Ensembling crowdsourced seizure prediction algorithms using long-term human intracranial EEG

Seizure prediction is feasible, but greater accuracy is needed to make seizure prediction clinically viable across a large group of patients. Recent work crowdsourced state-of-the-art prediction algorithms in a worldwide competition, yielding improvements in seizure prediction performance for patients whose seizures were previously found hard to anticipate. The aim of the current analysis was to explore potential performance improvements using an ensemble of the top competition algorithms. The results suggest that minor increments in performance may be possible; however, the outcomes of statistical testing limit the confidence in these increments. Our results suggest that for the specific algorithms, evaluation framework, and data considered here, incremental improvements are achievable but there may be upper bounds on machine learning–based seizure prediction performance for some patients whose seizures are challenging to predict. Other more tailored approaches that, for example, take into account a deeper understanding of preictal mechanisms, patient-specific sleep-wake rhythms, or novel measurement approaches, may still offer further gains for these types of patients.     

Transitions across cognitive states and death among older adults in relation to education: A multistate survival model using data from six longitudinal studies

Introduction

This study examines the role of educational attainment, an indicator of cognitive reserve, on transitions in later life between cognitive states (normal Mini-Mental State Examination (MMSE), mild MMSE impairment, and severe MMSE impairment) and death.

Autophagy as a therapeutic target in cardiovascular disease

The epidemic of heart failure continues apace, and development of novel therapies with clinical efficacy has lagged. Now, important insights into the molecular circuitry of cardiovascular autophagy have raised the prospect that this cellular pathway of protein quality control may be a target of clinical relevance. Whereas basal levels of autophagy are required for cell survival, excessive levels - or perhaps distinct forms of autophagic flux - contribute to disease pathogenesis. Our challenge will be to distinguish mechanisms that drive adaptive versus maladaptive autophagy and to manipulate those pathways for therapeutic gain. Recent evidence suggests this may be possible. Here, we review the fundamental biology of autophagy and its role in a variety of forms of cardiovascular disease. We discuss ways in which this evolutionarily conserved catabolic mechanism can be manipulated, discuss studies presently underway in heart disease, and provide our perspective on where this exciting field may lead in the future. This article is part of a special issue entitled ‘‘Key Signaling Molecules in Hypertrophy and Heart Failure.’’     

Hyperspectral imaging: a new approach to the diagnosis of hemorrhagic shock

BACKGROUND: Skin color changes and mottling are frequently described signs of hemorrhagic shock (HEM). Based on this, we developed a noninvasive, noncontact hyperspectral imaging system (HSI), which quantifies and depicts the surface tissue saturation of oxygen (SHSIO2) for each pixel in a region of interest (ROI). Our purpose was to assess HSI in a porcine HEM model. We hypothesized that HEM would cause decreases in SHSIO2 of the skin. METHODS: The HyperMed HSI system employs a spectral separator to vary the wavelength of light admitted to a digital camera. During image acquisition, a "hypercube" of images, each at a separate wavelength, is generated (at 5-nm intervals, from 500 to 600 nm). Then, the visible light spectrum for each pixel in the hypercube is compared by linear regression to standard spectra for oxyhemoglobin (OxyHb) and deoxyhemoglobin (DeoxyHb). The resulting fit coefficients for OxyHb and DeoxyHb are used to calculate SHSIO2 values for each pixel in the ROI. The mean values for OxyHb, DeoxyHb, and SHSIO2 across the ROI are calculated. Grayscale SHSIO2 pictures of the ROI are also generated, in which the brightness of each pixel is proportional to its value. Seventeen pigs, 36.4 +/- 0.11 kg, underwent standard preparation, and were maintained on ketamine and isoflurane. Normothermia was maintained (37 degrees C to 39 degrees C). The hemorrhage group (HEM, n = 9) underwent three blood withdrawals, each 10 mL/kg, with 15 minutes between withdrawals. After the third withdrawal, animals were resuscitated with lactated Ringer's and then shed blood. The control group (CTRL, n = 8) received intravenous fluids at 100 mL/h. HSI images were obtained of the inner hindlimb throughout. RESULTS: All HEM animals showed linear decreases in both mean SHSIO2 and OxyHb values with blood loss, which were reversed by resuscitation. These changes were evident on the grayscale SHSIO2 pictures, but not to the naked eye, and paralleled those of invasively obtained arterial base excess and mixed venous oxygen saturation. CONCLUSIONS: HSI is a promising noninvasive and noncontact tool for quantifying changes in skin oxygenation during HEM and resuscitation.     

Acute respiratory distress syndrome secondary to inhalation of chlorine gas in sheep

BACKGROUND: Toxic industrial chemicals (TICs) are potential terrorist weapons. Several TICs, such as chlorine, act primarily on the respiratory tract, but knowledge of the pathophysiology and treatment of these injuries is inadequate. This study aims to characterize the acute respiratory distress syndrome (ARDS) caused by chlorine gas (Cl2) inhalation in a large-animal model. METHODS: Anesthetized female sheep were ventilated with 300 L of a Cl2/air/oxygen mixture for 30 minutes. In phase 1 (n = 35), doses were 0 ppm (Group 1, n = 6); 120 ppm (Group 2, n = 6); 240 to 350 ppm (Group 3, n = 11); and 400 to 500 ppm (Group 4, n = 12). In phase 2 (n = 17), doses were 0 ppm (Group 5, n = 5); 60 ppm (Group 6, n = 5); and 90 ppm (Group 7, n = 7), and the multiple inert gas elimination technique (MIGET) was used to characterize the etiology of hypoxemia. Computed tomography (CT) scans were performed daily for all animals. RESULTS: In Phase 1, lung function was well maintained in Group 1; Cl2 caused immediate and sustained acute lung injury (PaO2-to-FiO2 ratio, PFR<3.0) in Group 2 and ARDS (PFR<2.0) in Groups 3 and 4. All animals in Groups 1 and 2 survived 96 hours. Kaplan-Meier analysis showed dose-related differences in survival (log-rank test, p < 0.0001). Logistic regression identified 280 ppm as the lethal dose 50%. CT and histopathology demonstrated lesions of both small airways and alveoli. In Phase 2, MIGET showed diversion of blood flow from normal to true-shunt lung compartments and, transiently, to poorly ventilated compartments. CONCLUSIONS: Cl2 causes severe, dose-related lung injury, with features seen in both smoke inhalation and in ARDS secondary to systemic disease. This model will be used to test new therapeutic modalities.     

Radiation-induced leukemia risk among those aged 0-20 at the time of the Chernobyl accident: a case-control study in the Ukraine

A case-control study was conducted to estimate the radiation-induced acute leukemia risk for the period 1987-1997 among residents aged 0-20 at the time of the Chernobyl accident in the most radioactively contaminated territories of the Ukraine (Rivno and Zhytomir regions). Data were collected on 272 leukemia cases diagnosed between 1 January 1987 and 31 December 1997. Of these, 98 cases were verified and interviewed. Verified cases were compared to 151 randomly selected controls matched by age, gender and type of settlement. The mean value of the estimated accumulated equivalent dose to the bone marrow was 4.5 mSv, and the maximum value was 101 mSv. A statistically significant increased risk of leukemia was found among males whose estimated radiation exposure was higher than 10 mSv. This association was statistically significant for acute leukemia cases that occurred in the period 1993-1997, particularly for acute lymphoblastic leukemia. A similar association was found for acute myeloid leukemia, diagnosed in the period 1987-1992.     

Image registration for targeted MRI-guided transperineal prostate biopsy

Purpose:

To develop and evaluate image registration methodology for automated re‐identification of tumor‐suspicious foci from preprocedural MR exams during MR‐guided transperineal prostate core biopsy.

Materials and Methods:

A hierarchical approach for automated registration between planning and intra‐procedural T2‐weighted prostate MRI was developed and evaluated on the images acquired during 10 consecutive MR‐guided biopsies. Registration accuracy was quantified at image‐based landmarks and by evaluating spatial overlap for the manually segmented prostate and sub‐structures. Registration reliability was evaluated by simulating initial mis‐registration and analyzing the convergence behavior. Registration precision was characterized at the planned biopsy targets.

Results:

The total computation time was compatible with a clinical setting, being at most 2 min. Deformable registration led to a significant improvement in spatial overlap of the prostate and peripheral zone contours compared with both rigid and affine registration. Average in‐slice landmark registration error was 1.3 ± 0.5 mm. Experiments simulating initial mis‐registration resulted in an estimated average capture range of 6 mm and an average in‐slice registration precision of ±0.3 mm.

Conclusion:

Our registration approach requires minimum user interaction and is compatible with the time constraints of our interventional clinical workflow. The initial evaluation shows acceptable accuracy, reliability and consistency of the method. J. Magn. Reson. Imaging 2012;36:987–992. © 2012 Wiley Periodicals, Inc.