Search Results (5)

Atrial fibrillation (AF) is a type of cardiac arrhythmia with a worldwide prevalence of more than 37 million among the adult population. This elusive disease is a major risk factor for ischemic stroke, along with increased rates of significant morbidity and eventual mortality. It is clinically diagnosed using medical-grade electrocardiogram (ECG) sensors in ambulatory settings. The recent emergence of consumer-grade wearables equipped with photoplethysmography (PPG) sensors has exhibited considerable promise for non-intrusive continuous monitoring in free-living conditions. However, the scarcity of large-scale public PPG datasets acquired from wearable devices hinders the development of intelligent automatic AF detection algorithms unaffected by motion artifacts, saturated ambient noise, inter- and intra-subject differences, or limited training data. In this work, we present a deep learning framework that leverages convolutional layers with a bidirectional long short-term memory (CNN-BiLSTM) network and an attention mechanism for effectively classifying raw AF rhythms from normal sinus rhythms (NSR). We derive and feed heart rate variability (HRV) and pulse rate variability (PRV) features as auxiliary inputs to the framework for robustness. A larger teacher model is trained using the MIT-BIH Arrhythmia ECG dataset. Through transfer learning (TL), its learned representation is adapted to a compressed student model (32x smaller) variant by using knowledge distillation (KD) for classifying AF with the UMass and MIMIC-III datasets of PPG signals. This results in the student model yielding average improvements in accuracy, sensitivity, F1 score, and Matthews correlation coefficient of 2.0%, 15.05%, 11.7%, and 9.85%, respectively, across both PPG datasets. Additionally, we employ Gradient-weighted Class Activation Mapping (Grad-CAM) to confer a notion of interpretability to the model decisions. We conclude that through a combination of techniques such as TL and KD, i.e., pre-trained initialization, we can utilize learned ECG concepts for scarcer PPG scenarios. This can reduce resource usage and enable deployment on edge devices. Full article

Compression is the cornerstone of treatment for lymphoedema; however, in patients with Stage III lymphoedema that causes a significant alteration of limb shape and size, additional creative problem solving is required in order to find a solution that fits the person, their limb and their personal circumstances. Medical adaptive compression (MAC) systems provide solutions to complex situations. We present two cases of individuals who completed an inpatient rehabilitation stay for intensive treatment of their lymphoedema and were fitted with MAC devices to ensure long-term success. Full article

Hydrocephalus is a neurological disorder characterized by an abnormal accumulation of cerebrospinal fluid (CSF) within the ventricular system of the brain, leading to cerebral ventricular dilation, brain parenchyma compression, and neuronal cell loss. Surgery is an effective method of draining excessive amounts of CSF. Ventriculoperitoneal shunt (VPS) allows excess CSF to divert into the abdomen; this device is the most commonly used in the treatment of hydrocephalus both in veterinary and human patients. This study aims to describe the application of two types of VPS, low-pressure valve and medium-pressure valve, using a frameless stereotactic neuronavigational system in eight severe hydrocephalus in dogs and, in particular, analyze the prevalence of postoperative overshunting. Non-communicating hydrocephalus was found in seven dogs, whereas the rest of them had communicating hydrocephalus caused by traumatic brain injury with a skull fracture. The criteria for pressure valve selection depended on the intraoperative intraventricular pressure (IVP) that was determined by the adaptive manometer, according to the human protocol. Low-pressure valve placement was performed in five dogs, and the others received medium-pressure valve placement. The follow-up period was 2 weeks, 4–12 weeks, and 12 weeks to 12 months. Pre- and postoperative information including neurological signs, CT-Scan or MRI, medical treatment, complications, and ventricular volume were compared in all dogs. Seven dogs showed neurological improvement within 2 weeks after surgery. Overshunting was seen in four dogs who received low-pressure valve placement. Three of them had shunt infections within 4 to 6 weeks after surgery. One dog underwent shunt revision from a low-pressure valve to a medium-pressure valve caused by severe overshunting and progressive neurological signs. In addition, cognitive and learning improvements were evaluated based on the owners’ feedback, and neurological signs were examined during the follow-up period in two dogs that received low-pressure valve placement. We conclude that a medium-pressure valve is recommended for overshunting prevention. However, low-pressure valve placement seems to improve cognitive function and learning ability, which is related to an increase in the brain parenchyma observed during long-term monitoring. Moreover, we also report our experience and surgical procedure for frameless stereotactic ventricular shunt placement (FSVSP) in VPS surgery in dogs affected by hydrocephalus. Full article

Auxetic materials exhibit a negative Poisson’s ratio under tension or compression, and such counter-intuitive behavior leads to enhanced mechanical properties such as shear resistance, impact resistance, and shape adaptability. Auxetic materials with these excellent properties show great potential applications in personal protection, medical health, sensing equipment, and other fields. However, there are still many limitations in them, from laboratory research to real applications. There have been many reported studies applying auxetic materials or structures to the development of sensing devices in anticipation of improving sensitivity. This review mainly focuses on the use of auxetic materials or auxetic structures in sensors, providing a broad review of auxetic-based sensing devices. The material selection, structure design, preparation method, sensing mechanism, and sensing performance are introduced. In addition, we explore the relationship between the auxetic mechanism and the sensing performance and summarize how the auxetic behavior enhances the sensitivity. Furthermore, potential applications of sensors based on the auxetic mechanism are discussed, and the remaining challenges and future research directions are suggested. This review may help to promote further research and application of auxetic sensing devices. Full article

The optimal airway management strategy during cardiopulmonary resuscitation is uncertain. In the case of out-of-hospital cardiac arrest, a high chest compression fraction is paramount to obtain the return of spontaneous circulation and improve survival and neurological outcomes. To improve this fraction, providing continuous chest compressions should be more effective than using the conventional 30:2 ratio. Airway management should, however, be adapted, since face-mask ventilation can hardly be carried out while continuous compressions are administered. The early insertion of a supraglottic device could therefore improve the chest compression fraction by allowing ventilation while maintaining compressions. This is a protocol for a multicenter, parallel, randomized simulation study. Depending on randomization, each team made up of paramedics and emergency medical technicians will manage the 10-min scenario according either to the standard approach (30 compressions with two face-mask ventilations) or to the experimental approach (continuous manual compressions with early insertion of an i-gel^® supraglottic device to deliver asynchronous ventilations). The primary outcome will be the chest compression fraction during the first two minutes of cardiopulmonary resuscitation. Secondary outcomes will be chest compression fraction (per cycle and overall), compressions and ventilations quality, time to first shock and to first ventilation, user satisfaction, and providers’ self-assessed cognitive load. Full article