A blueprint for precise and fault-tolerant analog neural networks.

Analog computing has reemerged as a promising avenue for accelerating deep neural networks (DNNs) to overcome the scalability challenges posed by traditional digital architectures. However, achieving high precision using analog technologies is challenging, as high-precision data converters are costly and impractical. In this work, we address this challenge by using the residue number system (RNS) and composing high-precision operations from multiple low-precision operations, thereby eliminating the need for high-precision data converters and information loss. Our study demonstrates that the RNS-based approach can achieve ≥99% FP32 accuracy with 6-bit integer arithmetic for DNN inference and 7-bit for DNN training. The reduced precision requirements imply that using RNS can achieve several orders of magnitude higher energy efficiency while maintaining the same throughput compared to conventional analog hardware with the same precision. We also present a fault-tolerant dataflow using redundant RNS to protect the computation against noise and errors inherent within analog hardware.

Changes in substance P levels of inferior turbinate in patients with mucosal contact headache.

INTRODUCTION: Mucosal contact headache is a referred pain that arises from contact between the nasal septum and the lateral nasal wall. Evidence supports the role of substance P in a contact headache such that release of substance P from sensory nerve endings causes inflammation and allergy. OBJECTIVES: This study aimed to determine possible differences in substance P levels in inferior turbinate hypertrophy creating a contact headache. METHODS: 28 patients who had contact headaches (study group) and 16 volunteers with no complaints were included in the study. Substance P levels in the inferior turbinate tissue samples were quantified using a commercially available substance P EIA kit. RESULTS: In the study group average substance P levels were 2.65±0.27pg/mg tissue (range: 0.61-5.44) and in the control group it was 1.77±0.27pg/mg tissue (range: 0.11-4.35). The difference was statistically significant between the two groups (p=0.0215). Average preoperative headache group visual analog scale scores was 5.93±0.38 (2-9) and the turbinate volume was 6.56±0.35cm3 (3.50-10.30). The control group turbinate volume was 4.71±0.39cm3 (2.50-7.70). We found a correlation between the visual analog scale scores and substance P levels such that substance P levels were higher in visual analog scale scores above 5 (p=0.001). CONCLUSION: This study demonstrates the relationship between intranasal contact headaches and increased mucosal substance P levels. We also found that there is no correlation with substance P levels and volume of the inferior turbinate.

Can we use serum copeptin levels as a biomarker in obstructive sleep apnea syndrome?

OBJECTIVE: The aim of this study was to compare serum copeptin levels in patients with obstructive sleep apnea syndrome (OSA) and simple snorers without sleep apnea; and to investigate relationships between copeptin levels and polysomnographic parameters. METHODS: Serum copeptin levels were determined using enzyme-linked immunosorbent assay in 47 patients with OSA and 12 patients without OSA (control group). Full-night polysomnography was performed in each patient. Patients with OSA were divided into three groups according to their Apnea Hypopnea Index (AHI) scores: mild OSA (5 < AHI < 15), moderate OSA (15 < AHI < 30), and severe OSA (AHI > 30). RESULTS: A total of 59 patients were included in the study. There were 23 female (39.0%) and 36 male (61.0%) subjects. The range of ages of study subjects was between 27 and 63 (mean 44.75 ± 9.64) years. According to the AHI values, patients were classified into four groups: simple snoring (n = 13), mild OSA (n = 10), moderate OSA (n = 15), and severe OSA (n = 21). Statistically significant differences between AHI groups in terms of age, Epworth score, and neck circumference. According to multiple comparison results for age, the difference between simple snoring and moderate OSA was statistically significant. According to multiple comparison results for Epworth score, the difference between simple snoring and severe OSA was statistically significant. According to multiple comparison results for neck circumference, a similar result was found like Epworth Sleepiness Scale score. The difference between AHI groups by gender was tested by a Pearson χ(2) test and was found to be statistically significant. There was no statistically significant difference among AHI groups in terms of copeptin. There was a statistically significant correlation of copeptin with AHI during rapid eye movement (REM) sleep; however, the correlation coefficient was not sufficiently large. CONCLUSIONS: Increased serum copeptin concentration may reflect a response to stress in some diseases. This is well documented especially in cardiovascular diseases; however, we could not find any difference in OSA groups in terms of copeptin levels.