Convolutional Neural Network Based Breast Cancer Risk Stratification Using a Mammographic Dataset

Abstract Background External validation of risk models is critical for risk-stratified breast cancer prevention. We used the Individualized Coherent Absolute Risk Estimation (iCARE) as a flexible tool for risk model development and comparative model validation and to make projections for population risk stratification. Methods Performance of two recently developed models, one based on the Breast and Prostate Cancer Cohort Consortium analysis (iCARE-BPC3) and another based on a literature review (iCARE-Lit), were compared with two established models (Breast Cancer Risk Assessment Tool and International Breast Cancer Intervention Study Model) based on classical risk factors in a UK-based cohort of 64 874 white non-Hispanic women (863 patients) age 35–74 years. Risk projections in a target population of US white non-Hispanic women age 50–70 years assessed potential improvements in risk stratification by adding mammographic breast density (MD) and polygenic risk score (PRS). Results The best calibrated models were iCARE-Lit (expected to observed number of cases [E/O] = 0.98, 95% confidence interval [CI] = 0.87 to 1.11) for women younger than 50 years, and iCARE-BPC3 (E/O = 1.00, 95% CI = 0.93 to 1.09) for women 50 years or older. Risk projections using iCARE-BPC3 indicated classical risk factors can identify approximately 500 000 women at moderate to high risk (>3% 5-year risk) in the target population. Addition of MD and a 313-variant PRS is expected to increase this number to approximately 3.5 million women, and among them, approximately 153 000 are expected to develop invasive breast cancer within 5 years. Conclusions iCARE models based on classical risk factors perform similarly to or better than BCRAT or IBIS in white non-Hispanic women. Addition of MD and PRS can lead to substantial improvements in risk stratification. However, these integrated models require independent prospective validation before broad clinical applications.

Download Full-text

The performance of artificial neural network, logistic regression, and random forest models to predict breast cancer in patients with type 2 diabetes mellitus (Preprint)

10.2196/preprints.14027 ◽

2019 ◽

Author(s):

Chia-Hung Kao

Keyword(s):

Breast Cancer ◽

Neural Network ◽

Diabetes Mellitus ◽

Type 2 Diabetes ◽

Artificial Neural Network ◽

Type 2 Diabetes Mellitus ◽

Logistic Regression ◽

Breast Cancer Risk ◽

Cancer Risk

BACKGROUND Breast cancer incidence may be higher among patients with type 2 diabetes mellitus (T2DM) compared with the general population. This study evaluated the performance of three models for predicting breast cancer risk in patients with T2DM. OBJECTIVE This study evaluated the performance of three models for predicting breast cancer risk in patients with T2DM. METHODS In total, 1,267,867 patients with newly diagnosed T2DM between 2000 and 2012 were identified from Taiwan National Health Insurance Research Database. By employing their data, we created prediction models for detecting an increased risk of subsequent breast cancer development in T2DM patients. The available potential risk factors for breast cancer were also collected for adjustment in the analyses. The Synthetic Minority Oversampling Technique (SMOTE) was used to augment data points in the minority class. Each data point was randomly allocated to the training and test sets at a ratio of approximate 39:1. The performance of artificial neural network (ANN), logistic regression (LR), and random forest (RF) models were determined using the recall, precision, F1 score, and area under receiver operating characteristic curve (AUC). RESULTS The AUCs of all three models were significantly higher than the area of 0.5 for the null hypothesis (0.959, 0.865, and 0.834 for RF, ANN, and LR models, respectively). The RF model has the largest AUC among all models; moreover, it had the highest values in all other metrics. CONCLUSIONS Although all three models could accurately predict high breast cancer risk in patients with T2DM in Taiwan, the RF model demonstrated the best performance. CLINICALTRIAL This is not a chinical trial.

Download Full-text