Predicting the molecular subtype of breast cancer and identifying interpretable imaging features using machine learning algorithms

Ma, Mengwei; Liu, Renyi; Chen, Wen; Xu, Weimin; Xu, Zeyuan; Wang, Sina; Wu, Jiefang; Pan, Derun; Zheng, Bowen; Qin, Genggeng; Chen, Weiguo

doi:10.1007/s00330-021-08271-4

Cited by 56 publications

(36 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, to understand how the models yielded their predictions, SHAP values have been proposed as the most effective method for a visual explanation of the model and for presenting properties of local accuracy and consistency [ 46 , 47 ]. Indeed, some studies have used SHAP values to select important features for predicting BC molecular subtypes from images [ 48 ].…”

Section: Discussionmentioning

confidence: 99%

Gradient Boosting Machine Identified Predictive Variables for Breast Cancer Patients Pre- and Post-Radiotherapy: Preliminary Results of an 8-Year Follow-Up Study

et al. 2022

View full text Add to dashboard Cite

Radiotherapy (RT) is part of the standard treatment of breast cancer (BC) because of its effects on relapse reduction and survival. However, response to treatment is highly variable, and some patients may develop disease progression (DP), a second primary cancer, or may succumb to the disease. Antioxidant systems and inflammatory processes are associated with the onset and development of BC and play a role in resistance to treatment. Here, we report our investigation into the clinical evolution of BC patients, and the impact of RT on the circulating levels of the antioxidant enzyme paraoxonase-1 (PON1), cytokines, and other standard biochemical and hematological variables. Gradient Boosting Machine (GBM) algorithm was used to identify predictive variables. This was a retrospective study in 237 patients with BC. Blood samples were obtained pre- and post-RT, with samples of healthy women used as control subjects. Results showed that 24 patients had DP eight years post-RT, and eight patients developed a second primary tumor. The algorithm identified interleukin-4 and total lymphocyte counts as the most relevant indices discriminating between BC patients and control subjects, while neutrophils, total leukocytes, eosinophils, very low-density lipoprotein cholesterol, and PON1 activity were potential predictors of fatal outcome.

show abstract

Section: Discussionmentioning

confidence: 99%

Gradient Boosting Machine Identified Predictive Variables for Breast Cancer Patients Pre- and Post-Radiotherapy: Preliminary Results of an 8-Year Follow-Up Study

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Recently, Ma et al used ML to differentiate between breast cancer molecular subtypes based on mammography and ultrasound. Both clinical data and imaging signs based on the BI-RADS lexicon served as inputs for the ML models [37] . In another breast cancer study, perfusion MRI radiomics were used to infer tumor infiltrating lymphocytes [38] .…”

Section: Imaging Acquisition Optimizationmentioning

confidence: 99%

Artificial intelligence in oncologic imaging

Chen

Terzić²,

Becker

et al. 2022

European Journal of Radiology Open

View full text Add to dashboard Cite

“…Each subtype has different clinical characteristics, prognosis, and response to treatment. Different imaging features of each subtype are also reported in the literature [ 3 , 4 , 5 ].…”

Section: Introductionmentioning

confidence: 99%

Correlation of Shear-Wave Elastography and Apparent Diffusion Coefficient Values in Breast Cancer and Their Relationship with the Prognostic Factors

Örgüç

Açar

2022

Diagnostics

View full text Add to dashboard Cite

Background: Diffusion-weighted imaging and elastography are widely accepted methods in the evaluation of breast masses, however, there is very limited data comparing the two methods. The apparent diffusion coefficient is a measure of the diffusion of water molecules obtained by diffusion-weighted imaging as a part of breast MRI. Breast elastography is an adjunct to conventional ultrasonography, which provides a noninvasive evaluation of the stiffness of the lesion. Theoretically, increased tissue density and stiffness are related to each other. The purpose of this study is to compare MRI ADC values of the breast masses with quantitative elastography based on ultrasound shear wave measurements and to investigate their possible relation with the prognostic factors and molecular subtypes. Methods: We retrospectively evaluated histopathologically proven 147 breast lesions. The molecular classification of malignant lesions was made according to the prognostic factors. Shear wave elastography was measured in kiloPascal (kPa) units which is a quantitative measure of tissue stiffness. DWI was obtained using a 1.5-T MRI system. Results: ADC values were strongly inversely correlated with elasticity (r = −0.662, p < 0.01) according to Pearson Correlation. In our study, the cut-off value of ADC was 1.00 × 10−3 cm2/s to achieve a sensitivity of 84.6% and specificity of 75.4%, and the cut-off value of elasticity was 105.5 kPa to achieve the sensitivity of 96.3% and specificity 76.9% to discriminate between the malignant and benign breast lesions. The status of prognostic factors was not correlated with the ADC values and elasticity. Conclusions: Elasticity and ADC values are correlated. Both cannot predict the status of prognostic factors and differentiate between molecular subtypes.

show abstract

Predicting the molecular subtype of breast cancer and identifying interpretable imaging features using machine learning algorithms

Cited by 56 publications

References 38 publications

Gradient Boosting Machine Identified Predictive Variables for Breast Cancer Patients Pre- and Post-Radiotherapy: Preliminary Results of an 8-Year Follow-Up Study

Gradient Boosting Machine Identified Predictive Variables for Breast Cancer Patients Pre- and Post-Radiotherapy: Preliminary Results of an 8-Year Follow-Up Study

Artificial intelligence in oncologic imaging

Correlation of Shear-Wave Elastography and Apparent Diffusion Coefficient Values in Breast Cancer and Their Relationship with the Prognostic Factors

Contact Info

Product

Resources

About