The purpose of this multicenter retrospective study was to evaluate radiomics analysis coupled with machine learning (ML) of dynamic contrast-enhanced (DCE) and diffusion-weighted imaging (DWI) radiomics models separately and combined as multiparametric MRI for improved breast cancer detection. Consecutive patients (Memorial Sloan Kettering Cancer Center, January 2018–March 2020; Medical University Vienna, from January 2011–August 2014) with a suspicious enhancing breast tumor on breast MRI categorized as BI-RADS 4 and who subsequently underwent image-guided biopsy were included. In 93 patients (mean age: 49 years ± 12 years; 100% women), there were 104 lesions (mean size: 22.8 mm; range: 7–99 mm), 46 malignant and 58 benign. Radiomics features were calculated. Subsequently, the five most significant features were fitted into multivariable modeling to produce a robust ML model for discriminating between benign and malignant lesions. A medium Gaussian support vector machine (SVM) model with five-fold cross validation was developed for each modality. A model based on DWI-extracted features achieved an AUC of 0.79 (95% CI: 0.70–0.88), whereas a model based on DCE-extracted features yielded an AUC of 0.83 (95% CI: 0.75–0.91). A multiparametric radiomics model combining DCE- and DWI-extracted features showed the best AUC (0.85; 95% CI: 0.77–0.92) and diagnostic accuracy (81.7%; 95% CI: 73.0–88.6). In conclusion, radiomics analysis coupled with ML of multiparametric MRI allows an improved evaluation of suspicious enhancing breast tumors recommended for biopsy on clinical breast MRI, facilitating accurate breast cancer diagnosis while reducing unnecessary benign breast biopsies.

Background: This study investigated the performance of simultaneous ¹⁸F-FDG PET/MRI of the breast as a platform for comprehensive radiomics analysis for breast cancer subtype analysis, hormone receptor status, proliferation rate and lymphonodular and distant metastatic spread. Methods: One hundred and twenty-four patients underwent simultaneous ¹⁸F-FDG PET/MRI. Breast tumors were segmented and radiomic features were extracted utilizing CERR software following the IBSI guidelines. LASSO regression was employed to select the most important radiomics features prior to model development. Five-fold cross validation was then utilized alongside support vector machines, resulting in predictive models for various combinations of imaging data series. Results: The highest AUC and accuracy for differentiation between luminal A and B was achieved by all MR sequences (AUC 0.98; accuracy 97.3). The best results in AUC for prediction of hormone receptor status and proliferation rate were found based on all MR and PET data (ER AUC 0.87, PR AUC 0.88, Ki-67 AUC 0.997). PET provided the best determination of grading (AUC 0.71), while all MR and PET analyses yielded the best results for lymphonodular and distant metastatic spread (0.81 and 0.99, respectively). Conclusion: ¹⁸F-FDG PET/MRI enables comprehensive high-quality radiomics analysis for breast cancer phenotyping and tumor decoding, utilizing the perks of simultaneously acquired morphologic, functional and metabolic data.

Background: The aim of this study was to assess whether multiparametric ¹⁸F-FDG PET/MRI-based radiomics analysis is able to predict pathological complete response in breast cancer patients and hence potentially enhance pretherapeutic patient stratification. Methods: A total of 73 female patients (mean age 49 years; range 27–77 years) with newly diagnosed, therapy-naive breast cancer underwent simultaneous ¹⁸F-FDG PET/MRI and were included in this retrospective study. All PET/MRI datasets were imported to dedicated software (ITK-SNAP v. 3.6.0) for lesion annotation using a semi-automated method. Pretreatment biopsy specimens were used to determine tumor histology, tumor and nuclear grades, and immunohistochemical status. Histopathological results from surgical tumor specimens were used as the reference standard to distinguish between complete pathological response (pCR) and noncomplete pathological response. An elastic net was employed to select the most important radiomic features prior to model development. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated for each model. Results: The best results in terms of AUCs and NPV for predicting complete pathological response in the entire cohort were obtained by the combination of all MR sequences and PET (0.8 and 79.5%, respectively), and no significant differences from the other models were observed. In further subgroup analyses, combining all MR and PET data, the best AUC (0.94) for predicting complete pathologic response was obtained in the HR+/HER2− group. No difference between results with/without the inclusion of PET characteristics was observed in the TN/HER2+ group, each leading to an AUC of 0.92 for all MR and all MR + PET datasets. Conclusion: ¹⁸F-FDG PET/MRI enables comprehensive high-quality radiomics analysis for the prediction of pCR in breast cancer patients, especially in those with HR+/HER2− receptor status.

Biopsy is the standard for assessment of bone marrow involvement in mantle cell lymphoma (MCL). We investigated whether [18F]FDG-PET radiomic texture features can improve prediction of bone marrow involvement in MCL, compared to standardized uptake values (SUV), and whether combination with laboratory data improves results. Ninety-seven MCL patients were retrospectively included. SUVmax, SUVmean, SUVpeak and 16 co-occurrence matrix texture features were extracted from pelvic bones on [18F]FDG-PET/CT. A multi-layer perceptron neural network was used to compare three combinations for prediction of bone marrow involvement—the SUVs, a radiomic signature based on SUVs and texture features, and the radiomic signature combined with laboratory parameters. This step was repeated using two cut-off values for relative bone marrow involvement: REL > 5% (>5% of red/cellular bone marrow); and REL > 10%. Biopsy demonstrated bone marrow involvement in 67/97 patients (69.1%). SUVs, the radiomic signature, and the radiomic signature with laboratory data showed AUCs of up to 0.66, 0.73, and 0.81 for involved vs. uninvolved bone marrow; 0.68, 0.84, and 0.84 for REL ≤ 5% vs. REL > 5%; and 0.69, 0.85, and 0.87 for REL ≤ 10% vs. REL > 10%. In conclusion, [18F]FDG-PET texture features improve SUV-based prediction of bone marrow involvement in MCL. The results may be further improved by combination with laboratory parameters.

This multicenter retrospective study compared the performance of radiomics analysis coupled with machine learning (ML) with that of radiologists for the classification of breast tumors. A total of 93 consecutive women (mean age: 49 ± 12 years) with 104 histopathologically verified enhancing lesions (mean size: 22.8 ± 15.1 mm), classified as suspicious on multiparametric breast MRIs were included. Two experienced breast radiologists assessed all of the lesions, assigning a Breast Imaging Reporting and Database System (BI-RADS) suspicion category, providing a diffusion-weighted imaging (DWI) score based on lesion signal intensity, and determining the apparent diffusion coefficient (ADC). Ten predictive models for breast lesion discrimination were generated using radiomic features extracted from the multiparametric MRI. The area under the receiver operating curve (AUC) and the accuracy were compared using McNemar’s test. Multiparametric radiomics with DWI score and BI-RADS (accuracy = 88.5%; AUC = 0.93) and multiparametric radiomics with ADC values and BI-RADS (accuracy= 88.5%; AUC = 0.96) models showed significant improvements in diagnostic accuracy compared to the multiparametric radiomics (DWI + DCE data) model (p = 0.01 and p = 0.02, respectively), but performed similarly compared to the multiparametric assessment by radiologists (accuracy = 85.6%; AUC = 0.03; p = 0.39). In conclusion, radiomics analysis coupled with the ML of multiparametric MRI could assist in breast lesion discrimination, especially for less experienced readers of breast MRIs.