• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.19 no.1
• /
• pp.163-169
• /
• 1989

The purpose of this study is to investigate on the clinical and radiographic patterns of residual cyst of the jaw for early diagnosis and treatment. The auther studied 87 cases of residual cyst with regard to age, sex distribution, the site of the lesion and several radiographic features. The results were as follows: 1. The average age was found to be 42.6 years, with a range of 15 to 84 years. The incidence was highest in the third and fourth decades(50.6%) and total 87 cases consist of 47 males and 38 females. 2. The common clinical symptoms were pus discharge, swelling, pain and no symptoms was presented in 5 cases(12.5%). 3. Residual cysts were found to be 46.0% maxillary anterior region, 18.4% maxillary molar region, 17.2% mandibular molar region and to be more common in the maxilla(70.1%) than in the mandible(29.9%) 4. Most of residual cysts were unilocular type(86 cases, 98.8%), showing distinct border(62 cases, 71.3%) with smooth margin(78 cases, 89.7%). 5. The adjacent teeth showed root resorption in 13 cases(14.9%), and root divergence in 16 cases(18.4%). 6. The residual cysts extended to the nasal fossa(22 cases, 22.5%), the maxillary sinus(19 cases, 19.4%) and caused the displacement of the mandibular canal wall (11 cases, 11.2%)

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.4
• /
• pp.87-94
• /
• 2007

Imbalances in focus, luminance and color between stereo Pairs could cause disparity vector estimation error and increment of transmission data. If the distribution of errors in residual image is large, it may influence to lowering of compression performance. Therefore, in this paper, we propose an efficient balancing method between stereo pairs to reduce the effect. For this, we registrated stereo images using a FFT based method to consider the pixels in the occluded region, we eliminated the pixels of blocks which has large error of disparity vector estimation in balancing function estimation. The balancing function has estimated using histogram specification, local information of target image and residual image between stereo images. Experiments show that the proposed method is effective in error distribution, PSNR and disparity vector estimation. We expect that our method can be improving compression efficiency in stereo coding system.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.10
• /
• pp.403-410
• /
• 2019

Image de-snowing aims at eliminating the negative influence by snow particles and improving scene understanding in images. In this paper, a perceptual generative adversarial network based a single image snow removal method is proposed. The residual U-Net is designed as a generator to generate the snow free image. In order to handle various sizes of snow particles, the inception module with different filter kernels is adopted to extract multiple resolution features of the input snow image. Except the adversarial loss, the perceptual loss and total variation loss are employed to improve the quality of the resulted image. Experimental results indicate that our method can obtain excellent performance both on synthetic and realistic snow images in terms of visual observation and commonly used visual quality indices.

• Journal of Digital Policy
• /
• v.3 no.3
• /
• pp.9-14
• /
• 2024

In this paper, proposes a method using a multi block structure composed of residual blocks with adaptive weights to improve the quality of results in single image super resolution. In the process of generating super resolution images using deep learning, the most critical factor for enhancing quality is feature extraction and application. While extracting various features is essential for restoring fine details that have been lost due to low resolution, issues such as increased network depth and complexity pose challenges in practical implementation. Therefore, the feature extraction process was structured efficiently, and the application process was improved to enhance quality. To achieve this, a multi block structure was designed after the initial feature extraction, with nested residual blocks inside each block, where adaptive weights were applied. Additionally, for final high resolution reconstruction, a multi kernel image reconstruction process was employed, further improving the quality of the results. The performance of the proposed method was evaluated by calculating PSNR and SSIM values compared to the original image, and its superiority was demonstrated through comparisons with existing algorithms.

• Progress in Medical Physics
• /
• v.18 no.1
• /
• pp.35-41
• /
• 2007

In this work practical considerations of a pulsed arterial spin labeling MRI are presented to reliable multi-slice perfusion measurements In the human brain. Three parameters were considered in this study. First, In order to improve slice profile and Inversion efficiency of a labeling pulse a high power Inversion pulse of adiabatic hyperbolic secant was designed. A $900^{\circ}$ rotation of the flip angle was provided to make a good slice profile and excellent Inversion efficiency. Second, to minimize contributions of a residual magnetization be4ween Interleaved scans of control and labeling we tested three different conditions which were applied 1) only saturation pulses, 2) only spotter gradients, and 3) combinations of saturation pulses and spotter gradients Applications of bo4h saturation pulses and spoiler gradients minimized the residual magnetization. Finally, to find a minimum gap between a tagged plane and an imaging plane we tested signal changes of the subtracted image between control and labeled Images with varying the gap. The optimum gap was about 20mm. In conclusion, In order to obtain high quality of perfusion Images In human brain It Is Important to use optimum parameters. Before routinely using In clinical studios, we recommend to make optimizations of sequence parameters.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.8
• /
• pp.67-73
• /
• 2015

In a distribution of digital image, there is a serious problem that is the image alteration by a forger. For the problem solution, this paper proposes the forensic decision algorithm of a median filtering (MF) image using the feature vector based on a coefficient of variation (c.v.) of Fourier transform. In the proposed algorithm, we compute Fourier transform (FT) coefficients of row and column line respectively of an image first, then c.v. between neighboring lines is computed. Subsquently, 10 Dim. feature vector is defined for the MF detection. On the experiment of MF detection, the proposed scheme is compared to MFR (Median Filter Residual) and Rhee's MF detection schemes that have the same 10 Dim. feature vector both. As a result, the performance is excellent at Unaltered, JPEG (QF=90), Down scaling (0.9) and Up scaling (1.1) images, and it showed good performance at Gaussian filtering ($3{\times}3$) image. However, in the performance evaluation of all measured items of the proposed scheme, AUC (Area Under ROC (Receiver Operating Characteristic) Curve) by the sensitivity and 1-specificity approached to 1 thus, it is confirmed that the grade of the performance evaluation is rated as 'Excellent (A)'.

• Journal of Broadcast Engineering
• /
• v.16 no.5
• /
• pp.793-796
• /
• 2011

H.264/AVC compresses video data by applying DCT transform, quantization and entropy coding processes to the residual signal obtained by inter/intra prediction. This paper proposes a method enhancing an existing DC offset adjustment technology which uses information of neighboring blocks to reduce residual information for improving coding efficiency. DC offset information is not sent over bitstreams, but calculated in the same way both in the decoder and in the encoder. Experimental results show that the proposed method enhances coding efficiency by 0.25% in average BD-Rate compared to H.264/AVC and gives better or worse coding efficiency compared to the existing DC offset method depending on video sequences with coding efficiency degradation by 0.09% in average BD-Rate. This experimental results also show that further coding efficiency improvement is possible by applying the proposed method adaptively to slice or macroblock coding units.

• Korean Journal of Digital Imaging in Medicine
• /
• v.16 no.2
• /
• pp.1-7
• /
• 2014

This study is to increase the accuracy of the diagnosis of benign prostatic hyperplasia by presenting a method that can accurately measure the residual urine amount of the bladder by using an ultrasound image. Agar powder, Propanol and distilled water were used as materials for making a phantom. In order to measure the volume, a $10m{\ell}$ cylinder, syringe and beaker were used. The image was obtained by scanning phantoms produced into six shapes. Each constant value was obtained by using the expression designed to measure the residual urine amount of the bladder and was compared and analyzed. The measuring method of Bladder volume was presented and a constant value for each shape was obtained and five observers measured it five times. According to the results of clinical application, the errors of Ellipse-beanbag, Shield-shield were 11.0%, 18.2%, respectively. Constant values depending on the shape of each phantom were presented in order to accurately measure the volume of the bladder in measuring the amount of residual urine for the diagnosis of benign prostatic hyperplasia. The accuracy of the volume using this was verified statistically(p > 0.05). Therefore, it is considered to be useful in diagnosing benign prostatic hyperplasia by using the ultrasound imaging measuring method presented.

• Journal of radiological science and technology
• /
• v.46 no.4
• /
• pp.277-285
• /
• 2023

In this study, pneumonia identification networks with the small number of layers were constructed by using chest X-ray images. The networks had similar trainable-parameters, and the performance of the trained models was quantitatively evaluated with the modification of the network architectures. A total of 6 networks were constructed: convolutional neural network (CNN), VGGNet, GoogleNet, residual network with identity blocks, ResNet with bottleneck blocks and ResNet with identity and bottleneck blocks. Trainable parameters for the 6 networks were set in a range of 273,921-294,817 by adjusting the output channels of convolution layers. The network training was implemented with binary cross entropy (BCE) loss function, sigmoid activation function, adaptive moment estimation (Adam) optimizer and 100 epochs. The performance of the trained models was evaluated in terms of training time, accuracy, precision, recall, specificity and F1-score. The results showed that the trained models with the small number of layers precisely detect pneumonia from chest X-ray images. In particular, the overall quantitative performance of the trained models based on the ResNets was above 0.9, and the performance levels were similar or superior to those based on the CNN, VGGNet and GoogleNet. Also, the residual blocks affected the performance of the trained models based on the ResNets. Therefore, in this study, we demonstrated that the object detection networks with the small number of layers are suitable for detecting pneumonia using chest X-ray images. And, the trained models based on the ResNets can be optimized by applying appropriate residual-blocks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.9
• /
• pp.2341-2348
• /
• 1996

This paper proposes an image compression algorithm that adopts projection scheme on mean-residual metod. Sub-blocks of an image are encoded using mean-residual method where mean value is predicted according to that of neighboring blocks. Projection scheme with 8 directions is applied to the compression of residual signals of blocks. Projection vectors are finite-state vector quantized according to the projection angle of nighboring blocks in order to exploit the correlation among them. Side information to represent the repetition of projection is run-length coded while the information for projection direction is compressed using entropy encoding. The proposed scheme apears to be better in PSNR performance when compared with conventional projection scheme as well as in subjective quality preserving the edges of images better than most tranform methods which usually require heavy computation load.