Artificial intelligence/machine learning for epilepsy and seizure diagnosis.

Accurate seizure and epilepsy diagnosis remains a challenging task due to the complexity and variability of manifestations, which can lead to delayed or missed diagnosis. Machine learning (ML) and artificial intelligence (AI) is a rapidly developing field, with growing interest in integrating and applying these tools to aid clinicians facing diagnostic uncertainties. ML algorithms, particularly deep neural networks, are increasingly employed in interpreting electroencephalograms (EEG), neuroimaging, wearable data, and seizure videos. This review discusses the development and testing phases of AI/ML tools, emphasizing the importance of generalizability and interpretability in medical applications, and highlights recent publications that demonstrate the current and potential utility of AI to aid clinicians in diagnosing epilepsy. Current barriers of AI integration in patient care include dataset availability and heterogeneity, which limit studies' quality, interpretability, comparability, and generalizability. ML and AI offer substantial promise in improving the accuracy and efficiency of epilepsy diagnosis. The growing availability of diverse datasets, enhanced processing speed, and ongoing efforts to standardize reporting contribute to the evolving landscape of AI applications in clinical care.

Methods and Impact for Using Federated Learning to Collaborate on Clinical Research.

BACKGROUND: The development of accurate machine learning algorithms requires sufficient quantities of diverse data. This poses a challenge in health care because of the sensitive and siloed nature of biomedical information. Decentralized algorithms through federated learning (FL) avoid data aggregation by instead distributing algorithms to the data before centrally updating one global model. OBJECTIVE: To establish a multicenter collaboration and assess the feasibility of using FL to train machine learning models for intracranial hemorrhage (ICH) detection without sharing data between sites. METHODS: Five neurosurgery departments across the United States collaborated to establish a federated network and train a convolutional neural network to detect ICH on computed tomography scans. The global FL model was benchmarked against a standard, centrally trained model using a held-out data set and was compared against locally trained models using site data. RESULTS: A federated network of practicing neurosurgeon scientists was successfully initiated to train a model for predicting ICH. The FL model achieved an area under the ROC curve of 0.9487 (95% CI 0.9471-0.9503) when predicting all subtypes of ICH compared with a benchmark (non-FL) area under the ROC curve of 0.9753 (95% CI 0.9742-0.9764), although performance varied by subtype. The FL model consistently achieved top three performance when validated on any site's data, suggesting improved generalizability. A qualitative survey described the experience of participants in the federated network. CONCLUSION: This study demonstrates the feasibility of implementing a federated network for multi-institutional collaboration among clinicians and using FL to conduct machine learning research, thereby opening a new paradigm for neurosurgical collaboration.

Maternal mental health and substance use disorders in sudden unexpected death in infancy using routinely collected health data in New Zealand, 2000-2016.

BACKGROUND: Mortality from sudden unexpected death in infancy (SUDI) has declined dramatically since the 'Back to Sleep' campaign. Deaths now are more prevalent in those with socioeconomic disadvantage. The investigation of SUDI frequently identifies parents that have mental health or drug, alcohol and addiction problems. AIMS: To estimate the prevalence of maternal mental health and substance use disorders and assess the magnitude of their risk for SUDI. METHODS: We conducted a population-based cohort study using data from the Integrated Data Infrastructure (IDI), a large research database containing linked data from a range of government agencies. The study population was all live births and their mothers in New Zealand from 2000 to 2016. The exposures of interest were maternal mental health problems and maternal substance use disorders in the year prior to the birth. The outcome was deaths from SUDI. RESULTS: The total population was 1086 504 live births and of these 1078 811 (99.3%) were able to be linked to other data sets within the IDI. The prevalence of maternal mental health problems in the total population was 5.2% and substance use disorder was 0.7%. There were 42 deaths from SUDI (0.75/1000) that were exposed to maternal mental illness and 864 deaths (0.84/1000) that were not exposed (adjusted relative risk (aRR)=1.23, 95% CI 0.90 to 1.68). There were 21 deaths from SUDI (2.67/1000) that were exposed to maternal substance use disorders and 885 (0.83/1000) that were not exposed (aRR=1.82, 95% CI 1.17 to 2.83). CONCLUSIONS: Maternal substance use disorders, but not maternal mental health problems, in the year prior to the child's birth was associated with an increased risk of SUDI. However, the numbers that are affected are small and the effect size moderate. This group of women should receive additional SUDI prevention services and Safe Sleep advice.

START: a system for flexible analysis of hundreds of genomic signal tracks in few lines of SQL-like queries.

BACKGROUND: A genomic signal track is a set of genomic intervals associated with values of various types, such as measurements from high-throughput experiments. Analysis of signal tracks requires complex computational methods, which often make the analysts focus too much on the detailed computational steps rather than on their biological questions. RESULTS: Here we propose Signal Track Query Language (STQL) for simple analysis of signal tracks. It is a Structured Query Language (SQL)-like declarative language, which means one only specifies what computations need to be done but not how these computations are to be carried out. STQL provides a rich set of constructs for manipulating genomic intervals and their values. To run STQL queries, we have developed the Signal Track Analytical Research Tool (START, http://yiplab.cse.cuhk.edu.hk/start/ ), a system that includes a Web-based user interface and a back-end execution system. The user interface helps users select data from our database of around 10,000 commonly-used public signal tracks, manage their own tracks, and construct, store and share STQL queries. The back-end system automatically translates STQL queries into optimized low-level programs and runs them on a computer cluster in parallel. We use STQL to perform 14 representative analytical tasks. By repeating these analyses using bedtools, Galaxy and custom Python scripts, we show that the STQL solution is usually the simplest, and the parallel execution achieves significant speed-up with large data files. Finally, we describe how a biologist with minimal formal training in computer programming self-learned STQL to analyze DNA methylation data we produced from 60 pairs of hepatocellular carcinoma (HCC) samples. CONCLUSIONS: Overall, STQL and START provide a generic way for analyzing a large number of genomic signal tracks in parallel easily.

Chronic Hepatitis C Treatment in Patients with Drug Injection History: Findings of the INTEGRATE Prospective, Observational Study.

INTRODUCTION: People who inject drugs represent an under-treated chronic hepatitis C virus (HCV)-infected patient population. METHODS: INTEGRATE was a prospective, observational study investigating the effectiveness, safety, and adherence in routine clinical practice to telaprevir in combination with peg-interferon and ribavirin (Peg-IFN/RBV) in patients with history of injecting drug use chronically infected with genotype 1 HCV. RESULTS: A total of 46 patients were enrolled and included in the intent-to-treat (ITT) population. Among heroin and/or cocaine users (n = 37; 80%), 22% reported use in the past month; 74% (34/46) of patients were on opioid substitution therapy in the pre-treatment phase, and 43% (20/46) discontinued HCV treatment prematurely. Sustained virologic response rate was 54% (25/46) in the ITT population and 74% (25/34) in the per protocol (evaluable-for-effectiveness) population. The main reason for failure in the ITT analysis was loss to follow-up (n = 8; 17%). Adverse events occurred in 91% (42/46) of patients. Mean patient-reported adherence to study drugs was >89% at Week 4, Week 12 and end of treatment. CONCLUSION: Despite a high rate of treatment discontinuation (including loss to follow-up), self-reported adherence to treatment was good and virologic cure rates were similar to those reported in large real-world cohorts. Our findings suggest that people with a history of injecting drug use should be considered for treatment of chronic HCV infection, and highlight the need for improvements in patient support to boost retention in care and, in turn, help to prevent reinfection and transmission. CLINICAL TRIAL REGISTRATION: Clinicaltrials.gov identifier, NCT01980290. FUNDING: Janssen Pharmaceuticals.

Treating Older HIV-1-infected Subjects With Cobicistat-boosted Darunavir in a 48-week Phase 3 Trial.

BACKGROUND: An aging HIV-1-infected population warrants examination of the acceptability of individual antiretroviral regimens. In a previous study of ritonavir-boosted darunavir (ARTEMIS), similar safety/efficacy profiles were observed in younger (≤45 years) and older (>45 years) HIV-1-infected subjects. OBJECTIVE: To evaluate safety and efficacy outcomes in HIV-1-infected younger versus older subjects treated with cobicistat-boosted darunavir. METHOD: In a 48-week, phase 3b, open-label trial, HIV-1-infected adults were administered darunavir 800 mg and cobicistat 150 mg once-daily with 2 nucleos(t)ide reverse transcriptase inhibitors (N[t]RTIs). Post hoc analyses examined safety and efficacy outcomes in subjects ≤45 and >45 years. RESULTS: Of 313 subjects, 76% were ≤45 years (median [range] age, 31 [18-45]) and 24% were >45 years (49 [46-72]). Baseline median (range) viral loads were 4.75 (2.6-6.8) and 4.83 (2.7-7.0) log10 copies/mL, and CD4+ counts were 379.0 (5-1473) and 310.5 (6-757) cells/mm3 in younger and older subjects, respectively. Through Week 48, similar proportions of younger and older subjects had ≥1 adverse event (AE; 93% vs 88%), ≥1 grade 2-4 AE possibly related to study drug (13% vs 15%), and discontinued study due to AE (3% vs 3%). At Week 48, 82% of younger and 78% of older subjects had viral load <50 copies/mL (95% CI of the difference: -7.4% to 13.8%). A higher proportion of older versus younger subjects took >4 concomitant medications during the study (69% vs 57%). CONCLUSION: Safety and efficacy profiles of cobicistat-boosted darunavir with 2 N(t)RTIs were similar in HIV-1-infected subjects ≤45 and >45 years.

Telaprevir-based therapy for treatment of HIV-1 and hepatitis C virus co-infected patients: An early access programme.

OBJECTIVES: HPC3005 is a multicentre, open-label, telaprevir trial in HCV/HIV coinfected patients with severe fibrosis or compensated cirrhosis. METHODS: Patients were treated with telaprevir 750 mg every 8 h (1125 mg if on efavirenz) plus pegylated interferon-alpha (PEG-IFN, 180 µg once-weekly) and ribavirin (RBV, 800 mg/day) for 12 weeks, followed by 36 weeks of PEG-IFN/RBV. RESULTS: Mean age was 44 years, 97/118 patients were male and all were Caucasian, 68 had severe fibrosis and 50 had cirrhosis. Seventy-eight had HCV RNA levels ≥800 000 IU/mL, 72 had HCV genotype 1a, baseline HIV RNA was <50 copies/mL in 112 patients. Overall, 114/118 patients continued antiretroviral treatment, 4 were untreated. Seventy-five patients received tenofovir and 74 emtricitabine; in addition 53 received atazanavir/ritonavir, 43 raltegravir, and 24 efavirenz. By intention-to-treat, 78 (66%) patients achieved SVR24. Nineteen discontinued telaprevir, 8 for virological endpoint, 5 for adverse events (2 anaemia, 2 rash, 1 asthenia), 5 for non-compliance and 1 withdrew consent. The most common adverse events were anaemia (36 patients), thrombocytopaenia (33), rash (26), bilirubin increase (17), and neutropenia (16). CONCLUSIONS: In this early access programme in coinfected patients with severe fibrosis or cirrhosis, 66% of patients achieved SVR. The most common adverse events were haematological. CLINICAL TRIAL NUMBER: NCT01500616.

Aging and masculinity: portrayals in men's magazines.

Textual and visual representations of age are instructive as they suggest ideals towards which individuals should strive and influence how we perceive age. The purpose of our study was to investigate textual and visual representations of later life in the advertisements and interest stories of six widely read North American male-oriented magazines (namely, Esquire, GQ, Maxim, Men's Health, Men's Journal, and Zoomer). Through a content analysis and a visual textual analysis, we examined how older men were depicted in the magazine images and accompanying texts. Our findings revealed that older men were largely absent, and when portrayed, were positively depicted as experienced and powerful celebrities or as healthy and happy unknown individuals. The magazine advertisements and interest stories collectively required individuals to engage in consumer culture in order to achieve age and masculinity ideals and stave off the transition from the Third Age to the Fourth Age. We consider our findings in relation to theorizing about ageism, age relations, the Third and Fourth Ages, and idealized aging masculinity.

Memory failures for everyday tasks in caregivers of children with autism.

The stress of caring for a loved one with chronic illness has been linked with impairments in cognitive processes such as attention and problem solving, though few studies have examined the impact on memory. Compromised cognition, in particular, aspects of everyday functioning such as remembering medical instructions and appointments, might affect caregivers' ability to maintain the consistency and quality of care needed by the child. A sample of 31 caregivers of children with autism and 51 parents of neuro-typical children completed an electronic survey assessing their levels of psychological distress and everyday memory. Perceived stress scores were higher in the caregiver group, as were self-reported memory failures for everyday tasks. The negative impact of caregiver stress on everyday memory was particularly salient among caregivers experiencing higher perceived levels of stress. These findings have implications for interventions that aim to improve caregivers' cognitive well being through targeting the psychological sequelae associated with the caregiving experience.

Electrical or photocontrol of the rotary motion of a metallacarborane.

Rotary motion around a molecular axis has been controlled by simple electron transfer processes and by photoexcitation. The basis of the motion is intramolecular rotation of a carborane cage ligand (7,8-dicarbollide) around a nickel axle. The Ni(III) metallacarborane structure is a transoid sandwich with two pairs of carbon vertices reflected through a center of symmetry, but that of the Ni(IV) species is cisoid. The interconversion of the two provides the basis for controlled, rotational, oscillatory motion. The energies of the Ni(III) and Ni(IV) species are calculated as a function of the rotation angle.

Molecular recognition and conductance in crown ethers.

Crown ethers have the remarkable property of recognizing and binding specific metal cations in complex mixtures. We propose to combine molecular recognition with molecular electric conductance. The question we address is: can the event of binding a cation be sensed by a change in conductance? Specifically, we study a short molecular wire (MW) containing a crown-6 molecule connected via sulfur atoms to two gold atomic wires acting as metallic leads. Upon binding a cation, the density of states of the system is only slightly affected. This reflects the fact that the cation binding is largely electrostatic in nature and is accompanied by little electronic reorganization. Yet, the cationic binding does significantly lower conductance. We also identify strong interference affecting the conductance. A striking feature is the insensitivity of conductance to the type of ligand with the exception of the proton.