• 대한방사선기술학회지:방사선기술과학
• /
• 제47권1호
• /
• pp.29-37
• /
• 2024

This study aims to evaluate the performance of the U-Net based learning model that may vary depending on the histogram equalization algorithm. The subject of the experiment were 17 radiology students of this college, and 1,727 data sets in which the region of interest was set in the thyroid after acquiring ultrasound image data were used. The training set consisted of 1,383 images, the validation set consisted of 172 and the test data set consisted of 172. The equalization algorithm was divided into Histogram Equalization(HE) and Contrast Limited Adaptive Histogram Equalization(CLAHE), and according to the clip limit, it was divided into CLAHE8-1, CLAHE8-2. CLAHE8-3. Deep Learning was learned through size control, histogram equalization, Z-score normalization, and data augmentation. As a result of the experiment, the Attention U-Net showed the highest performance from CLAHE8-2 to 0.8355, and the U-Net and BSU-Net showed the highest performance from CLAHE8-3 to 0.8303 and 0.8277. In the case of mIoU, the Attention U-Net was 0.7175 in CLAHE8-2, the U-Net was 0.7098 and the BSU-Net was 0.7060 in CLAHE8-3. This study attempted to confirm the effects of U-Net, Attention U-Net, and BSU-Net models when histogram equalization is performed on ultrasound images. The increase in Clip Limit can be expected to increase the ROI match with the prediction mask by clarifying the boundaries, which affects the improvement of the contrast of the thyroid area in deep learning model learning, and consequently affects the performance improvement.

• 한국방사선학회논문지
• /
• 제17권5호
• /
• pp.663-670
• /
• 2023

본 연구에서는 기존의 노이즈 제거 알고리즘을 적용한 영역 확장 기반의 분할 방법과 U-Net을 이용한 분할 방법의 성능을 정량적 평가인자를 이용하여 비교평가 하고자 하였다. 먼저, 전산화단층검사 영상에 median filter, median modified Wiener filter, fast non-local means algorithm을 모델링하여 적용한 뒤 영역 확장 기반의 분할을 수행하였다. 그리고 U-Net 기반의 분할 모델로 훈련을 진행하여 분할을 수행하였다. 그 후, 노이즈 제거 알고리즘을 사용한 경우와 U-Net을 사용한 경우의 분할 성능을 비교 평가하기 위해 평균 제곱근 편차 (root mean square error, RMSE), 최대 신호 대 잡음비 (peak signal to noise ratio, PSNR), universal quality image index (UQI), 그리고 dice similarity coefficient (DSC)를 측정하였다. 실험 결과, U-Net을 이용하여 분할을 수행했을 때 분할 성능이 가장 향상되었다. RMSE, PSNR, UQI, 그리고 DSC 값은 각각 약 0.063, 72.11, 0.864, 그리고 0.982로 noisy한 영상에 비해 각각 1.97배, 1.09배, 5.30배, 그리고 1.99배 개선된 것을 확인할 수 있었다. 결론적으로, 전산화단층검사영상에서 U-Net이 노이즈 제거 알고리즘에 비해 분할 성능 향상에 효과적임을 입증하였다.

• 방송공학회논문지
• /
• 제27권6호
• /
• pp.840-847
• /
• 2022

심층 신경망(Deep Neural Network) 기반 음성 향상에서 입력 음성의 글로벌 정보와 로컬 정보를 활용하는 것은 모델의 성능과 밀접한 연관성을 갖는다. 최근에는 다중 스케일을 사용하여 입력 데이터의 글로벌 정보와 로컬 정보를 활용하는 중첩 U-Net 구조가 제안되었으며, 이러한 중첩 U-Net은 음성 향상 분야에도 적용되어 매우 우수한 성능을 보였다. 그러나 중첩 U-Net에서 사용되는 단일 skip connection은 중첩된 구조에 알맞게 변형되어야 할 필요성이 있다. 본 논문은 중첩 U-Net 기반 음성 향상 알고리즘의 성능을 최적화하기 위하여 다중 레벨 skip connection(multi-level skip connection, MLS)을 제안하였다. 실험 결과, 제안된 MLS는 기존의 skip connection과 비교하여 다양한 객관적 평가 지표에서 큰 성능 향상을 보이며 이를 통해 MLS가 중첩 U-Net 기반 음성 향상 알고리즘의 성능을 최적화시킬 수 있음을 확인하였다. 또한, 최종 제안 모델은 다른 심층 신경망 기반 음성 향상 모델과 비교하여서도 매우 우수한 성능을 보인다.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.273-275
• /
• 2022

Deep Learning is used nowadays in Nuclei segmentation. While recent developments in theory and open-source software have made these tools easier to implement, expert knowledge is still required to choose the exemplary model architecture and training setup. We compare two popular segmentation frameworks, U-Net and Mask-RCNN, in the nuclei segmentation task and find that they have different strengths and failures. we compared both models aiming for the best nuclei segmentation performance. Experimental Results of Nuclei Medical Images Segmentation using U-NET algorithm Outperform Mask R-CNN Algorithm.

• 한국측량학회지
• /
• 제40권4호
• /
• pp.251-260
• /
• 2022

소나무 재선충 피해나무는 줄어들고 있으나, 피해 지역은 전국으로 확대되고 있다. 최근에 딥러닝 기술이 발전하면서 소나무재선충 고사목 탐지 연구에 적용이 빠르게 시도되고 있다. 본 연구의 목적은 딥러닝 학습데이터의 효과적인 취득과 정확한 참값을 확보하고, 학습을 통해 U-Net 모델의 탐지능력을 보다 향상시키기 위함이다. 이러한 목적달성을 위해 단계별 딥러닝 알고리즘을 적용한 필터링 방법을 이용하여 딥러닝 모델의 불명확한 분석 근거를 최소화하고, 효율적인 분석 및 판단을 할 수 있도록 하였다. 분석결과 U-Net알고리즘을 이용한 소나무재선충 고사목 탐지 및 성능향상에 있어 기간별로 분석한 참값을 이용한 U-Net 모델이 기존에 제공하였던 참값을 이용한 U-Net 모델보다 재현율(Recall)은 -0.5%p, 정밀도(Precision)은 7.6%p, F-1 score는 4.1%p로 분석되었다. 향후 다양한 필터링 기법을 적용하여 재선충 탐지 정밀도를 높일 수 있는 가능성이 있을 것으로 판단되며, 드론 정사영상과 인공지능을 이용한 드론 예찰방법이 소나무재선충 방제 사업에 활용 가능할 것으로 판단된다.

• Smart Structures and Systems
• /
• 제29권1호
• /
• pp.1-16
• /
• 2022

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

• 대한의용생체공학회:의공학회지
• /
• 제42권3호
• /
• pp.100-106
• /
• 2021

Measuring Intima-media thickness (IMT) with ultrasound images can help early detection of coronary artery disease. As a result, numerous machine learning studies have been conducted to measure IMT. However, most of these studies require several steps of pre-treatment to extract the boundary, and some require manual intervention, so they are not suitable for on-site treatment in urgent situations. in this paper, we propose to use deep learning networks U-Net, Attention U-Net, and Pretrained U-Net to automatically segment the intima-media complex. This study also applied the HE, HS, and CLAHE preprocessing technique to wireless portable ultrasound diagnostic device images. As a result, The average dice coefficient of HE applied Models is 71% and CLAHE applied Models is 70%, while the HS applied Models have improved as 72% dice coefficient. Among them, Pretrained U-Net showed the highest performance with an average of 74%. When comparing this with the mean value of IMT measured by Conventional wired ultrasound equipment, the highest correlation coefficient value was shown in the HS applied pretrained U-Net.

• Korean Chemical Engineering Research
• /
• 제57권2호
• /
• pp.274-282
• /
• 2019

최근 빅데이터 과학은 사회현상 모델링을 통한 예측은 물론 강화학습과 결합하여 산업분야 자동제어까지 응용범위가 확대되고 있다. 이러한 추세 가운데 이미지 영상 데이터 활용연구는 화학, 제조, 농업, 바이오산업 등 다양한 산업분야에서 활발히 진행되고 있다. 본 논문은 신경망 기술을 활용하여 영상 데이터의 시맨틱 분할 성능을 개선하고자, U-Net의 계산효율성을 개선한 DeepU-Net 신경망에 AutoML 강화학습 알고리즘을 구현한 NASNet을 결합하였다. BRATS2015 MRI 데이터을 활용해 성능 검증을 수행하였다. 학습을 수행한 결과 DeepU-Net은 U-Net 신경망 구조보다 계산속도 향상 뿐 아니라 예측 정확도도 동등 이상의 성능이 있음을 확인하였다. 또한 이미지 시맨틱 분할 성능을 개선하기 위해서는 일반적으로 적용하는 드롭아웃 층을 빼고, DeepU-Net에 강화학습을 통해 구한 커널과 필터 수를 신경망의 하이퍼 파라미터로 선정했을 때 DeepU-Net보다 학습정확도는 0.5%, 검증정확도는 0.3% 시맨틱 분할 성능을 개선할 수 있었다. 향후 본 논문에서 시도한 자동화된 신경망을 활용해 MRI 뇌 영상진단은 물론, 열화상 카메라를 통한 이상진단, 비파괴 검사 진단, 화학물질 누출감시, CCTV를 통한 산불감시 등 다양한 분야에 응용될 수 있을 것으로 판단된다.

• Korean Journal of Radiology
• /
• 제20권8호
• /
• pp.1275-1284
• /
• 2019

Objective: To develop algorithms using convolutional neural networks (CNNs) for automatic segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) and compare them with conventional algorithms, including a thresholding-based segmentation. Materials and Methods: Between September 2005 and August 2015, 429 patients presenting with acute cerebral ischemia (training:validation:test set = 246:89:94) were retrospectively enrolled in this study, which was performed under Institutional Review Board approval. Ground truth segmentations for acute ischemic lesions on DWI were manually drawn under the consensus of two expert radiologists. CNN algorithms were developed using two-dimensional U-Net with squeeze-and-excitation blocks (U-Net) and a DenseNet with squeeze-and-excitation blocks (DenseNet) with squeeze-and-excitation operations for automatic segmentation of acute ischemic lesions on DWI. The CNN algorithms were compared with conventional algorithms based on DWI and the apparent diffusion coefficient (ADC) signal intensity. The performances of the algorithms were assessed using the Dice index with 5-fold cross-validation. The Dice indices were analyzed according to infarct volumes (< 10 mL, ≥ 10 mL), number of infarcts (≤ 5, 6-10, ≥ 11), and b-value of 1000 (b1000) signal intensities (< 50, 50-100, > 100), time intervals to DWI, and DWI protocols. Results: The CNN algorithms were significantly superior to conventional algorithms (p < 0.001). Dice indices for the CNN algorithms were 0.85 for U-Net and DenseNet and 0.86 for an ensemble of U-Net and DenseNet, while the indices were 0.58 for ADC-b1000 and b1000-ADC and 0.52 for the commercial ADC algorithm. The Dice indices for small and large lesions, respectively, were 0.81 and 0.88 with U-Net, 0.80 and 0.88 with DenseNet, and 0.82 and 0.89 with the ensemble of U-Net and DenseNet. The CNN algorithms showed significant differences in Dice indices according to infarct volumes (p < 0.001). Conclusion: The CNN algorithm for automatic segmentation of acute ischemic lesions on DWI achieved Dice indices greater than or equal to 0.85 and showed superior performance to conventional algorithms.

• 한국멀티미디어학회논문지
• /
• 제24권11호
• /
• pp.1492-1499
• /
• 2021

Skin prick test is widely used in diagnosing allergic sensitization to common inhalant or food allergens, in which positivities are manually determined by calculating the areas or mean diameters of wheals and erythemas provoked by allergens pricked into patients' skin. In this work, we propose a segmentation algorithm over U-Net, one of the FCN models of deep learning, to help us more objectively grasp the erythema boundaries. The performance of the model is analyzed by comparing the results of automatic segmentation of the test data to U-Net with the results of manual segmentation. As a result, the average Dice coefficient value was 94.93%, the average precision and sensitivity value was 95.19% and 95.24% respectively. We find that the proposed algorithm effectively discriminates the skin's erythema boundaries. We expect this algorithm to play an auxiliary role in skin prick test in real clinical trials in the future.