Interpretation of lung disease classification with light attention connected module

Choi, Youngjin; Lee, Hong-Chul

doi:10.1016/j.bspc.2023.104695

Cited by 18 publications

(4 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Model CNN yang diusulkan menunjukkan tingkat akurasi 90%. Diharapkan dapat berperan penting sebagai alat bantu untuk membantu staf medis apakah ada kelainan pernapasan selama pola inhalasi atau ekspirasi (Choi & Lee, 2023).…”

Section: Penelitian Terkaitunclassified

Utilization of Spiking Neural Network (SNN) in X-Ray Image for Lung Disease Detection

Ningtias,

Rofi’i,

Pramudita

2024

JITE

View full text Add to dashboard Cite

The large number of cases of lung disease means that doctors have difficulty in making initial diagnoses, making them prone to misdiagnoses. One type of lung disease that is included in the vulnerable category is pneumonia. Early detection of the condition of the lungs affected by bacterial pneumonia can be carried out by screening using the X-Ray examination modality, namely Digital Radiography (DR). However, in practice, the diagnosis process on Citra DR takes a long time because it requires competent medical personnel (specialists). A system is needed that can help medical personnel to speed up the process of diagnosing lung disease and get accurate results so that misdiagnosis does not occur. The aim of this research is to utilize the Spiking Neural Network (SNN) method for classifying lung disease from DR images. The system was created using MATLAB with the initial step of creating a read data program, namely reading DR image secondary data in .jpg format taken from Kaggle.com. This research uses DR image data totaling 200 images. Next, standardize the size to 50 x 50 pixels. Then segmenting the image divides the gray level histogram into two different parts of the image automatically without requiring user assistance to enter threshold values for normal and pneumonia images. Then convert the image to 1 dimension and create a manual program for the training data using 50 normal images and 50 pneumonia images. Lastly, create a program to test the data using 100 normal images and 100 pneumonia images. Based on the results of data testing, a confusion matrix was obtained from 200 images with sensitivity of 87%, specificity of 69%, precision of 73.7288%, recall of 69%, and accuracy of 78%

show abstract

Section: Penelitian Terkaitunclassified

Utilization of Spiking Neural Network (SNN) in X-Ray Image for Lung Disease Detection

Ningtias,

Rofi’i,

Pramudita

2024

JITE

View full text Add to dashboard Cite

show abstract

“…The basic models can be referred to [ 33 , 126 , 127 ]. Preferably, the model undergoes some tailoring or tuning of its structure based on the classification task and optimization strategy [ 24 , 128 , 129 ]. For example, the FNN-based method transforms the lung sound into a combination representation of acoustic characteristics, then feeds it to the FNN for abnormal sound identification [ 18 ].…”

Section: Deep Learning In Lung Sound Analysismentioning

confidence: 99%

Deep learning-based lung sound analysis for intelligent stethoscope

Huang,

Qiao

et al. 2023

Military Med Res

View full text Add to dashboard Cite

Auscultation is crucial for the diagnosis of respiratory system diseases. However, traditional stethoscopes have inherent limitations, such as inter-listener variability and subjectivity, and they cannot record respiratory sounds for offline/retrospective diagnosis or remote prescriptions in telemedicine. The emergence of digital stethoscopes has overcome these limitations by allowing physicians to store and share respiratory sounds for consultation and education. On this basis, machine learning, particularly deep learning, enables the fully-automatic analysis of lung sounds that may pave the way for intelligent stethoscopes. This review thus aims to provide a comprehensive overview of deep learning algorithms used for lung sound analysis to emphasize the significance of artificial intelligence (AI) in this field. We focus on each component of deep learning-based lung sound analysis systems, including the task categories, public datasets, denoising methods, and, most importantly, existing deep learning methods, i.e., the state-of-the-art approaches to convert lung sounds into two-dimensional (2D) spectrograms and use convolutional neural networks for the end-to-end recognition of respiratory diseases or abnormal lung sounds. Additionally, this review highlights current challenges in this field, including the variety of devices, noise sensitivity, and poor interpretability of deep models. To address the poor reproducibility and variety of deep learning in this field, this review also provides a scalable and flexible open-source framework that aims to standardize the algorithmic workflow and provide a solid basis for replication and future extension: https://github.com/contactless-healthcare/Deep-Learning-for-Lung-Sound-Analysis.

show abstract

“…Learning the distribution of weights among feature maps, thus improving the feature extraction capability of the model. Efficient channel attention requires only a few parameters to produce remarkable results [42]. Due to the superior performance of ECA-Net, many studies have been conducted to use it for the adaptation of channel feature weights [43,44].…”

Section: Eca-netmentioning

confidence: 99%

Deep transfer learning rolling bearing fault diagnosis method based on convolutional neural network feature fusion

Yu,

Fu,

Song

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Current deep-learning methods are often based on significantly large quantities of labeled fault data for supervised training. In practice, it is difficult to obtain samples of rolling bearing failures. In this paper, a transfer learning-based feature fusion convolutional neural network approach for bearing fault diagnosis is proposed. Specifically, the raw vibration signal features and the corresponding time-frequency image features of the input data are extracted by a one-dimensional convolutional neural network and a pre-trained ConvNeXt, respectively, and connected by a feature fusion strategy. Then, the fine-tuning method based on transfer learning can effectively reduce the reliance on labeled samples in the target domain. A wide convolution kernel is introduced in the time-domain signal feature extraction to increase the receptive field, which is combined with the channel attention mechanism to further optimize the feature quality. Finally, two common bearing datasets are utilized for fault diagnosis experiments. The experimental results show that the proposed model achieves an average accuracy of more than 98.63% in both cross-working conditions and cross-device diagnosis tasks. Meanwhile, anti-noise experiments and ablation experiments further validate the accuracy and robustness of the proposed method.

show abstract

Interpretation of lung disease classification with light attention connected module

Cited by 18 publications

References 54 publications

Utilization of Spiking Neural Network (SNN) in X-Ray Image for Lung Disease Detection

Utilization of Spiking Neural Network (SNN) in X-Ray Image for Lung Disease Detection

Deep learning-based lung sound analysis for intelligent stethoscope

Deep transfer learning rolling bearing fault diagnosis method based on convolutional neural network feature fusion

Contact Info

Product

Resources

About