http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Deep Neural Network based Disease Severity Regression for Diagnosis of Abdominal Aortic Aneurysm
Joo Hyeon Im(임주현),Sooho Kim(김수호),In Chan Lee(이인찬),Jin-Oh Hahn(한진오),Byeng D. Youn(윤병동) 대한기계학회 2020 대한기계학회 춘추학술대회 Vol.2020 No.12
This paper proposes an abdominal aortic aneurysm severity regression algorithm by combining a physical model and deep learning. There are three typical problems in providing personalized medical service: 1) insufficiency of data, 2) dispersion of disease severity, 3) lack of individuality. In this study, blood pressure waveform data was obtained by implementing AAA in arterial tree simulation model. The data was acquired when there was a disease and when it was normal, and individual diversity was given through parameter change. Also, the shape of the aortic aneurysm was made into four types. The data showed a similar tendency to the blood pressure waveform in the presence of an aneurysm in the literature. Severity regression was performed through a deep neural network using data from blood pressure waveforms with various severity levels as input. As a result, it was confirmed that the regression performed well through the deep neural network.
FDM 방식 3D 프린터에 대한 딥러닝 기반 물성 추정
임주현(Joo Hyeon Im),김원곤(Wongon Kim),안성훈(Sung-Hoon Ahn),윤병동(Byeng Dong Youn) 대한기계학회 2022 대한기계학회 춘추학술대회 Vol.2022 No.11
Additive manufacturing (AM) is playing a major role in Industry 4.0. AM can simplify the fabrication of complex shapes while minimizing manufacturing time and cost. However, the poor surface quality and weak mechanical properties of the output hinder the broad application of AM. In this study, we propose a deep learning-based method for estimating the mechanical property in a fused deposition modeling-type 3D printer. Acceleration signals were acquired in the process of outputting a tensile specimen through construction of a test-bed. In addition, a tensile test was performed using the specimen to collect information on mechanical properties. After that, features were extracted from the divided signals to analyze the correlation with the mechanical properties. Finally, the quality of highly correlated property was estimated through deep neural network. We confirmed that deep learning-based method is good at estimating property through mean square error loss and root mean square error.
FDM 방식 3D 프린터에 대한 딥러닝 기반 물성 추정
임주현(Joo Hyeon Im),김원곤(Wongon Kim),안성훈(Sung-Hoon Ahn),윤병동(Byeng Dong Youn) 대한기계학회 2022 대한기계학회 춘추학술대회 Vol.2022 No.11
Additive manufacturing (AM) is playing a major role in Industry 4.0. AM can simplify the fabrication of complex shapes while minimizing manufacturing time and cost. However, the poor surface quality and weak mechanical properties of the output hinder the broad application of AM. In this study, we propose a deep learning-based method for estimating the mechanical property in a fused deposition modeling-type 3D printer. Acceleration signals were acquired in the process of outputting a tensile specimen through construction of a test-bed. In addition, a tensile test was performed using the specimen to collect information on mechanical properties. After that, features were extracted from the divided signals to analyze the correlation with the mechanical properties. Finally, the quality of highly correlated property was estimated through deep neural network. We confirmed that deep learning-based method is good at estimating property through mean square error loss and root mean square error.
비침습적 말초동맥질환 진단을 위한 CNN 기반 앙상블 모델
이동휴(Donghyu Lee),임주현(Joo Hyeon Im),이인찬(In Chan Lee),김수호(Sooho Kim),한진오(Jin-Oh Hahn),윤병동(Byeng D. Youn) 대한기계학회 2021 대한기계학회 춘추학술대회 Vol.2021 No.11
In order to develop affordable and non-invasive peripheral arterial disease (PAD) screening methods, machine learning-based diagnostic techniques are receiving increasing interest. A recent study of CNN-based pulse waveform analysis techniques for PAD diagnosis demonstrated promise and potential as a diagnostic tool. Accordingly, in this study, Bayesian hyperparameter optimization and ensemble techniques are applied to enhance the performance of CNN-based PAD diagnosis models. For training and evaluation of the PAD diagnosis model, the blood pressure waveform data of virtual PAD patients were generated from a validated transmission line model. The trained diagnostic model was evaluated by various evaluation metrics of the classification performance according to the severity and location of the disease.