An improved SqueezeNet model for the diagnosis of lung cancer in CT scans

Tsivgoulis, Michail; Papastergiou, Thomas; Megalooikonomou, V.

doi:10.1016/j.mlwa.2022.100399

Cited by 13 publications

(4 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performance and efficiency of the suggested classifier is evaluated with existing methodologies such as E‐DBN (Siddiqui et al, 2023a), 3D CNN (Joshua et al, 2021), Squeeze Nodule Net (Tsivgoulis et al, 2022), 3D‐DCNN with multi‐layered filter (Siddiqui et al, 2023b), CCDC‐HNN (Wankhade & Vigneshwari, 2023) and 2D CNN with TPO (Lin et al, 2020). The parameters like accuracy, sensitivity, specificity and F ‐1 score are used to evaluate the efficiency of the proposed approach.…”

Section: Results and Analysismentioning

confidence: 99%

“…However, the analysation for the suggested model for class activation functions are not performed well. Tsivgoulis et al (2022) introduced A Squeeze Nodule Net which is a light weight CNN architecture that has the capability to categorize the nodules as malignant, benign and a mid-range. The ultimate goal of squeeze net is to minimize the computational power and the runtime during the period of classifying the lung images.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Lung cancer computed tomography image classification using Attention based Capsule Network with dispersed dynamic routing

Paramasivam,

Patil,

Konda

et al. 2024

Expert Systems

View full text Add to dashboard Cite

Lung cancer is relying as one of the significant and leading cause for the deaths which are based on cancer. So, an effective diagnosis is a crucial step to save the patients who are all dying due to lung cancer. Moreover, the diagnosis must be performed based on the severity of lung cancer and the severity can be addressed with the help of an optimal classification approach. So, this research introduced an Attention based Capsule Network (A‐Caps Net) with dispersed dynamic routing to perform in‐depth classification of the disease affected partitions of the image and results in better classification results. The attention layer with dispersed dynamic routing evaluates the digit capsule from feature vector in a constant manner. As the first stage, data acquisitioned from datasets such as Lung Nodule Analysis‐16 (LUNA‐16), The Cancer Imaging Archive (TCIA) dataset and Lung Image Database Consortium and Image Database Resource Initiative (LIDC‐IDRI). After acquisitioning data, pre‐processing is done to enhance the resolution of the image using Generative Adversarial Network. The pre‐processed output is given as output for extraction of features that takes place using GLCM and VGG‐16 which extracts the low level features and high level features respectively. Finally, categorization of lung cancer is performed using Attention based Capsule Network (A‐Caps Net) with dispersed dynamic routing which categorize the lung cancer as benign and malignant. The results obtained through experimental analysis exhibits that proposed approach attained better accuracy of 99.57%, 99.91% and 99.29% for LUNA‐16, LIDC‐IDRI and TCIA dataset respectively. The classification accuracy achieved by the proposed approach for LUNA‐16 dataset is 99.57% which is comparably higher than DBN, 3D CNN, Squeeze Nodule Net and 3D‐DCNN with multi‐layered filter with accuracies of 99.16%, 97.17% and 94.1% respectively.

show abstract

Section: Results and Analysismentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Lung cancer computed tomography image classification using Attention based Capsule Network with dispersed dynamic routing

Paramasivam,

Patil,

Konda

et al. 2024

Expert Systems

View full text Add to dashboard Cite

show abstract

“…Current designs for lightweight networks were mainly applied in the following areas: the first was the lightweight design of convolutional layers, such as deep separable convolution [ 68 , 69 , 70 ]. The second was the design of convolutional modules, e.g., the annealing module used in Squeeze Net to achieve light-weighting by reducing the network parameters [ 71 , 72 ]. The third was to use searching mechanisms in the network architecture to disconnect invalid neurons and retain the valid connections alone [ 73 , 74 , 75 ].…”

Section: Methodsmentioning

confidence: 99%

Table Tennis Track Detection Based on Temporal Feature Multiplexing Network

Liu

et al. 2023

Sensors

View full text Add to dashboard Cite

Recording the trajectory of table tennis balls in real-time enables the analysis of the opponent’s attacking characteristics and weaknesses. The current analysis of the ball paths mainly relied on human viewing, which lacked certain theoretical data support. In order to solve the problem of the lack of objective data analysis in the research of table tennis competition, a target detection algorithm-based table tennis trajectory extraction network was proposed to record the trajectory of the table tennis movement in video. The network improved the feature reuse rate in order to achieve a lightweight network and enhance the detection accuracy. The core of the network was the “feature store & return” module, which could store the output of the current network layer and pass the features to the input of the network layer at the next moment to achieve efficient reuse of the features. In this module, the Transformer model was used to secondarily process the features, build the global association information, and enhance the feature richness of the feature map. According to the designed experiments, the detection accuracy of the network was 96.8% for table tennis and 89.1% for target localization. Moreover, the parameter size of the model was only 7.68 MB, and the detection frame rate could reach 634.19 FPS using the hardware for the tests. In summary, the network designed in this paper has the characteristics of both lightweight and high precision in table tennis detection, and the performance of the proposed model significantly outperforms that of the existing models.

show abstract

“…The potential applications of SqueezeNet techniques are various in the field of image and computer vision tasks. The main versatile application of SqueezeNet in healthcare [64] and self-driving cars [66], where compact efficient models are highly desirable. Self-driving cars rely heavily on real-time object detection to safely navigate through their environment.…”

Section: Quantitative Analysis Of Run Time Versus Psnrmentioning

confidence: 99%

SENext: Squeeze-and-ExcitationNext for Single Image Super-Resolution

2023

View full text Add to dashboard Cite

Recent research on image and video processing using convolutional neural networks has shown remarkable improvements, especially in the area of single image super-resolution(SISR). The primary target of SISR is to recover the visually appealing high-resolution (HR) output image from the original degraded low-resolution (LR) input image. However, most recent convolutional neural networks (CNNs)based image super-resolution frameworks often used a deeper and broader network architecture that requires a sizeable computational resource, risk of overfitting, increases computational complexity, and more memory consumption, as well as takes more processing time during the evaluations. To address these issues, we have presented a Squeeze-and-ExcitationNext for Single Image Super-Resolution approach, known as SENext. In brief, the squeeze-and-excitation blocks (SEB) are used in our network architecture with a view to reduce the computational cost and adopt the channel-wise feature mappings to recalibrate the features adaptively. Furthermore, local, sub-local, and global skip connections are employed between each SEB to enable the feature reusability and stabilize training convergence smoothly. Instead of hand-designed bicubic upsampling at pre-processing step, we have performed post-upsampling at the later stage to reconstruct the high-resolution image. Extensive quantitative and qualitative experiments are performed on the benchmark test dataset, including Set5, Set14, BSDS100, Urban100, and Manga109. These experimental evaluations validate the superiority of the SENext over other deep CNN image SR methods in terms of PSNR/SSIM, FLOPs, Number of parameters, processing speed, and visually pleasing effect.

show abstract

An improved SqueezeNet model for the diagnosis of lung cancer in CT scans

Cited by 13 publications

References 47 publications

Lung cancer computed tomography image classification using Attention based Capsule Network with dispersed dynamic routing

Lung cancer computed tomography image classification using Attention based Capsule Network with dispersed dynamic routing

Table Tennis Track Detection Based on Temporal Feature Multiplexing Network

SENext: Squeeze-and-ExcitationNext for Single Image Super-Resolution

Contact Info

Product

Resources

About