations of the acute ischemic lesion on NCCT is greatly reduced using semiautomated rNCCT. The rNCCT map may therefore aid acute infarct detection and provide more reliable infarct estimates for clinicians with less experience.

Gastrointestinal bleeding with direct oral anticoagulants in patients with atrial fibrillation and anaemia.

INTRODUCTION: A high risk of gastrointestinal bleeding has been reported with the use of some direct oral anticoagulants (DOACs). This risk may be of particular concern in individuals with associated anaemia. The aim of this study is to investigate potential differences in the risks of gastrointestinal bleeding and stroke among the four available DOACs in patients with atrial fibrillation (AF) and moderate or severe anaemia. MATERIALS AND METHODS: All Danish patients diagnosed with incident AF who had a baseline haemoglobin measurement and subsequently initiated DOAC therapy between 2012 and 2021 were identified through administrative registries. Only patients with moderate or severe anaemia (N = 7269) were included and evaluated regarding the risk of hospitalization for gastrointestinal bleeding and stroke. Standardized absolute 1-year risks of stroke and gastrointestinal bleeding were calculated from multivariable Cox regression analyses. DOACs were compared pairwise RESULTS: Compared with apixaban, both dabigatran and rivaroxaban were associated with a significantly increased risk of gastrointestinal bleeding with standardized 1-year risk ratios of 1.73 (95 % confidence interval [CI], 1.10-2.35) and 1.56 (95 % CI, 1.18-1.93), respectively, while no significant difference was seen in the comparison of apixaban with edoxaban 1.32 (95 % CI, 0.41-2.32). No significant differences in gastrointestinal bleeding were observed with pairwise comparisons of dabigatran, rivaroxaban and edoxaban. Finally, no significant difference in stroke risk among the four DOACs was observed. CONCLUSION: In AF patients with moderate or severe anaemia, apixaban was associated with a significantly lower risk of gastrointestinal bleeding than dabigatran and rivaroxaban. No significant difference in stroke risk was observed across all four available DOACs.

Factors Associated With Fast Early Infarct Growth in Patients With Acute Ischemic Stroke With a Large Vessel Occlusion.

BACKGROUND AND OBJECTIVES: The optimal methods for predicting early infarct growth rate (EIGR) in acute ischemic stroke with a large vessel occlusion (LVO) have not been established. We aimed to study the factors associated with EIGR, with a focus on the collateral circulation as assessed by the hypoperfusion intensity ratio (HIR) on perfusion imaging, and determine whether the associations found are consistent across imaging modalities. METHODS: Retrospective multicenter international study including patients with anterior circulation LVO-related acute stroke with witnessed stroke onset and baseline perfusion imaging (MRI or CT) performed within 24 hours from symptom onset. To avoid selection bias, patients were selected from (1) the prospective registries of 4 comprehensive stroke centers with systematic use of perfusion imaging and including both thrombectomy-treated and untreated patients and (2) 1 prospective thrombectomy study where perfusion imaging was acquired per protocol, but treatment decisions were made blinded to the results. EIGR was defined as infarct volume on baseline imaging divided by onset-to-imaging time and fast progressors as EIGR ≥10 mL/h. The HIR, defined as the proportion of time-to-maximum (Tmax) >6 second with Tmax >10 second volume, was measured on perfusion imaging using RAPID software. The factors independently associated with fast progression were studied using multivariable logistic regression models, with separate analyses for CT- and MRI-assessed patients. RESULTS: Overall, 1,127 patients were included (CT, n = 471; MRI, n = 656). Median age was 74 years (interquartile range [IQR] 62-83), 52% were male, median NIH Stroke Scale was 16 (IQR 9-21), median HIR was 0.42 (IQR 0.26-0.58), and 415 (37%) were fast progressors. The HIR was the primary factor associated with fast progression, with very similar results across imaging modalities: The proportion of fast progressors was 4% in the first HIR quartile (i.e., excellent collaterals), ∼15% in the second, ∼50% in the third, and ∼77% in the fourth (p < 0.001 for each imaging modality). Fast progression was independently associated with poor 3-month functional outcome in both the CT and MRI cohorts (p < 0.001 and p = 0.030, respectively). DISCUSSION: The HIR is the primary factor associated with fast infarct progression, regardless of imaging modality. These results have implication for neuroprotection trial design, as well as informing triage decisions at primary stroke centers.

Non-inferiority of deep learning ischemic stroke segmentation on non-contrast CT within 16-hours compared to expert neuroradiologists.

We determined if a convolutional neural network (CNN) deep learning model can accurately segment acute ischemic changes on non-contrast CT compared to neuroradiologists. Non-contrast CT (NCCT) examinations from 232 acute ischemic stroke patients who were enrolled in the DEFUSE 3 trial were included in this study. Three experienced neuroradiologists independently segmented hypodensity that reflected the ischemic core on each scan. The neuroradiologist with the most experience (expert A) served as the ground truth for deep learning model training. Two additional neuroradiologists' (experts B and C) segmentations were used for data testing. The 232 studies were randomly split into training and test sets. The training set was further randomly divided into 5 folds with training and validation sets. A 3-dimensional CNN architecture was trained and optimized to predict the segmentations of expert A from NCCT. The performance of the model was assessed using a set of volume, overlap, and distance metrics using non-inferiority thresholds of 20%, 3 ml, and 3 mm, respectively. The optimized model trained on expert A was compared to test experts B and C. We used a one-sided Wilcoxon signed-rank test to test for the non-inferiority of the model-expert compared to the inter-expert agreement. The final model performance for the ischemic core segmentation task reached a performance of 0.46 ± 0.09 Surface Dice at Tolerance 5mm and 0.47 ± 0.13 Dice when trained on expert A. Compared to the two test neuroradiologists the model-expert agreement was non-inferior to the inter-expert agreement, [Formula: see text]. The before, CNN accurately delineates the hypodense ischemic core on NCCT in acute ischemic stroke patients with an accuracy comparable to neuroradiologists.

EEG pre-burst suppression: characterization and inverse association with preoperative cognitive function in older adults.

The most common complication in older surgical patients is postoperative delirium (POD). POD is associated with preoperative cognitive impairment and longer durations of intraoperative burst suppression (BSup) - electroencephalography (EEG) with repeated periods of suppression (very low-voltage brain activity). However, BSup has modest sensitivity for predicting POD. We hypothesized that a brain state of lowered EEG power immediately precedes BSup, which we have termed "pre-burst suppression" (preBSup). Further, we hypothesized that even patients without BSup experience these preBSup transient reductions in EEG power, and that preBSup (like BSup) would be associated with preoperative cognitive function and delirium risk. Data included 83 32-channel intraoperative EEG recordings of the first hour of surgery from 2 prospective cohort studies of patients ≥age 60 scheduled for ≥2-h non-cardiac, non-neurologic surgery under general anesthesia (maintained with a potent inhaled anesthetic or a propofol infusion). Among patients with BSup, we defined preBSup as the difference in 3-35 Hz power (dB) during the 1-s preceding BSup relative to the average 3-35 Hz power of their intraoperative EEG recording. We then recorded the percentage of time that each patient spent in preBSup, including those without BSup. Next, we characterized the association between percentage of time in preBSup and (1) percentage of time in BSup, (2) preoperative cognitive function, and (3) POD incidence. The percentage of time in preBSup and BSup were correlated (Spearman's ρ [95% CI]: 0.52 [0.34, 0.66], p < 0.001). The percentage of time in BSup, preBSup, or their combination were each inversely associated with preoperative cognitive function (ß [95% CI]: -0.10 [-0.19, -0.01], p = 0.024; -0.04 [-0.06, -0.01], p = 0.009; -0.04 [-0.06, -0.01], p = 0.003, respectively). Consistent with prior literature, BSup was significantly associated with POD (odds ratio [95% CI]: 1.34 [1.01, 1.78], p = 0.043), though this association did not hold for preBSup (odds ratio [95% CI]: 1.04 [0.95, 1.14], p = 0.421). While all patients had ≥1 preBSup instance, only 20.5% of patients had ≥1 BSup instance. These exploratory findings suggest that future studies are warranted to further study the extent to which preBSup, even in the absence of BSup, can identify patients with impaired preoperative cognition and/or POD risk.

USE-Evaluator: Performance metrics for medical image segmentation models supervised by uncertain, small or empty reference annotations in neuroimaging.

Performance metrics for medical image segmentation models are used to measure the agreement between the reference annotation and the predicted segmentation. Usually, overlap metrics, such as the Dice, are used as a metric to evaluate the performance of these models in order for results to be comparable. However, there is a mismatch between the distributions of cases and the difficulty level of segmentation tasks in public data sets compared to clinical practice. Common metrics used to assess performance fail to capture the impact of this mismatch, particularly when dealing with datasets in clinical settings that involve challenging segmentation tasks, pathologies with low signal, and reference annotations that are uncertain, small, or empty. Limitations of common metrics may result in ineffective machine learning research in designing and optimizing models. To effectively evaluate the clinical value of such models, it is essential to consider factors such as the uncertainty associated with reference annotations, the ability to accurately measure performance regardless of the size of the reference annotation volume, and the classification of cases where reference annotations are empty. We study how uncertain, small, and empty reference annotations influence the value of metrics on a stroke in-house data set regardless of the model. We examine metrics behavior on the predictions of a standard deep learning framework in order to identify suitable metrics in such a setting. We compare our results to the BRATS 2019 and Spinal Cord public data sets. We show how uncertain, small, or empty reference annotations require a rethinking of the evaluation. The evaluation code was released to encourage further analysis of this topic https://github.com/SophieOstmeier/UncertainSmallEmpty.git.

CT perfusion to measure venous outflow in acute ischemic stroke in patients with a large vessel occlusion.

BACKGROUND: Robust venous outflow (VO) profiles, measured by degree of venous opacification on pre-thrombectomy CT angiography (CTA) studies, are strongly correlated with favorable outcomes in patients with large vessel occlusion acute ischemic stroke treated by thrombectomy. However, VO measurements are laborious and require neuroimaging expertise. OBJECTIVE: To develop a semi-automated method to measure VO using CTA and CT perfusion imaging studies. METHODS: We developed a graphical interface using The Visualization Toolkit, allowing for voxel selection at the confluence and bilateral internal cerebral veins on CTA along with arterial input functions (AIFs) from both internal carotid arteries. We extracted concentration-time curves from the CT perfusion study at the corresponding locations associated with AIF and venous output function (VOF). Outcome analyses were primarily conducted by the Mann-Whitney U and Jonckheere-Terpstra tests. RESULTS: Segmentation at the pre-selected AIF and VOF locations was performed on a sample of 97 patients. 65 patients had favorable VO (VO+) and 32 patients had unfavorable VO (VO-). VO+ patients were found to have a significantly shorter VOF time to peak (8.26; 95% CI 7.07 to 10.34) than VO- patients (9.44; 95% CI 8.61 to 10.91), P=0.007. No significant difference was found in VOF curve width and the difference in time between AIF and VOF peaks. CONCLUSIONS: Time to peak of VOF at the confluence of sinuses was significantly associated with manually scored venous outflow. Further studies should aim to understand better the association between arterial inflow and venous outflow, and capture quantitative metrics of venous outflow at other locations.

Exploring I-FABP, endothelin-1 and L-lactate as biomarkers of acute intestinal necrosis: a case-control study.

OBJECTIVE: Acute intestinal necrosis (AIN) is a disease with devastating high mortality. AIN due to obstructed arterial blood flow has a blurred clinical presentation. Timely diagnosis is paramount, and a blood-based biomarker is warranted to increase patient survival. We aimed to assess intestinal fatty acid binding protein (I-FABP) and endothelin-1 as diagnostic biomarkers for AIN. To our knowledge, this is the first study exploring endothelin-1 in AIN patients from a general surgical population. DESIGN: We conducted a single-centre nested case-control study comparing acutely admitted AIN patients to age- and sex-matched non-AIN patients during 2015-2016. I-FABP and endothelin-1 were analysed using an enzyme-linked immunosorbent assay. L-lactate levels were also measured in all patients. Cut-offs were estimated using receiver operator characteristic curves, and the diagnostic performance was estimated using the area under the receiver operator characteristic curve (AUC). RESULTS: We identified 43 AIN patients and included 225 matched control patients. Median levels of I-FABP, endothelin-1 and L-lactate were 3550 (IQR: 1746-9235) pg/ml, 3.91 (IQR: 3.33-5.19) pg/ml and 0.92 (IQR: 0.74-1.45) mM in AIN patients and 1731 (IQR: 1124-2848) pg/ml, 2.94 (IQR: 2.32-3.82) pg/ml and 0.85 (IQR: 0.64-1.21) mM in control patients, respectively. The diagnostic performances of endothelin-1 and of I-FABP + endothelin-1 combined were moderate. Endothelin-1 alone revealed an AUC of 0.74 (0.67; 0.82). The sensitivity and specificity of endothelin-1 were 0.81 and 0.64, respectively. CONCLUSION: I-FABP and endothelin-1 are promising biomarkers for AIN, with moderate diagnostic performance compared with the commonly used biomarker L-lactate. PREREGISTRATION: ClinicalTrials.gov: NCT05665946.

Anatomical predictors of need for decompressive craniectomy after stroke using voxel-based lesion symptom mapping.

BACKGROUND AND PURPOSE: Malignant cerebral edema (MCE) secondary to ischemic stroke is a highly morbid condition. Decompressive craniectomy (DC) is the only treatment for MCE that has been shown to reduce mortality. We examined whether early infarction and/or hypoperfusion in specific topographic regions was predictive of the need for later DC. METHODS: A retrospective database of patients evaluated for large vessel occlusion (LVO) stroke at Stanford between 2010 and 2019 was used. Thirty patients with LVO and baseline perfusion MRI who underwent DC were evaluated. Propensity matching based on age, lesion size, and recanalization status was performed on the remaining cohort. Baseline masks of apparent diffusion coefficient (ADC) + Tmax  >6 seconds lesions were generated using automated perfusion software. Voxel-based lesion symptom maping was used to perform logistic regression at each voxel to generate statistical maps of lesion location associated with DC. Hemispheres were combined to increase statistical power. RESULTS: Sixty patients were analyzed. After adjusting for age, lesion size, and recanalization status as covariates, scattered cortical regions, predominately within the temporal and frontal lobe, were mildly to moderately predictive of the need for DC (z-scores: 2.4-6.74, p < .01). CONCLUSIONS: Scattered temporal and frontal lobe regions on baseline diffusion and perfusion MRI were found to be mildly to moderately predictive of the need for subsequent DC in patients with LVO stroke.