Amended Convolutional Neural Network with Global Average Pooling for Image Classification

Al-Sabaawi, Aiman; Ibrahim, Hassan M.; Arkah, Zinah Mohsin; Al-Amidie, Muthana; Alzubaidi, Laith

doi:10.1007/978-3-030-71187-0_16

Cited by 20 publications

(9 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this setup, all layers were frozen except for the final fully connected layer. The output tensor of the base model was processed through a series of layers: 1) a global average pooling 2D layer to reduce spatial dimensions (31); 2) a flattened layer to convert the 2D matrix into a 1D vector; 3) a dense layer with 512 units where ReLU activation was employed; and 4) a final output layer with a SoftMax activation function designed to predict subtypes was implemented.…”

Section: Model Selectionmentioning

confidence: 99%

OmicsFootPrint: a framework to integrate and interpret multi-omics data using circular images and deep neural networks

Tang,

Prodduturi,

Thompson

et al. 2024

Preprint

View full text Add to dashboard Cite

Background: The integration and interpretation of multi-omics data are crucial in elucidating the pathophysiological mechanisms of complex diseases. While Deep Neural Networks (DNNs) and Graph Neural Networks (GNNs) have revolutionized the analysis of such data in cancer and neurodegenerative diseases, there remains a critical need for improved multi-omics analytic methodologies. Methods: We developed OmicsFootPrint, a novel framework for transforming multi-omics data into two-dimensional circular images for each sample, enabling intuitive representation and analysis. Utilizing DNNs for analysis, this framework incorporates the SHapley Additive exPlanations (SHAP) algorithm for model interpretation. We benchmarked OmicsFootPrint performance with data sets from The Cancer Genome Atlas (TCGA); we classified subtypes of lung cancer and breast cancer, specifically the PAM50 subtype and histological classifications. Additionally, we applied OmicsFootPrint to predict drug response data in cancer cell lines. Results: Evaluation of lung cancer subtypes using TCGA data sets, OmicsFootPrint achieved an average Area Under Curve (AUC) of over 0.90 for distinguishing between lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) subtypes. Notably, the EfficientNetV2 model reached an AUC of 0.98±0.02. For the TCGA BRCA (breast cancer) PAM50 subtype classification, encompassing Luminal A, Luminal B, Basal, and HER2 subtypes (836 samples in total), the framework reported an average AUC of 0.83±0.07. Furthermore, the application of OmicsFootPrint to the BRCA data set for distinguishing between invasive lobular carcinoma (ILC) and invasive ductal carcinoma (IDC) showed an average AUC of 0.87±0.04 with four-omics datatypes. Additionally, in drug response studies, OmicsFootPrint outperformed published algorithms, achieving the highest average AUC of 0.74 across all data sets. OmicsFootPrint also demonstrated robust performance with reduced training sample sizes. Conclusions: OmicsFootPrint represents a significant advancement, offering a comprehensive, efficient, and interpretable framework for multi-omics data analysis. Its innovative approach in transforming extensive multi-omics data sets into compact, circular images not only reduces memory requirements but also enhances the interpretability of complex biological data. This approach facilitates personalized treatment strategies and enriches our understanding of disease mechanisms, marking a significant step forward in the field of multi-omics research.

show abstract

Section: Model Selectionmentioning

confidence: 99%

OmicsFootPrint: a framework to integrate and interpret multi-omics data using circular images and deep neural networks

Tang,

Prodduturi,

Thompson

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…The GAP operation is an abbreviation for Global Average Pooling, which performs average pooling on each channel of the input feature map to obtain a new feature map of size [1 × 1 × c]. The 1 × 1 convolution is a convolution operation with a kernel size of 1 × 1, which can reduce or increase the dimension of the input feature map [39]. In the Slim Neck module, the 1 × 1 convolution is used to calculate the weight feature map.…”

Section: Overhead Power Line Damage Detection Based On Slim Neckmentioning

confidence: 99%

Overhead Power Line Damage Detection: An Innovative Approach Using Enhanced YOLOv8

Wu,

Liao,

Chen

et al. 2024

Electronics

View full text Add to dashboard Cite

This paper presents an enhanced version of YOLOv8 specifically designed for detecting damage in overhead power lines. Firstly, to improve the model’s robustness, an adaptive threshold mechanism is introduced that can dynamically adjust the detection threshold based on the brightness, contrast, and other characteristics of the input image. Secondly, a novel convolution method, GSConv, is adopted in the YOLOv8 framework, which balances the model’s running speed and accuracy. Finally, a lightweight network structure, Slim Neck, is introduced, effectively reducing the model’s complexity and computational load while maintaining good performance. These improvements enable our YOLOv8 model to achieve excellent performance in detecting ‘thunderbolt’ and ‘break’ types of cable damage. Experimental results show that the improved YOLOv8 network model has an average detection accuracy (mAP) of 90.2%, a recall rate of 91.6%, and a precision of 89.8% on the ‘Cable Damage Detection’ dataset from RoboFlow for ‘thunderbolt’. For ‘break’, the mAP is 86.5%, the recall rate is 84.1%, and the precision is 86.1%. Compared with the original YOLOv8 model, these indicators have been significantly improved, highlighting the high practical value and strong generalization ability of the proposed algorithm in detecting damage to overhead power lines. This also demonstrates the high practical value of the method in future research directions.

show abstract

“…By inputting a chest radiograph of size 256 × 256, the R-CNN outputs PAWP, quantitatively. Inputted images are convoluted by VGG16 [21]-based convolutional layers, and the convoluted data are flattened by the global average pooling layer (GAP layer) [22]. The model was trained by cross validating 748 samples.…”

Section: Developed R-cnnmentioning

confidence: 99%

“…= 1.752 + 0.248 2 = 1.876 (22) as the threshold defined in Equation (20). This value is α th in Fig.…”

Section: Parameters Settingmentioning

confidence: 99%

Reliability Metrics of Explainable CNN based on Wasserstein Distance for Cardiac Evaluation

Omae

Kakimoto

Saito

et al. 2022

Preprint

View full text Add to dashboard Cite

In recent works, convolutional neural networks (CNN) have been used in the non-invasive examination of the cardiac region for estimating pulmonary artery wedge pressure (PAWP) from chest radiographs.Moreover, because CNNs are able to output activated regions, physicians can estimate PAWP along with reasons. However, when new patient radiograph data are fed into the CNN, there is a possibility that activated regions that contain areas other than the cardiac appear.In this case, although we expect a large estimation error, it is not well known.Therefore, we verify this hypothesis by distance theory and statistic approaches.In particular, we build the probability distributions for the cardiac region and the regression activation map (RAM) and measure the similarity between these distributions by Wasserstein distance (WSD).When the CNN estimates PAWP from areas other than the cardiac region, the WSD value is high.Therefore, WSD is a reliability metrics for explainable CNN.We created two groups, normal and anomaly classes, based on WSD values.Chest radiographs which had a high WSD were assigned to the anomaly class, and those with a low WSD were assigned to the normal class. By comparing the normal and anomaly classes based on the PAWP estimation error, we confirmed that the errors from the anomaly class were higher than those of the normal class.Therefore, physicians need to be aware that there might be large estimation errors when activated regions contain areas other than the cardiac.

show abstract

Amended Convolutional Neural Network with Global Average Pooling for Image Classification

Cited by 20 publications

References 18 publications

OmicsFootPrint: a framework to integrate and interpret multi-omics data using circular images and deep neural networks

OmicsFootPrint: a framework to integrate and interpret multi-omics data using circular images and deep neural networks

Overhead Power Line Damage Detection: An Innovative Approach Using Enhanced YOLOv8

Reliability Metrics of Explainable CNN based on Wasserstein Distance for Cardiac Evaluation

Contact Info

Product

Resources

About