RI is a powerful tool for diagnosis and screening of breast cancer (1). However, widespread use of breast MRI has been restricted because of the limited availability of sites that offer this method. One major reason for limited availability is the lack of radiologists who can offer substantial expertise in interpreting breast MR images. Sophisticated machine learning approaches show promise in complementing human diagnosis (2). Broadly speaking, machine learning can be divided into two major classes: one is radiomic analysis (RA), where handmade image features are extracted; and the other is the concept of convolutional neural networks (CNN), in which the computer learns to recognize image features on its own, usually on the basis of a set of labeled training examples. Both approaches have been pursued with considerable success for image interpretation, although in different areas: In the field of diagnostic radiology, RA has been successfully used to further classify tumor types (3,4). However, CNNs require a larger pool of training images before they achieve a clinically useful performance. Within radiology, breast imaging, specifically mammographic screening, lends itself to be used with CNNs because similarly large data sets are available (5,6). With such large mammographic data sets, and with the advent of increased computing power, deep learning may have the potential to outperform regular computer-assisted diagnosis systems for mammographic interpretation (5). Studies are limited regarding the use of RA or CNNs for diagnostic classification of contrast agent-enhancing breast lesions (ie, for differential diagnosis of benign vs malignant lesions). Bickelhaupt et al (7) used machine learning for further characterization of lesions suspicious for cancer that were found on digital mammographic images and used unenhanced and diffusion-weighted MRI for this purpose. However, the use of RA or CNNs for classification of enhancing lesions observed at regular, clinical, dynamic contrast agent-enhanced, or multiparametric breast MRI is not established. Because breast MRI is performed less than mammographic screening, available breast MRI data sets are smaller

No abstract

identifying image features that are robust with respect to segmentation variability is a tough challenge in radiomics. So far, this problem has mainly been tackled in test-retest analyses. in this work we analyse radiomics feature reproducibility in two phases: first with manual segmentations provided by four expert readers and second with probabilistic automated segmentations using a recently developed neural network (pHiseg). We test feature reproducibility on three publicly available datasets of lung, kidney and liver lesions. We find consistent results both over manual and automated segmentations in all three datasets and show that there are subsets of radiomic features which are robust against segmentation variability and other radiomic features which are prone to poor reproducibility under differing segmentations. By providing a detailed analysis of robustness of the most common radiomics features across several datasets, we envision that more reliable and reproducible radiomic models can be built in the future based on this work.

Computer vision (CV) has the potential to change medicine fundamentally. Expert knowledge provided by CV can enhance diagnosis. Unfortunately, existing algorithms often remain below expectations, as databases used for training are usually too small, incomplete, and heterogeneous in quality. Moreover, data protection is a serious obstacle to the exchange of data. To overcome this limitation, we propose to use generative models (GMs) to produce high-resolution synthetic radiographs that do not contain any personal identification information. Blinded analyses by CV and radiology experts confirmed the high similarity of synthesized and real radiographs. The combination of pooled GM improves the performance of CV algorithms trained on smaller datasets, and the integration of synthesized data into patient data repositories can compensate for underrepresented disease entities. By integrating federated learning strategies, even hospitals with few datasets can contribute to and benefit from GM training.

Prediction of overall survival based on multimodal MRI of brain tumor patients is a difficult problem. Although survival also depends on factors that cannot be assessed via preoperative MRI such as surgical outcome, encouraging results for MRI-based survival analysis have been published for different datasets. We assess if and how established radiomic approaches as well as novel methods can predict overall survival of brain tumor patients on the BraTS challenge dataset. This dataset consists of multimodal preoperative images of 211 glioblastoma patients from several institutions with reported resection status and known survival. In the official challenge setting, only patients with a reported gross total resection (GTR) are taken into account. We therefore evaluated previously published methods as well as different machine learning approaches on the BraTS dataset. For different types of resection status, these approaches are compared to a baseline, a linear regression on patient age only. This naive approach won the 3rd place out of 26 participants in the BraTS survival prediction challenge 2018. Previously published radiomic signatures show significant correlations and predictiveness to patient survival for patients with a reported subtotal resection. However, for patients with reported GTR, none of the evaluated approaches was able to outperform the age-only baseline in a cross-validation setting, explaining the poor performance of approaches based on radiomics in the BraTS challenge 2018.