EEG based automated seizure detection - A survey of medical professionals.

Diagnosing and managing seizures presents substantial challenges for clinicians caring for patients with epilepsy. Although machine learning (ML) has been proposed for automated seizure detection using EEG data, there is little evidence of these technologies being broadly adopted in clinical practice. Moreover, there is a noticeable lack of surveys investigating this topic from the perspective of medical practitioners, which limits the understanding of the obstacles for the development of effective automated seizure detection. Besides the issue of generalisability and replicability seen in a small amount of studies, obstacles to the adoption of automated seizure detection remain largely unknown. To understand the obstacles preventing the application of seizure detection tools in clinical practice, we conducted a survey targeting medical professionals involved in the management of epilepsy. Our study aimed to gather insights on various factors such as the clinical utility, professional sentiment, benchmark requirements, and perceived barriers associated with the use of automated seizure detection tools. Our key findings are: I) The minimum acceptable sensitivity reported by most of our respondents (80%) seems achievable based on studies reported from most currently available ML-based EEG seizure detection algorithms, but replication studies often fail to meet this minimum. II) Respondents are receptive to the adoption of ML seizure detection tools and willing to spend time in training. III) The top three barriers for usage of such tools in clinical practice are related to availability, lack of training, and the blackbox nature of ML algorithms. Based on our findings, we developed a guide that can serve as a basis for developing ML-based seizure detection tools that meet the requirements of medical professionals, and foster the integration of these tools into clinical practice.

Nickel oxide morphology synthesized with a hydrothermal method for inverted perovskite solar cells.

In this paper, a hydrothermal method is used to synthesize a nickel oxide nanostructure (nano-NiO) for its application to inverted perovskite solar cells. These pore nanostructures were employed to increase both the contact and channel between the hole transport and perovskite layers of an ITO/nano-N i O/C H 3 N H 3 P b I 3/P C B M/A g device. The purpose of this research is twofold. First, three different nano-NiO morphologies were synthesized at temperatures of 140°C, 160°C, and 180°C. Then, a Raman spectrometer was used to check the phonon vibration and magnon scattering characteristics after an annealing temperature of 500°C. Second, nano-NiO powders were dispersed in isopropanol for subsequent spin coating on the inverted solar cells. The nano-NiO morphologies were multi-layer flakes, microspheres, and particles at synthesis temperatures of 140°C, 160°C, and 180°C, respectively. When the microsphere nano-NiO was used as the hole transport layer, the perovskite layer had a larger coverage of 83.9%. The grain size of the perovskite layer was analyzed by x-ray diffraction, and strong crystal orientations of (110) and (220) peaks were found. Despite this, the power conversion efficiency could affect the promotion, which is 1.37 times higher than the poly(3,4-ethylenedioxythiophene) polystyrene sulfonate element conversion efficiency of the planar structure.

Estimating Patient-Level Uncertainty in Seizure Detection Using Group-Specific Out-of-Distribution Detection Technique.

Epilepsy is a chronic neurological disorder affecting around 1% of the global population, characterized by recurrent epileptic seizures. Accurate diagnosis and treatment are crucial for reducing mortality rates. Recent advancements in machine learning (ML) algorithms have shown potential in aiding clinicians with seizure detection in electroencephalography (EEG) data. However, these algorithms face significant challenges due to the patient-specific variability in seizure patterns and the limited availability of high-quality EEG data for training, causing erratic predictions. These erratic predictions are harmful, especially for high-stake domains in healthcare, negatively affecting patients. Therefore, ensuring safety in AI is of the utmost importance. In this study, we propose a novel ensemble method for uncertainty quantification to identify patients with low-confidence predictions in ML-based seizure detection algorithms. Our approach aims to mitigate high-risk predictions in previously unseen seizure patients, thereby enhancing the robustness of existing seizure detection algorithms. Additionally, our method can be implemented with most of the deep learning (DL) models. We evaluated the proposed method against established uncertainty detection techniques, demonstrating its effectiveness in identifying patients for whom the model's predictions are less certain. Our proposed method managed to achieve 87%, 89% and 75% in accuracy, specificity and sensitivity, respectively. This study represents a novel attempt to improve the reliability and robustness of DL algorithms in the domain of seizure detection. This study underscores the value of integrating uncertainty quantification into ML algorithms for seizure detection, offering clinicians a practical tool to gauge the applicability of ML models for individual patients.

Preference and usage pattern of mobile medical apps for drug information purposes among hospital pharmacists in Sarawak, Malaysia.

INTRODUCTION: Pharmacists are frequent users of mobile medical apps (MMA) for drug information (DI) and clinical decision-making purposes. However, the wide range of available MMA may be of variable credibility and results in heterogeneous recommendations. The need for subscription may also influence choice of apps. OBJECTIVE: The objective of this study was to determine the usage pattern of MMA among hospital pharmacists, including their perceptions and factors affecting their choice of apps. METHODS: This cross-sectional study required respondents to fill in an online questionnaire. The questionnaire included sections on respondents' demographic data, MMA usage pattern, perceived usefulness and opinion on subscription fees. Items were adapted from available literature and validated locally. It was made accessible for 6 weeks starting November 2019 for all pharmacists working in the 23 public hospitals in Sarawak to response (universal sampling). Collected data were analysed using descriptive and inferential statistics. RESULTS: A response rate of 37.2% was achieved (n = 162). Respondents were heavily reliant on MMA, with 78.4% accessing them multiple times daily. The majority also agreed that MMA contain correct and up-to-date information. A median of 5 apps were downloaded, suggesting an ultimate app catering for all DI needs was lacking. The Malaysian Drug Formulary was the most downloaded app (88.3%), whereas Lexicomp® was the most "well-rounded" in terms of functionality. Clinical pharmacists were significantly more likely to purchase MMA, in particular UpToDate® (p < 0.01) due to their need to access clinical updates. Respondents highly recommended institutional access for either UpToDate® or Lexicomp® be made available. Pre-registration pharmacists should be guided on judicious MMA usage, as they downloaded significantly more apps and were more likely to indicate not knowing which DI recommendation to follow (both p < 0.01). CONCLUSION: MMA has become an indispensable tool for hospital pharmacists, however there was a tendency to download multiple apps for DI needs. Institutional access can be considered for credible apps identified to ensure accuracy and uniformity of DI recommendations, with purchase decision made after surveying the needs and preferences of end users.

Full-Privacy Secured Search Engine Empowered by Efficient Genome-Mapping Algorithms.

Since the 90s, keyword-based search engines have been the only option for people to locate relevant web content through a simple query comprising one to a few keywords. These engines, whether free or paid, retained users' search queries and preferences, often to deliver targeted ads. Additionally, user-uploaded articles for plagiarism detection can further be stored as part of service providers' expanding databases for profit. Essentially, users could not search without exposing their queries to these providers. We present a new solution here: a method for searching the internet using a full article as a query without disclosing the content. Our Sapiens Aperio Veritas Engine (S.A.V.E.) uses an encoding scheme and an FM-index search, borrowed from next-generation human genome sequencing. Each word in a user's query is transformed into one of 12 "amino acids" to create a pseudo-biological sequence (PBS) on the user's device. Plagiarism checks are done by users submitting their locally created PBSs to our cloud service. This detects identical content in our database, which includes all English and Chinese Wikipedia articles and Open Access journals up to April 2021. PBSs, longer than 12 "amino acids", show accurate results with less than 0.8% false positives. Performance-wise, S.A.V.E. runs at a similar genome-mapping speed as Bowtie and is >5 orders faster than BLAST. With both standard and private modes, S.A.V.E. offers a revolutionary, privacy-first search and plagiarism check system. We believe this sets an exciting precedent for future search engines prioritizing user confidentiality. S.A.V.E. can be accessed at https://dyn.life.nthu.edu.tw/SAVE/.

EEG datasets for seizure detection and prediction- A review.

Electroencephalogram (EEG) datasets from epilepsy patients have been used to develop seizure detection and prediction algorithms using machine learning (ML) techniques with the aim of implementing the learned model in a device. However, the format and structure of publicly available datasets are different from each other, and there is a lack of guidelines on the use of these datasets. This impacts the generatability, generalizability, and reproducibility of the results and findings produced by the studies. In this narrative review, we compiled and compared the different characteristics of the publicly available EEG datasets that are commonly used to develop seizure detection and prediction algorithms. We investigated the advantages and limitations of the characteristics of the EEG datasets. Based on our study, we identified 17 characteristics that make the EEG datasets unique from each other. We also briefly looked into how certain characteristics of the publicly available datasets affect the performance and outcome of a study, as well as the influences it has on the choice of ML techniques and preprocessing steps required to develop seizure detection and prediction algorithms. In conclusion, this study provides a guideline on the choice of publicly available EEG datasets to both clinicians and scientists working to develop a reproducible, generalizable, and effective seizure detection and prediction algorithm.

MiR34 contributes to spinal muscular atrophy and AAV9-mediated delivery of MiR34a ameliorates the motor deficits in SMA mice.

Spinal muscular atrophy (SMA) is a neurodegenerative disease characterized by the selective loss of spinal motor neurons (MNs) and concomitant muscle weakness. Mutation of SMN1 is known to cause SMA, and restoring SMN protein levels via antisense oligonucleotide treatment is effective for ameliorating symptoms. However, this approach is hindered by exorbitant costs, invasive procedures, and poor treatment responses of some patients. Here, we seek to circumvent these hurdles by identifying reliable biomarkers that could predict treatment efficacy. We uncovered that MiR34 exhibits consistent downregulation during SMA progression in both human and rodent contexts. Importantly, Mir34 family-knockout mice display axon swelling and reduced neuromuscular junction (NMJ) endplates, recapitulating SMA pathology. Introducing MiR34a via scAAV9 improved the motor ability of SMNΔ7 mice, possibly by restoring NMJ endplate size. Finally, we observed a consistent decreasing trend in MiR34 family expression in the cerebrospinal fluid (CSF) of type I SMA patients during the loading phase of nusinersen treatment. Baseline CSF MiR34 levels before nusinersen injection proved predictive of patient motor skills 1 year later. Thus, we propose that MiR34 may serve as a biomarker of SMA since it is associated with the pathology and can help evaluate the therapeutic effects of nusinersen.

Evolution of Initial Pharmacologic Treatment of Newly Diagnosed Parkinson's Disease Patients over a Decade in Singapore.

OBJECTIVE: The aim of this study is to compare Parkinson's disease (PD) treatment practices by movement disorder (MD) specialists across a decade, and to determine the factors that influence drug choice for the motor symptoms of PD in newly diagnosed drug-naïve patients. METHODS: This prospective temporal analysis included patients seen at the National Neuroscience Institute in Singapore and diagnosed with PD by MD specialists in the years 2007 and 2017. Primary outcomes were use of specific PD drugs and changes in drug-prescribing patterns. Descriptive analyses and multivariable logistic regression models determined the extent to which patient characteristics were associated with type of PD treatment. RESULTS: Of 230 patients with PD (mean (SD) age, 66.7 (10.3) years), 131 (57.0%) were male. From 2007 to 2017, the use of ergot dopamine agonists and anticholinergics decreased from 19.3% to 2.0% (P < 0.001) and from 12.0% to 2.7% (P = 0.004), respectively. The use of monoamine oxidase B inhibitors (MAOBI) increased from 13.3% to 25.2% (P = 0.033). The use of levodopa (LD)-sparing strategies decreased nonsignificantly from 33.7% to 24.5% (P = 0.133). Overall, 196 (85.2%) patients were initiated on symptomatic monotherapy, with LD being the most commonly prescribed. MAOBI was the most common drug used in combination therapy. Age ≤70 (adjusted OR, 11.9; 95% CI, 4.5-31.5) and Hoehn and Yahr (HY) stage <2 (adjusted OR, 3.4; 95% CI, 1.5-7.7) were independent factors for LD-sparing strategies. Non-LD prescriptions (13 of 92; 14.1%) were more likely to be discontinued compared to LD ones (6 of 149; 4.0%) (P = 0.005). CONCLUSIONS: Drug-prescribing patterns in PD have changed significantly through the last decade, influenced by emerging evidence and reports of adverse drug effects. Choosing drugs based on the patient's age and disease severity remain sound guiding principles across the years. It is important that international and national guidelines for pharmacotherapy in PD be updated consistently throughout different socioeconomic settings to optimize care.

A gate-free monolayer WSe2 pn diode.

Interest in bringing p- and n-type monolayer semiconducting transition metal dichalcogenides (TMD) into contact to form rectifying pn diode has thrived since it is crucial to control the electrical properties in two-dimensional (2D) electronic and optoelectronic devices. Usually this involves vertically stacking different TMDs with pn heterojunction or, laterally manipulating carrier density by gate biasing. Here, by utilizing a locally reversed ferroelectric polarization, we laterally manipulate the carrier density and created a WSe2 pn homojunction on the supporting ferroelectric BiFeO3 substrate. This non-volatile WSe2 pn homojunction is demonstrated with optical and scanning probe methods and scanning photoelectron micro-spectroscopy. A homo-interface is a direct manifestation of our WSe2 pn diode, which can be quantitatively understood as a clear rectifying behavior. The non-volatile confinement of carriers and associated gate-free pn homojunction can be an addition to the 2D electron-photon toolbox and pave the way to develop laterally 2D electronics and photonics.

Spatio-temporal skin strain distributions evoke low variability spike responses in cuneate neurons.

A common method to explore the somatosensory function of the brain is to relate skin stimuli to neurophysiological recordings. However, interaction with the skin involves complex mechanical effects. Variability in mechanically induced spike responses is likely to be due in part to mechanical variability of the transformation of stimuli into spiking patterns in the primary sensors located in the skin. This source of variability greatly hampers detailed investigations of the response of the brain to different types of mechanical stimuli. A novel stimulation technique designed to minimize the uncertainty in the strain distributions induced in the skin was applied to evoke responses in single neurons in the cat. We show that exposure to specific spatio-temporal stimuli induced highly reproducible spike responses in the cells of the cuneate nucleus, which represents the first stage of integration of peripheral inputs to the brain. Using precisely controlled spatio-temporal stimuli, we also show that cuneate neurons, as a whole, were selectively sensitive to the spatial and to the temporal aspects of the stimuli. We conclude that the present skin stimulation technique based on localized differential tractions greatly reduces response variability that is exogenous to the information processing of the brain and hence paves the way for substantially more detailed investigations of the brain's somatosensory system.