Real-time arrhythmia detection using convolutional neural networks.

Cardiovascular diseases, such as heart attack and congestive heart failure, are the leading cause of death both in the United States and worldwide. The current medical practice for diagnosing cardiovascular diseases is not suitable for long-term, out-of-hospital use. A key to long-term monitoring is the ability to detect abnormal cardiac rhythms, i.e., arrhythmia, in real-time. Most existing studies only focus on the accuracy of arrhythmia classification, instead of runtime performance of the workflow. In this paper, we present our work on supporting real-time arrhythmic detection using convolutional neural networks, which take images of electrocardiogram (ECG) segments as input, and classify the arrhythmia conditions. To support real-time processing, we have carried out extensive experiments and evaluated the computational cost of each step of the classification workflow. Our results show that it is feasible to achieve real-time arrhythmic detection using convolutional neural networks. To further demonstrate the generalizability of this approach, we used the trained model with processed data collected by a customized wearable sensor from a lab setting, and the results shown that our approach is highly accurate and efficient. This research provides the potentials to enable in-home real-time heart monitoring based on 2D image data, which opens up opportunities for integrating both machine learning and traditional diagnostic approaches.

Advances in translational neuropathic research: example of enantioselective pharmacokinetic-pharmacodynamic modeling of ketamine-induced pain relief in complex regional pain syndrome.

Historically, complex regional pain syndrome (CRPS) was poorly defined, which meant that scientists and clinicians faced much uncertainty in the study, diagnosis, and treatment of the syndrome. The problem could be attributed to a nonspecific diagnostic criteria, unknown pathophysiologic causes, and limited treatment options. The two forms of CRPS still are painful, debilitating disorders whose sufferers carry heavy emotional burdens. Current research has shown that CRPS I and CRPS II are distinctive processes, and the presence or absence of a partial nerve lesion distinguishes them apart. Ketamine has been the focus of various studies involving the treatment of CRPS; however, currently, there is incomplete data from evidence-based studies. The question as to why ketamine is effective in controlling the symptoms of a subset of patients with CRPS and not others remains to be answered. A possible explanation to this phenomenon is pharmacogenetic differences that may exist in different patient populations. This review summarizes important translational work recently published on the treatment of CRPS using ketamine.

Critical failure of a percutaneous discectomy probe requiring surgical removal during disc decompression.

OBJECTIVE: We report a complication while performing a percutaneous disc decompression at the L4-L5 level using a Dekompressor Percutaneous Discectomy Probe. CASE REPORT: A 54-year-old male was referred to the pain clinic for possible percutaneous disc decompression. For the procedure the Dekompressor unit was inserted over the stylette and a percutaneous disc decompression was performed for 1 to 2 minutes using the channeling technique. Approximately (3/4) mL of disc nucleus was successfully removed. Upon withdrawal of the Dekompressor unit, it was noted that the probe was no longer connected to the device handle. Fluoroscopic imaging showed that approximately 4 inches of the probe remained in the patient and that the auger's distal end was still lodged in the disc. An incision was made, and the auger was successfully removed by a neurosurgeon. CONCLUSIONS: The patient had an uneventful recovery. Manipulation of the auger should be performed in a linear motion as best as possible and under fluoroscopic guidance.

Pilot evaluation of a prototype critical care blood glucose monitor in normal volunteers.

BACKGROUND: Availability of a highly accurate in-hospital automated blood glucose (BG) monitor could facilitate implementation of intensive insulin therapy protocols through effective titration of insulin therapy, improved BG control, and avoidance of hypoglycemia. We evaluated a functional prototype BG monitor designed to perform frequent automated blood sampling for glucose monitoring. METHODS: Sixteen healthy adult volunteer subjects had intravenous catheter insertions in a forearm or hand vein and were studied for 8 hours. The prototype monitor consisted of an autosampling unit with a precise computer-controlled reversible syringe pump and a glucose analytical section. BG was referenced against a Yellow Springs Instrument (YSI) laboratory analyzer. Sampling errors for automated blood draws were assessed by calculating the percent of failed draws, and BG data were analyzed using the Bland and Altman technique. RESULTS: Out of 498 total sample draws, unsuccessful draws were categorized as follow: 11 (2.2%) were due to autosampler technical problems, 21 (4.2%) were due to catheter-related failures, and 37 (7.4%) were BG meter errors confirmed by a glucometer-generated error code. Blood draw difficulties or failures related to the catheter site (e.g., catheter occlusion or vein collapse) occurred in 6/15 (40%) subjects. Mean BG bias versus YSI was 0.20 +/- 12.6 mg/dl, and mean absolute relative difference was 10.4%. CONCLUSIONS: Automated phlebotomy can be performed in healthy subjects using this prototype BG monitor. The BG measurement technology had suboptimal accuracy based on a YSI reference. A more accurate BG point-of-care testing meter and strip technology have been incorporated into the future version of this monitor. Development of such a monitor could alleviate the burden of frequent BG testing and reduce the risk of hypoglycemia in patients on insulin therapy.