Intracranial hemorrhage is a medical emergency that requires urgent diagnosis and immediate treatment to improve patient outcome. Machine learning algorithms can be used to perform medical image classification and assist clinicians in diagnosing radiological scans. In this paper, we apply 3-dimensional convolutional neural networks (3D CNN) to classify computed tomography (CT) brain scans into normal scans (N) and abnormal scans containing subarachnoid hemorrhage (SAH), intraparenchymal hemorrhage (IPH), acute subdural hemorrhage (ASDH) and brain polytrauma hemorrhage (BPH). The dataset used consists of 399 volumetric CT brain images representing approximately 12,000 images from the National Neuroscience Institute, Singapore. We used a 3D CNN to perform both 2-class (normal versus a specific abnormal class) and 4-class classification (between normal, SAH, IPH, ASDH). We apply image thresholding at the image pre-processing step, that improves 3D CNN classification accuracy and performance by accentuating the pixel intensities that contribute most to feature discrimination. For 2-class classification, the F1 scores for various pairs of medical diagnoses ranged from 0.706 to 0.902 without thresholding. With thresholding implemented, the F1 scores improved and ranged from 0.919 to 0.952. Our results are comparable to, and in some cases, exceed the results published in other work applying 3D CNN to CT or magnetic resonance imaging (MRI) brain scan classification. This work represents a direct application of a 3D CNN to a real hospital scenario involving a medically emergent CT brain diagnosis.

Interprofessional collaboration (IPC) is key to ensuring safe quality care for patients. However, IPC intervention outcomes are variable, leading to calls for systems theories to understand complex interactions in healthcare. Using networked ecological systems theory (NEST), we aimed to uncover facilitators and barriers impacting the interactions between nurses and physicians in a specialty healthcare center. A qualitative study involving 55 non-participant observations and 17 individual semi-structured interviews was conducted at the National Neuroscience Institute of Singapore from April 2019 to March 2021. Template analysis was used to analyze the data. The most important IPC facilitators were exosystemic institutional support and physicians’ willingness to engage in IPC in the microsystems that together enabled the establishment of disease-based outpatient programs fostering patient-centered interactions among different healthcare professionals (HCP). We also found that patient-, disease-, and systems-related knowledge played an important role in facilitating IPC. Macrosystemic entrenchments such as intraprofessional composition of ward rounds emerged as a significant barrier. However, microsystemic efforts such as chat groups connecting all HCP involved in the care of the patients in the wards have fostered IPC. Although still preliminary, these findings suggest NEST can be useful to inform systematic interventions to improve IPC.

Despite decades of research on the impact of interprofessional collaboration (IPC), we still lack definitive proof that team-based care can lead to a tangible effect on healthcare outcomes. Without return on investment (ROI) evidence, healthcare leaders cannot justifiably throw their weight behind IPC, and the institutional push for healthcare manpower reforms crucial for facilitating IPC will remain variable and fragmentary. The lack of proof for the ROI of IPC is likely due to a lack of a unifying conceptual framework and the over-reliance on the single-method study design. To address the gaps, this paper describes a protocol which uses as a framework the Quadruple Aim which examines the ROI of IPC using four dimensions: patient outcomes, patient experience, provider well-being, and cost of care. A multimethod approach is proposed whereby patient outcomes are measured using quantitative methods, and patient experience and provider well-being are assessed using qualitative methods. Healthcare costs will be calculated using the time-driven activity-based costing methodology. The study is set in a Singapore-based national and regional center that takes care of patients with neurological issues.

Objective. Micro-electrocorticographic (μECoG) arrays are able to record neural activities from the cortical surface, without the need to penetrate the brain parenchyma. Owing in part to small electrode sizes, previous studies have demonstrated that single-unit spikes could be detected from the cortical surface, and likely from Layer I neurons of the neocortex. Here we tested the ability to use μECoG arrays to decode, in rats, body position during open field navigation, through isolated single-unit activities. Approach. μECoG arrays were chronically implanted onto primary motor cortex (M1) of Wistar rats, and neural recording was performed in awake, behaving rats in an open-field enclosure. The signals were band-pass filtered between 300–3000 Hz. Threshold-crossing spikes were identified and sorted into distinct units based on defined criteria including waveform morphology and refractory period. Body positions were derived from video recordings. We used gradient-boosting machine to predict body position based on previous 100 ms of spike data, and correlation analyses to elucidate the relationship between position and spike patterns. Main results. Single-unit spikes could be extracted during chronic recording from μECoG, and spatial position could be decoded from these spikes with a mean absolute error of prediction of 0.135 and 0.090 in the x- and y- dimensions (of a normalized range from 0 to 1), and Pearson’s r of 0.607 and 0.571, respectively. Significance. μECoG can detect single-unit activities that likely arise from superficial neurons in the cortex and is a promising alternative to intracortical arrays, with the added benefit of scalability to cover large cortical surface with minimal incremental risks. More studies should be performed in human related to its use as brain–machine interface.