Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Feb 16;13(4):757.
doi: 10.3390/diagnostics13040757.

DTBV: A Deep Transfer-Based Bone Cancer Diagnosis System Using VGG16 Feature Extraction

G Suganeshwari  1 , R Balakumar  2 , Kalimuthu Karuppanan  3 , Sahaya Beni Prathiba  2   4 , Sudha Anbalagan  4 , Gunasekaran Raja  2
Affiliations

DTBV: A Deep Transfer-Based Bone Cancer Diagnosis System Using VGG16 Feature Extraction

G Suganeshwari et al. Diagnostics (Basel). .

Abstract

Among the many different types of cancer, bone cancer is the most lethal and least prevalent. More cases are reported each year. Early diagnosis of bone cancer is crucial since it helps limit the spread of malignant cells and reduce mortality. The manual method of detection of bone cancer is cumbersome and requires specialized knowledge. A deep transfer-based bone cancer diagnosis (DTBV) system using VGG16 feature extraction is proposed to address these issues. The proposed DTBV system uses a transfer learning (TL) approach in which a pre-trained convolutional neural network (CNN) model is used to extract features from the pre-processed input image and a support vector machine (SVM) model is used to train using these features to distinguish between cancerous and healthy bone. The CNN is applied to the image datasets as it provides better image recognition with high accuracy when the layers in neural network feature extraction increase. In the proposed DTBV system, the VGG16 model extracts the features from the input X-ray image. A mutual information statistic that measures the dependency between the different features is then used to select the best features. This is the first time this method has been used for detecting bone cancer. Once selected features are selected, they are fed into the SVM classifier. The SVM model classifies the given testing dataset into malignant and benign categories. A comprehensive performance evaluation has demonstrated that the proposed DTBV system is highly efficient in detecting bone cancer, with an accuracy of 93.9%, which is more accurate than other existing systems.

Keywords: X-ray image; bone cancer; convolutional neural network; health care; medical image processing; support vector machines; transfer learning.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Overview of the proposed DTBV system for the detection of bone cancer.
Figure 2
Figure 2
Healthy bones.
Figure 3
Figure 3
Malignant bones.
Figure 4
Figure 4
Pre-processing of X-ray images.
Figure 5
Figure 5
Confusion matrix of test data.
Figure 6
Figure 6
(a) Comparison of the proposed DTBV system with other ML models for classification. (b) The ROC curves of different classification models. (c) Precision–recall curve obtained for the different classification models.
Figure 7
Figure 7
Comparison of the proposed DTBV system with other existing systems.
See this image and copyright information in PMC

References

    1. Muhammad K., Khan S., Ser J.D., Albuquerque V.H.C.D. Deep Learning for Multigrade Brain Tumor Classification in Smart Healthcare Systems: A Prospective Survey. IEEE Trans. Neural Netw. Learn. Syst. 2021;32:507–522. doi: 10.1109/TNNLS.2020.2995800. - DOI - PubMed
    1. Rahman M., Rashid S.M., Nayem Ferdous Khan M., Biswas A., Mahmud A. Symptom Wise Age Prediction of Cancer Patients using Classifier Comparison and Feature Selection; Proceedings of the 2019 22nd International Conference on Computer and Information Technology (ICCIT); Dhaka, Bangladesh. 18–20 December 2019; pp. 1–6.
    1. Cancer Stat Facts. [(accessed on 1 January 2021)]; Available online: https://seer.cancer.gov/statfacts/html/bones.html.
    1. Gaume M., Chevret S., Campagna R., Larousserie F., Biau D. The appropriate and sequential value of standard radiograph, computed tomography and magnetic resonance imaging to characterize a bone tumor. Sci. Rep. 2022;12:6196. doi: 10.1038/s41598-022-10218-8. - DOI - PMC - PubMed
    1. Sun J., Xing F., Braun J., Traub F., Rommens P.M., Xiang Z., Ritz U. Progress of Phototherapy Applications in the Treatment of Bone Cancer. Int. J. Mol. Sci. 2021;22:11354. doi: 10.3390/ijms222111354. - DOI - PMC - PubMed

LinkOut - more resources