Automatic image detection of colonic polyps is still an unsolved problem due to the large variation in terms shape, texture, size, and color, existence various polyp-like mimics during colonoscopy. In this paper, we apply a recent region-based convolutional neural network (CNN) approach for automatic images videos obtained from colonoscopy examinations. We use deep-CNN model (Inception Resnet) as transfer learning scheme system. To overcome polyp obstacles small number images, examine...

Automatic polyp detection has been shown to be difficult due various polyp-like structures in the colon and high interclass variations size, color, shape, texture. An efficient method should not only have a correct rate (high sensitivity) but also low false precision specificity). The state-of-the-art methods include convolutional neural networks (CNN). However, CNNs vulnerable small perturbations noise; they sometimes miss same appearing neighboring frames produce number of positives. We...

Colonoscopy is a standard method for screening polyps by highly trained physicians. Miss-detected in colonoscopy are potential risk factor colorectal cancer. In this study, we investigate an automatic polyp classification framework. We aim to compare two different approaches named hand-craft feature and convolutional neural network (CNN) based deep learning method. Combined shape color features used hand craft extraction support vector machine (SVM) adopted classification. For CNN approach,...

Automatic polyp detection and segmentation are highly desirable for colon screening due to miss rate by physicians during colonoscopy, which is about 25%. However, this computerization still an unsolved problem various polyp-like structures in the high interclass variations terms of size, color, shape, texture. In paper, we adapt Mask R-CNN evaluate its performance with different modern convolutional neural networks (CNN) as feature extractor segmentation. We investigate improvement each...

Motor imagery (MI)-based brain-computer interface systems (BCIs) normally use a powerful spatial filtering and classification method to maximize their performance. The common pattern (CSP) algorithm is widely used for MI-based BCIs. In this work, we propose new sparse representation-based (SRC) scheme BCI applications. Sensorimotor rhythms are extracted from electroencephalograms classification. proposed SRC utilizes the frequency band power CSP extract features We analyzed performance of...

One of the major obstacles in automatic polyp detection during colonoscopy is lack labeled training images. In this paper, we propose a framework conditional adversarial networks to increase number samples by generating synthetic Using normal binary form mask which represents only position as an input conditioned image, realistic image generation difficult task generative approach. We edge filtering-based combined train our proposed networks. This enables generations while maintaining...

To decrease colon polyp miss-rate during colonoscopy, a real-time detection system with high accuracy is needed. Recently, there have been many efforts to develop models for detection, but work still required algorithms reliable results. We use single-shot feed-forward fully convolutional neural networks (F-CNN) an accurate system. F-CNNs are usually trained on binary masks object segmentation. propose the of 2D Gaussian instead enable these detect different types polyps more effectively and...

Despite the power of deep neural networks for a wide range tasks, an overconfident prediction issue has limited their practical use in many safety-critical applications. Many recent works have been proposed to mitigate this issue, but most them require either additional computational costs training and/or inference phases or customized architectures output confidence estimates separately. In paper, we propose method with novel loss function, named Correctness Ranking Loss, which regularizes...

A portable hybrid brain monitoring system is proposed to perform simultaneous 16-channel electroencephalogram (EEG) and 8-channel functional near-infrared spectroscopy (fNIRS) measurements. Architecture-optimized analog frontend integrated circuits (Texas Instruments ADS1299 ADS8688A) were used simultaneously achieve 24-bit EEG resolution reliable latency-less (<;0.85 μs) bio-optical Suppression of the noise crosstalk generated by digital circuit components flashing NIR light sources was...

Hypoxia-inducible factor (HIF)-2α and the zinc-ZIP8-MTF1 axis in chondrocytes serve as catabolic regulators of osteoarthritic cartilage destruction by regulating expression genes. We explored possible crosstalk between these signaling pathways its biological significance osteoarthritis (OA).Microarray analysis, various mRNA protein assays were conducted using primary cultured mouse articular experimental OA to reveal molecular mechanisms underlying HIF-2α axis. Experimental mice was induced...

The prediction reliability of neural networks is important in many applications. Specifically, safety-critical domains, such as cancer or autonomous driving, a reliable confidence model's critical for the interpretation results. Modern deep have achieved significant improvement performance different image classification tasks. However, these tend to be poorly calibrated terms output confidence. Temperature scaling an efficient post-processing-based calibration scheme and obtains well In this...

Dry contact electrode-based EEG acquisition is one of the easiest ways to obtain neural information from human brain, providing many advantages such as rapid installation, and enhanced wearability. However, high impedance due insufficient electrical coupling at electrode-scalp interface still remains a critical issue. In this paper, two-wired active dry electrode system proposed by combining finger-shaped spring-loaded probes buffer circuits. The shrinkable bootstrap topology-based circuitry...

Electroencephalogram (EEG) based brain-computer interface (BCI) provides a new communication and control channel for people with severe motor disabilities. Motor imagery sensorimotor rhythm (SMR) analysis is one of the widely used methods in BCI field. However, these signals are very noisy strongly depends on subjects. Therefore, it difficult to classify them thus more powerful classification needed. In this paper, we propose method sparse representation EEG ell-1 minimization. Using Mu...

In a frequency hopping spread spectrum (FHSS) network, the pattern plays an important role in user authentication at physical layer. However, recently, it has been possible to trace through blind estimation method for (FH) signals. If can be reproduced, attacker imitate FH signal and send fake data FHSS system. To prevent this situation, non-replicable system that targets layer of network is required. study, radio fingerprinting-based emitter identification targeting signals was proposed. A...

For the design of electroencephalography (EEG) based BCI systems, crucial issues are to acquire high fidelity EEG signals and provide convenient installation users. Electrodes key components which measure from user's scalp. In this paper, we introduce a dry electrodes for systems. The proposed equipped with six spring loaded probes. They capable acquiring good enough quality without usage conductive gels. To verify performance electrodes, contact impedance compare them those conventional wet...

RESET distribution of phase-change random access memory (PRAM) is highly related to heat fluctuations during write (RESET <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">W</sub> ). In this work we investigate the effect load resistance (R xmlns:xlink="http://www.w3.org/1999/xlink">L</sub> ) with constant voltage method and propose new an optimal R selection equation considering Joule heating thermoelectric effects. Since compensates for...

In this paper, we aim to introduce a design of active dry electrodes for EEG based BCI systems. The proposed consist probes and an circuit easy installation higher quality recording. To verify the evaluation electrodes, tried detect alpha rhythms as typical feature by using our own acquisition boards Matlab. Experimental results show that power rhythm reaches over 50 percent after 10 seconds when subject closes his eyes. We will performance through various analysis such detection visual...

Ball grid array (BGA) packages have been characterized from one port S-parameter measurements by shorting and opening the connection on ball side of BGA packages. Transmission line parameters (resistance, inductance capacitance) using /spl Gamma/ equivalent circuit model are extracted measured S/sub 11/ parameter. Extracted resistances strongly dependent frequency, but inductances capacitances nearly constant up to 500 MHz. well matched those an LCR meter calculated a three-dimensional (3-D)...