Analysis of breast cancer classification robustness with radiomics feature extraction and deep learning techniques

Rashid H. U., İBRİKÇİ T., PAYDAŞ S., BİNOKAY F., ÇEVİK U.

EXPERT SYSTEMS, cilt.39, sa.8, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 39 Sayı: 8
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1111/exsy.13018
  • Dergi Adı: EXPERT SYSTEMS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, ABI/INFORM, Applied Science & Technology Source, Biotechnology Research Abstracts, Business Source Elite, Business Source Premier, Compendex, Computer & Applied Sciences, INSPEC, Library, Information Science & Technology Abstracts (LISTA), Psycinfo
  • Anahtar Kelimeler: breast tumour classification, deep learning, radiomic features, IMAGE
  • Çukurova Üniversitesi Adresli: Evet

Özet

Breast cancer and breast imaging diagnostic procedures are typically carried out using a variety of imaging modalities, including mammography, MRI, and ultrasound. However, ultrasound and mammography have limitations and MRI is recognized as better than other procedures. Recent computational approaches, such as radiomics, applied to image analysis have shown remarkable progress in lowering diagnostic difficulties. This research analysed the robustness of breast tumour classification with feature extraction (radiomics) and a featureless method (deep learning). The proposal consists of two stages: the first stage introduced and explored radiomics-based steps. A total of 111 tumour lesions were used to derive 74 radiomic features consisting of shape, and three separate second-order metrics. Associations of these features were used to classify tumour lesions with four different kernels from support vector machine algorithm. In the confusion matrix analysis, the SVM-RBF kernel developed optimal diagnostic efficiency with a maximum test accuracy of 97.06% on the combination of feature analysis. The second stage developed with deep learning techniques (InceptionV3 and CNN-SVM). A total of 2998 images were used to create the models. In this portion, the CNN-SVM model achieved the highest accuracy, 95.28%, with an AUC of 0.974, where the pre-trained InceptionV3 achieved an AUC of only 0.932. Finally, the obtained result in both stages was discussed together and other related studies.