Evaluation of Parameter Fine-Tuning with Transfer Learning for Osteoporosis Classification in Knee Radiograph

Abubakar, Usman; Boukar, Moussa Mahamat; Adeshina, Steve A.

doi:10.14569/ijacsa.2022.0130829

Cited by 8 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The issue with medical imaging is the lack of huge data sets. Because it isn't recommended to start from scratch and build a DCNN, the medical images can be categorized by using the features learned through a process called transfer learning [26]. The ensemble architecture suggested here will make sure that all of the descriptors needed for picture classification are there, so that the process can go smoothly.…”

Section: Methodsmentioning

confidence: 99%

COVID-19 Disease Detection based on X-Ray Image Classification using CNN with GEV Activation Function

Mohamed¹,

Elsamahy²,

Salem³

2022

IJACSA

View full text Add to dashboard Cite

The globe was rocked by unprecedented levels of disruption, which had devastating effects on daily life, global health, and global economy. Since the COVID-19 epidemic started, methods for delivering accurate diagnoses for multicategory classification have been proposed in this work (COVID vs. normal vs. pneumonia). XceptionNet and Dense Net, two transfer learning pre-trained model networks, are employed in our CNN model. The low-level properties of the two DCNN structures were combined and used to a classifier for the final prediction. To get better results with unbalanced data, we used the GEV activation function (generalized extreme value) to augment the training dataset using data augmentation for validation accuracy, which allowed us to increase the training dataset while still maintaining validation accuracy with the output classifier. The model has been put through its paces in two distinct scenarios. In the first instance, the model was tested using Image Augmentation for train data and the GEV (generalized extreme value) function for output class, and it got a 94% accuracy second instance Model evaluations were conducted without data augmentation and yielded an accuracy rating of 95% for the output class.

show abstract

Section: Methodsmentioning

confidence: 99%

COVID-19 Disease Detection based on X-Ray Image Classification using CNN with GEV Activation Function

Mohamed¹,

Elsamahy²,

Salem³

2022

IJACSA

View full text Add to dashboard Cite

show abstract

“…This is achieved through the combination of depthwise separable convolution and residual connection techniques, which effectively reduce computational costs. Consequently, several studies have proposed the use of EfficientNet for osteoporosis screening [37][38][39]. On the other hand, ResNet focuses on learning the residual between the input and the desired output.…”

Section: State-of-the-art Cnn Models For Automatic Screening Of Athle...mentioning

confidence: 99%

“…For instance, Norio Yamamoto et al evaluated the predictive ability of EfficientNet on hip radiographs and achieved a sensitivity of 82.26%, specificity of 92.16%, and an AUC of 92.19% [37]. Usman Bello Abubakar et al employed Efficient-Net with fine-tuning techniques to classify osteoporosis in knee radiographs, achieving a sensitivity and specificity of 86% [39]. EfficientNet effectively manages model complexity and computational costs through composite scaling factors, while still maintaining high performance.…”

Section: Efficientnetmentioning

confidence: 99%

GLCM-Based FBLS: A Novel Broad Learning System for Knee Osteopenia and Osteoprosis Screening in Athletes

Chen,

Zheng,

Duan

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Due to the physical strain experienced during intense workouts, athletes are at a heightened risk of developing osteopenia and osteoporosis. These conditions not only impact their overall health but also their athletic performance. The current clinical screening methods for osteoporosis are limited by their high radiation dose, complex post-processing requirements, and the significant time and resources needed for implementation. This makes it challenging to incorporate them into athletes’ daily training routines. Consequently, our objective was to develop an innovative automated screening approach for detecting osteopenia and osteoporosis using X-ray image data. Although several automated screening methods based on deep learning have achieved notable results, they often suffer from overfitting and inadequate datasets. To address these limitations, we proposed a novel model called the GLCM-based fuzzy broad learning system (GLCM-based FBLS). Initially, texture features of X-ray images were extracted using the gray-level co-occurrence matrix (GLCM). Subsequently, these features were combined with the fuzzy broad learning system to extract crucial information and enhance the accuracy of predicting osteoporotic conditions. Finally, we applied the proposed method to the field of osteopenia and osteoporosis screening. By comparing this model with three advanced deep learning models, we have verified the effectiveness of GLCM-based FBLS in the automatic screening of osteoporosis for athletes.

show abstract

“… To categorize various food products using transfer learning, a recognition model was introduced in the paper [16].  The paper [18] gives an evaluation of pre-trained models for the detection of osteoporosis which is a bone disease in knee radiographs. VGG16 and VGG16 with fine-tuning were used in this study.…”

Section: Introductionmentioning

confidence: 99%

Investigating the Impact of Train / Test Split Ratio on the Performance of Pre-Trained Models with Custom Datasets

Bichri,

Chergui,

Hain

2024

IJACSA

View full text Add to dashboard Cite

The proper allocation of data between training and testing is a critical factor influencing the performance of deep learning models, especially those built upon pre-trained architectures. Having the suitable training set size is an important factor for the classification model's generalization performance. The main goal of this study is to find the appropriate training set size for three pre-trained networks using different custom datasets. For this aim, the study presented in this paper explores the effect of varying the train / test split ratio on the performance of three popular pre-trained models, namely MobileNetV2, ResNet50v2 and VGG19, with a focus on image classification task. In this work, three balanced datasets never seen by the models have been used, each containing 1000 images divided into two classes. The train / test split ratios used for this study are: 60-40, 70-30, 80-20 and 90-10. The focus was on the critical metrics of sensitivity, specificity and overall accuracy to evaluate the performance of the classifiers under the different ratios. Experimental results show that, the performance of the classifiers is affected by varying the training / testing split ratio for the three custom datasets. Moreover, with the three pretrained models, using more than 70% of the dataset images for the training task gives better performance.

show abstract

Evaluation of Parameter Fine-Tuning with Transfer Learning for Osteoporosis Classification in Knee Radiograph

Cited by 8 publications

References 15 publications

COVID-19 Disease Detection based on X-Ray Image Classification using CNN with GEV Activation Function

COVID-19 Disease Detection based on X-Ray Image Classification using CNN with GEV Activation Function

GLCM-Based FBLS: A Novel Broad Learning System for Knee Osteopenia and Osteoprosis Screening in Athletes

Investigating the Impact of Train / Test Split Ratio on the Performance of Pre-Trained Models with Custom Datasets

Contact Info

Product

Resources

About