The value of predicting breast cancer with a DBT 2.5D deep learning model

Abstract

Objective: To evaluate the accuracy and efficacy of a 2.5-dimensional deep learning (DL) model based on digital breast tomosynthesis (DBT) in predicting breast cancer.

Methods: Through a retrospective analysis of data from 361 patients with breast tumor lesions treated at Shandong Provincial Hospital Affiliated to Shandong First Medical University between 2018 and 2020, this study utilized deep convolutional neural networks (DCNN) to automatically extract key features from DBT images. By applying dimensionality reduction and feature fusion selection, a variety of machine learning predictive models based on a 2.5-dimensional feature set were constructed. Additionally, a comprehensive predictive model was developed by combining univariate and multivariate logistic regression analyses with clinical data. The model's performance was assessed using receiver operating characteristic (ROC) curves, area under the curve (AUC) values, and accuracy rates.

Results: In the test set, DBT 2.5D deep learning-based logistic regression, LightGBM, multilayer perceptron, and comprehensive models achieved accuracies of 72.2%, 75.0%, 79.2%, and 80.6%; AUCs of 0.826, 0.756, 0.859, and 0.871; sensitivities of 63.8%, 70.2%, 80.9%, and 87.2%; specificities of 88.0%, 84.0%, 76.0%, and 68.0%; PPVs of 90.9%, 89.2%, 86.4%, and 83.7%; NPVs of 56.4%, 60.0%, 67.9%, and 73.9%; and F1 scores of 75.0%, 78.6%, 83.5%, and 85.4%, respectively. These results underscore the high efficiency and potential of DBT 2.5D deep learning models in breast cancer diagnosis, particularly the comprehensive model's superior performance across key metrics.

Conclusion: The 2.5D deep learning model based on DBT shows good performance in preoperative breast cancer prediction, with its integration with clinical data further enhancing its effectiveness. The combination of deep learning and radiomics offers a viable approach for early breast cancer diagnosis, supporting the development of more accurate personalized diagnostic and treatment strategies.

Keywords: 2.5D deep learning; Breast cancer; Diagnosis; Digital breast tomosynthesis; Predictive model.

Fig. 1

Patient selection flowchart

Fig. 2

The delineation of ROI in the images from a patient of 52 years old and with infiltrating ductal carcinoma. A Craniocaudal DBT image; B tumor delineation schematic; C 3D tumor mask. D based on the 3D tumor mask, corresponding 2D DBT images are created, where “0” represents the layer with the largest cross-section of the tumor for each patient, and “1” and “− 1” depict the adjacent upper and lower slice images; E: Schematic of the segmented tumor 2D image

Fig. 3

The modeling process of deep learning prediction models, clinical prediction models, and integrated prediction models. Initially, a residual convolutional neural network is pretrained using the ImageNet dataset, followed by fine-tuning with digital breast tomosynthesis data to extract deep learning features. Subsequently, a clinical model is created based on clinical risk prediction factors. Finally, the deep learning model and the clinical model are combined to form an integrated comprehensive predictive model, culminating in the generation of a nomogram for assisting in diagnostic decisions

Fig. 4

ROC curves of eight machine learning algorithms for model prediction. A ROC curves of deep learning predictive models using 8 different machine learning algorithms in the training set; B ROC curves of deep learning predictive models using 8 different machine learning algorithms in the test set

Fig. 5

A–E Represent the features and their importance included in the Logistic Regression, Random Forest, ExtraTrees, XGBoost, and LightGBM models, respectively

Fig. 6

ROC curves of clinical prediction model (Clinic Signature), deep learning predictive model (Rad Signature) and combined model (Nomogram). A ROC curves of the clinical predictive model, deep learning predictive model, and combined predictive model in the training set; B ROC curves of the clinical predictive model, deep learning predictive model, and combined predictive model in the test set

Fig. 7

A Decision curves for the clinical prediction model, deep learning prediction models, and combined prediction model; B calibration curves for the clinical prediction model, deep learning prediction models, and combined prediction model

Fig. 8

Nomogram constructed based on deep learning features and clinical risk factors