Exploring the Potential Performance of Fibroscan for Predicting and Evaluating Metabolic Syndrome using a Feature Selected Strategy of Machine Learning

Metabolic syndrome (MetS) includes several conditions that can increase an individual's predisposition to high-risk cardiovascular events, morbidity, and mortality. Non-alcoholic fatty liver disease (NAFLD) is a predominant cause of cirrhosis, which is a global indicator of liver transplantation and is considered the hepatic manifestation of MetS. FibroScan^® provides an accurate and non-invasive method for assessing liver steatosis and fibrosis in patients with NAFLD, via a controlled attenuation parameter (CAP) and liver stiffness measurement (LSM or E) scores and has been widely used in current clinical practice. Several machine learning (ML) models with a recursive feature elimination (RFE) algorithm were applied to evaluate the importance of the CAP score. Analysis by ANOVA revealed that five symptoms at different CAP and E score levels were significant. All eight ML models had accuracy scores > 0.9, while treebags and random forest had the best kappa values (0.6439 and 0.6533, respectively). The CAP score was the most important variable in the seven ML models. Machine learning models with RFE demonstrated that using the CAP score to identify patients with MetS may be feasible. Thus, a combination of CAP scores and other significant biomarkers could be used for early detection in predicting MetS.

Intraoperative EMG recovery patterns and outcomes after RLN traction-related amplitude decrease during monitored thyroidectomy

OBJECTIVES

Traction injury is the most common type of recurrent laryngeal nerve (RLN) injury in thyroid surgery. Intraoperative neuromonitoring (IONM) facilitates early detection of adverse electromyography (EMG) effect, and this corrective maneuver can reduce severe and repeated nerve injury. This study aimed to evaluate intraoperative patterns and outcomes of EMG decrease and recovery by traction injury.

METHODS

644 patients received nerve monitored thyroidectomy with 1142 RLNs at risk were enrolled. Intermittent IONM with stimulating dissecting instrument (real-time during surgical procedure) and trans-thyroid cartilage EMG recording method (without electrode malpositioning issue) were used for nerve stimulation and signal recording. When an EMG amplitude showed a decrease of >50% during RLN dissection, the surgical maneuver was paused immediately. Nerve dissection was restarted when the EMG amplitude was stable.

RESULTS

44/1142 (3.9%) RLNs exhibited a >50% EMG amplitude decrease during RLN dissection and all (100%) showed gradual progressive amplitude recovery within a few minutes after releasing thyroid traction (10 recovered from LOS; 34 recovered from a 51-90% amplitude decrease). Three EMG recovery patterns were noted, A-complete EMG recovery (n=14, 32%); B-incomplete EMG recovery with an injury point (n=16, 36%); C-incomplete EMG recovery without an injury point (n=14, 32%). Patients with postoperative weak or fixed vocal cord mobility in A, B, and C were 0(0%), 7(44%), and 2(14%), respectively. Complete EMG recovery was found in 14 nerves, and incomplete recovery was found in another 30 nerves. Temporary vocal cord palsy was found in 6 nerves due to unavoidable repeated traction.

CONCLUSION

Early detection of traction-related RLN amplitude decrease allows monitoring of intraoperative EMG signal recovery during thyroid surgery. Different recovery patterns show different vocal cord function outcomes. To elucidate the recovery patterns can assist surgeons in the intraoperative decision making and postoperative management.

Gigahertz quantized charge pumping in graphene quantum dots

Single-electron pumps are set to revolutionize electrical metrology by enabling the ampere to be redefined in terms of the elementary charge of an electron. Pumps based on lithographically fixed tunnel barriers in mesoscopic metallic systems and normal/superconducting hybrid turnstiles can reach very small error rates, but only at megahertz pumping speeds that correspond to small currents of the order of picoamperes. Tunable barrier pumps in semiconductor structures are operated at gigahertz frequencies, but the theoretical treatment of the error rate is more complex and only approximate predictions are available. Here, we present a monolithic, fixed-barrier single-electron pump made entirely from graphene that performs at frequencies up to several gigahertz. Combined with the record-high accuracy of the quantum Hall effect and proximity-induced Josephson junctions, quantized-current generation brings an all-graphene closure of the quantum metrological triangle within reach. Envisaged applications for graphene charge pumps outside quantum metrology include single-photon generation via electron-hole recombination in electrostatically doped bilayer graphene reservoirs, single Dirac fermion emission in relativistic electron quantum optics and read-out of spin-based graphene qubits in quantum information processing.

Clinical characteristics and outcomes of patients with obstructive sleep apnoea requiring intensive care

We reviewed the clinical characteristics, required intervention and short- and long-term outcomes in obstructive sleep apnoea (OSA) patients requiring intensive care. A retrospective, single-centre, observational cohort study was undertaken in a multidisciplinary teaching medical and surgical intensive care unit. Adult patients with OSA (apnoea-hypopnoea index of 5 or higher) requiring intensive care from January 2000 to January 2005 were included. Thirty-seven OSA patients (age: 58 +/- 14 years, male:female 27:10) were admitted due to respiratory (n=12, 32%), cerebrovascular (n=8, 22%), cardiovascular (n=16, 43%) and infectious events (n=1, 2.7%). Comparing the clinical features, polysomnographic data and outcome among these groups, we found that OSA patients admitted due to respiratory events had significantly higher Acute Physiology and Chronic Health Evaluation II scores, lower arterial blood gas pH, higher PaCO2, a higher incidence of respiratory failure (92%) and required non-invasive ventilation after extubation (73%), and higher intensive care unit readmission rates than patients admitted due to cerebrovascular events and cardiovascular events (P < 0.05). No difference was found in the in-hospital and long-term mortality rate. The most common reason for intensive care unit admission in critically ill OSA patients was a cardiovascular event, followed by respiratory and cerebrovascular events. The baseline polysomnographic data of the OSA patients were not correlated with their clinical features and outcomes in the intensive care unit. A more complicated clinical course and higher intensive care unit readmission rate were encountered in OSA patients admitted due to respiratory events. Further studies would be required to evaluate the efficacy of non-invasive ventilation for facilitation of extubation in OSA patients presenting with hypercapnic respiratory failure.