ABSTRACT
Stroke is a major threat to life and health in modern society, especially in the aging population. Stroke may cause sudden death or severe sequela-like hemiplegia. Although computed tomography (CT) and magnetic resonance imaging (MRI) are standard diagnosis methods, and artificial intelligence models have been built based on these images, shortage in medical resources and the time and cost of CT/MRI imaging hamper fast detection, thus increasing the severity of stroke. Here, we developed a convolutional neural network model by integrating four networks, Xception, ResNet50, VGG19, and EfficientNetb1, to recognize stroke based on 2D facial images with a cross-validation area under curve (AUC) of 0.91 within the training set of 185 acute ischemic stroke patients and 551 age- and sex-matched controls, and AUC of 0.82 in an independent data set regardless of age and sex. The model computed stroke probability was quantitatively associated with facial features, various clinical parameters of blood clotting indicators and leukocyte counts, and, more importantly, stroke incidence in the near future. Our real-time facial image artificial intelligence model can be used to rapidly screen and prediagnose stroke before CT scanning, thus meeting the urgent need in emergency clinics, potentially translatable to routine monitoring.
ABSTRACT
Background and purpose: This study sought to improve methods to identify biomarkers in the neuroendocrine system related to stroke progression to improve the accuracy of traditional tools for evaluating stroke prognosis. Methods: Seventy-four stroke patients and 237 healthy controls were prospectively included. We measured urinary epinephrine (E), noradrenaline (NE), dopamine (DA) and cortisol (F) on days 1, 3, and 5 after stroke onset and plasma F, adrenocorticotropic hormone (ACTH), thyrotropin (TSH), prolactin (PRL), follicle-stimulating hormone (FSH), luteinizing hormone (LH) and growth hormone (GH). The correlation between these hormone levels and 90-day prognosis was analyzed, their value in assessing prognosis was compared with lesion volume and National Institutes of Health Stroke Scale (NIHSS) scores using receiver operating characteristic (ROC) curves, and their correlation with conventional clinical variables was assessed. Results: Levels of F, 24-h urinary free cortisol(UFC), E, NE, DA, and GH on days 1, 3, and 5 were significantly higher in stroke patients than in controls (P < 0.01), while ACTH and TSH decreased, gradually approaching normal within 5 days of onset. Levels of E, NE, F, and 24-h UFC were proportional to severity, and all gradually decreased within 5 days of onset in patients with a good prognosis and gradually increased or remained high in those with a poor prognosis. After adjustment for age, sex, NIHSS, or Glasgow Coma Scale (GCS) score, F > 13.6 µg/dL, ACTH > 22.02 pg/mL and NE > 123.5 µg/ 24 h were identified as risk factors for a poor prognosis 90 days after stroke (P < 0.05). The combination of F, ACTH, NE, white blood cell count (WBC), glucose (Glu), and hemoglobin (Hb) was significantly more accurate than lesion volume (AUC: 0.931 vs. 0.694 P = 0.019) and NIHSS score (AUC: 0.931 vs. 0.746 P = 0.034) in predicting poor prognosis of stroke 1 day after onset. Hormones and traditional clinical variables were correlated to varying degrees, with NE correlating most strongly with 24-h UFC (r = 0.54) and moderately positively with lesion volume (r = 0.40) and NIHSS score (r = 0.45). Conclusions: Stroke causes significant time-phased dynamic changes in the hypothalamic-pituitary-adrenal axis and sympathetic nervous system, and plasma F, ACTH, and urinary NE levels can be used to assess stroke severity and prognosis. Chinese clinical trial registry: Registration Number: ChiCTR1900024992. Registration Date: 2019/8/6.
ABSTRACT
BACKGROUND: Aortic dissection (AD) and pulmonary embolism (PE) are both life-threatening disorders. Because of their conflicting treatments, treatment becomes difficult when they occur together, and there is no standard treatment protocol. CASE SUMMARY: A 67-year-old man fell down the stairs due to syncope and was brought to our hospital as a confused and irritable patient who was uncooperative during the physical examination. Further examination of the head, chest and abdomen by computed tomography revealed a subdural hemorrhage, multiple rib fractures, a hemopneumothorax and a renal hematoma. He was admitted to the Emergency Intensive Care Unit and given a combination of oxygen therapy, external rib fixation, analgesia and enteral nutrition. The patient regained consciousness after 2 wk but complained of abdominal pain and dyspnea with an arterial partial pressure of oxygen of 8.66 kPa. Computed tomography angiograms confirmed that he had both AD and PE. We subsequently performed only nonsurgical treatment, including nasal high-flow oxygen therapy, nonsteroidal analgesia, amlodipine for blood pressure control, beta-blockers for heart rate control. Eight weeks after admission, the patient improved and was discharged from the hospital. CONCLUSION: Patients with AD should be alerted to the possibility of a combined PE, the development of which may be associated with aortic compression. In patients with type B AD combined with low-risk PE, a nonsurgical, nonanticoagulant treatment regimen may be feasible.
ABSTRACT
Objective: To investigate whether D-dimer levels and changes in D-dimer levels can be used as effective indexes to evaluate the risk of death among intraparenchymal haemorrhage (IPH) patients. Methods: A retrospective cohort study of 732 patients with IPH was conducted at an academic medical centre. The risk factors for adverse hospitalization outcomes were analysed, and logistic multivariate analysis was performed. Patients were divided into supratentorial, brainstem and cerebellum groups. According to whether intraventricular haematoma (IVH) and subarachnoid haematoma (SAH) co-occurred, the supratentorial group was divided into simple haematoma, combined IVH, combined SAH and combined IVH+SAH subgroups. The relationship between D-dimer levels and hospitalization outcome in each group/ subgroup was analysed. Results: Compared with survivors, the plasma D-dimer level of the nonsurvivors on the second day after admission was significantly higher (2.52 ± 3.89 µg/mL vs 0.77 ± 2.31 µg/mL, P = 0.032), and the difference in plasma D-dimer levels between the second day after admission and admission significantly increased (1.77 ± 3.70 µg/mL vs 0.26 ± 2.80 µg/mL, P = 0.049), and a D-dimer level on day 2 > 0.58 µg/mL was an independent risk factor for mortality among IPH patients (OR 3.114, 95% CI: 1.007). In the supratentorial group and the IVH subgroup, the level of D-dimer on day 2 was significantly higher among nonsurvivors than among survivors (2.18 ± 2.13 µg/mL vs 0.65 ± 1.04 µg/mL, P = 0.011; 2.45 ± 2.31 µg/mL vs 0.91 ± 1.26 µg/mL, P = 0.028, respectively). Conclusion: The increase in plasma D-dimer levels on day 2 is related to poor hospitalization outcomes of patients with IPH, and this correlation may exist only among patients with both IVH and IPH.
