• Journal of Dental Rehabilitation and Applied Science
• /
• v.38 no.4
• /
• pp.196-203
• /
• 2022

Purpose: This study aimed to evaluate the accuracy and clinical usability of an identification model using deep learning for 79 dental implant types. Materials and Methods: A total of 45396 implant fixture images were collected through panoramic radiographs of patients who received implant treatment from 2001 to 2020 at 30 dental clinics. The collected implant images were 79 types from 18 manufacturers. EfficientNet and Meta Pseudo Labels algorithms were used. For EfficientNet, EfficientNet-B0 and EfficientNet-B4 were used as submodels. For Meta Pseudo Labels, two models were applied according to the widen factor. Top 1 accuracy was measured for EfficientNet and top 1 and top 5 accuracy for Meta Pseudo Labels were measured. Results: EfficientNet-B0 and EfficientNet-B4 showed top 1 accuracy of 89.4. Meta Pseudo Labels 1 showed top 1 accuracy of 87.96, and Meta pseudo labels 2 with increased widen factor showed 88.35. In Top5 Accuracy, the score of Meta Pseudo Labels 1 was 97.90, which was 0.11% higher than 97.79 of Meta Pseudo Labels 2. Conclusion: All four deep learning algorithms used for implant identification in this study showed close to 90% accuracy. In order to increase the clinical applicability of deep learning for implant identification, it will be necessary to collect a wider amount of data and develop a fine-tuned algorithm for implant identification.

• Journal of Biomedical Engineering Research
• /
• v.42 no.4
• /
• pp.150-158
• /
• 2021

Intracranial hemorrhage (ICH) refers to acute bleeding inside the intracranial vault. Not only does this devastating disease record a very high mortality rate, but it can also cause serious chronic impairment of sensory, motor, and cognitive functions. Therefore, a prompt and professional diagnosis of the disease is highly critical. Noninvasive brain imaging data are essential for clinicians to efficiently diagnose the locus of brain lesion, volume of bleeding, and subsequent cortical damage, and to take clinical interventions. In particular, computed tomography (CT) images are used most often for the diagnosis of ICH. In order to diagnose ICH through CT images, not only medical specialists with a sufficient number of diagnosis experiences are required, but even when this condition is met, there are many cases where bleeding cannot be successfully detected due to factors such as low signal ratio and artifacts of the image itself. In addition, discrepancies between interpretations or even misinterpretations might exist causing critical clinical consequences. To resolve these clinical problems, we developed a diagnostic model predicting intracranial bleeding and its subtypes (intraparenchymal, intraventricular, subarachnoid, subdural, and epidural) by applying deep learning algorithms to CT images. We also constructed a visualization tool highlighting important regions in a CT image for predicting ICH. Specifically, 1) 27,758 CT brain images from RSNA were pre-processed to minimize the computational load. 2) Three different CNN-based models (ResNet, EfficientNet-B2, and EfficientNet-B7) were trained based on a training image data set. 3) Diagnosis performance of each of the three models was evaluated based on an independent test image data set: As a result of the model comparison, EfficientNet-B7's performance (classification accuracy = 91%) was a way greater than the other models. 4) Finally, based on the result of EfficientNet-B7, we visualized the lesions of internal bleeding using the Grad-CAM. Our research suggests that artificial intelligence-based diagnostic systems can help diagnose and treat brain diseases resolving various problems in clinical situations.

• Journal of Information Processing Systems
• /
• v.7 no.2
• /
• pp.271-286
• /
• 2011

In this paper, we propose efficient algorithms for parallel prefix computation and sorting on a recursive dual-net. The recursive dual-net $RDN^k$(B) for k > 0 has $(2n_o)^{2K}/2$ nodes and $d_0$ + k links per node, where $n_0$ and $d_0$ are the number of nod es and the node-degree of the base-network B, respectively. Assume that each node holds one data item, the communication and computation time complexities of the algorithm for parallel prefix computation on $RDN^k$(B), k > 0, are $2^{k+1}-2+2^kT_{comm}(0)$ and $2^{k+1}-2+2^kT_{comp}(0)$, respectively, where $T_{comm}(0)$ and $T_{comp}(0)$ are the communication and computation time complexities of the algorithm for parallel prefix computation on the base-network B, respectively. The algorithm for parallel sorting on $RDN^k$(B) is restricted on B = $Q_m$ where $Q_m$ is an m-cube. Assume that each node holds a single data item, the sorting algorithm runs in $O((m2^k)^2)$ computation steps and $O((km2^k)^2)$ communication steps, respectively.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.406-412
• /
• 2022

Scene classification of very high-resolution (VHR) imagery can attribute semantics to land cover in a variety of domains. Real-world application requirements have not been addressed by conventional techniques for remote sensing image classification. Recent research has demonstrated that deep convolutional neural networks (CNNs) are effective at extracting features due to their strong feature extraction capabilities. In order to improve classification performance, these approaches rely primarily on semantic information. Since the abstract and global semantic information makes it difficult for the network to correctly classify scene images with similar structures and high interclass similarity, it achieves a low classification accuracy. We propose a VHR remote sensing image classification model that uses extracts the global feature from the original VHR image using an EfficientNet-V2L CNN pre-trained to detect similar classes. The image is then classified using a multilayer perceptron (MLP). This method was evaluated using two benchmark remote sensing datasets: the 21-class UC Merced, and the 38-class PatternNet. As compared to other state-of-the-art models, the proposed model significantly improves performance.

• Korean Journal of Computational Design and Engineering
• /
• v.9 no.3
• /
• pp.203-211
• /
• 2004

In this paper, we present a method of extracting a regular 2-manifold triangular net from a triangular net including degenerate and self-intersected triangles. This method can be applied to obtaining an offset model without degenerate and self-intersected triangles. Then this offset model can be used to generate CL curves and extract machining features for CAPP The robust and efficient algorithm to detect valid triangles by growing regions from an initial valid triangle is presented. The main advantage of the algorithm is that detection of valid triangles is performed only in valid regions and their adjacent selfintersections, and omitted in the rest regions (invalid regions). This advantage increases robustness of the algorithm. As well as a k-d tree bucketing method is used to detect self-intersections efficiently.

• Journal of electromagnetic engineering and science
• /
• v.6 no.2
• /
• pp.103-109
• /
• 2006

This paper suggests an efficient and comprehensive algorithm of the protection ratio derivation and illustrates some calculated results applicable to the initial planning of frequency coordination in the fixed wireless access networks. The net filter discrimination associated with Tx spectrum mask and overall Rx filter characteristic has been also examined to show the effect of the adjacent channel interference. The calculations for co-channel and adjacent channel protection ratios are performed for the current microwave frequency band of 6.7 GHz including Tx spectrum mask and Rx filter response. According to results, fade margin and co-channel protection ratio reveal 41.4 and 75.2 dB, respectively, for 64-QAM and 60 km at BER $10^{-6}$. It is shown that the net filter discrimination with 40 MHz channel bandwidth provides 28.9 dB at the first adjacent channel, which yields 46.3 dB of adjacent channel protection ratio. In addition, the protection ratio of 38 GHz radio relay system is also reviewed for millimeter wave band applications. The proposed method gives some advantages of an easy and systematic extension for protection ratio calculation and is also applied to frequency coordination in fixed millimeter wave networks.

• ETRI Journal
• /
• v.46 no.2
• /
• pp.263-276
• /
• 2024

Different diseases occur in the brain. For instance, hereditary and progressive diseases affect and degenerate the white matter. Although addressing, diagnosing, and treating complex abnormalities in the brain is challenging, different strategies have been presented with significant advances in medical research. With state-of-art developments in artificial intelligence, new techniques are being applied to brain magnetic resonance images. Deep learning has been recently used for the segmentation and classification of brain images. In this study, we classified normal and pathological brain images using pretrained deep models through transfer learning. The EfficientNet-B5 model reached the highest accuracy of 98.39% on real data, 91.96% on augmented data, and 100% on pathological data. To verify the reliability of the model, fivefold cross-validation and a two-tier cross-test were applied. The results suggest that the proposed method performs reasonably on the classification of brain magnetic resonance images.

• Proceedings of the Korea Electromagnetic Engineering Society Conference
• /
• 2005.11a
• /
• pp.125-130
• /
• 2005

This paper suggests an efficient method of protection ratio calculation and shows some calculated results applicable to frequency coordination in microwave relay system networks, and the net filter discrimination (NFD) associated with Tx spectrum mask and overall Rx filter characteristics has been examined to obtain the adjacent channel protection ratio. The protection ratio comprises several factors such as C/N of modulation scheme, noise-to-interference ratio, multiple interference allowance, fade margins of multi-path and rain attenuation, and NFD. According to computed results for 6.7 GHz, 64-QAM, and 60 km at BER $10^{-6}$, fade margin and co-channel protection ratio are 41.1 and 75.2 dB, respectively, In addition, NFD for channel bandwidth of 40 MHz reveals 28.9 dB at the first adjacent channel, which results in adjacent channel protection ratio of 46.3 dB. The proposed method provides some merits of an easy calculation, systematic extension, and applying the same concept to frequency coordination in millimeter wave relay system networks.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.655-657
• /
• 2021

It is challenging to find breast ultrasound image training dataset to develop an accurate machine learning model due to various regulations, personal information issues, and expensiveness of acquiring the images. However, studies targeting transfer learning for ultrasound breast cancer images classification have not been able to achieve high performance compared to radiologists. Here, we propose an improved transfer learning model for ultrasound breast cancer classification using publicly available dataset. We argue that with a proper combination of ImageNet pre-trained model and optimizer, a better performing model for ultrasound breast cancer image classification can be achieved. The proposed model provided a preliminary test accuracy of 99.5%. With more experiments involving various hyperparameters, the model is expected to achieve higher performance when subjected to new instances.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.1 s.343
• /
• pp.133-142
• /
• 2006

This paper suggests an efficient method of protection ratio(PR) calculation and shows some results of point-to-point radio relay system for frequency coordination. The proposed PR can be expressed as a function of C/N of modulation scheme, noise-to-interference ratio(N/I), multiple interference allowance, fade margins of multi-Path and rain attenuation and net filter discrimination. And PR calculation is performed in view of fade margin, modulation scheme, distance, and interference for actual point-to-point radio relay frequency. According to results for 6.2 GHz, 64-QAM and 60 km at BER 10-6, fade margin and co-channel Protection ratio reveal 41.1dB and 74.9 dB, respectively The merit of presented method provides a systematic and easy calculation by means of PR correction factor related with various parameters and can apply the same concept to frequency coordination of millimeter wave radio relay system.