The Image Biomarker Standardization Initiative validated consensus-based reference values for 169 radiomics features, thus enabling calibration and verification of radiomics software. Key results: • research teams found agreement for calculation of 169 radiomics features derived from a digital phantom and a human lung cancer on CT scan. • Of these 169 candidate radiomics features, good to excellent reproducibility was achieved for 167 radiomics features using MRI, 18F-FDG PET and CT images obtained in 51 patients with soft-tissue sarcoma.

Delineation of the left ventricular cavity, myocardium, and right ventricle from cardiac magnetic resonance images (multi-slice 2-D cine MRI) is a common clinical task to establish diagnosis. The automation of the corresponding tasks has thus been the subject of intense research over the past decades. In this paper, we introduce the "Automatic Cardiac Diagnosis Challenge" dataset (ACDC), the largest publicly available and fully annotated dataset for the purpose of cardiac MRI (CMR) assessment. The dataset contains data from 150 multi-equipments CMRI recordings with reference measurements and classification from two medical experts. The overarching objective of this paper is to measure how far state-of-the-art deep learning methods can go at assessing CMRI, i.e., segmenting the myocardium and the two ventricles as well as classifying pathologies. In the wake of the 2017 MICCAI-ACDC challenge, we report results from deep learning methods provided by nine research groups for the segmentation task and four groups for the classification task. Results show that the best methods faithfully reproduce the expert analysis, leading to a mean value of 0.97 correlation score for the automatic extraction of clinical indices and an accuracy of 0.96 for automatic diagnosis. These results clearly open the door to highly accurate and fully automatic analysis of cardiac CMRI. We also identify scenarios for which deep learning methods are still failing. Both the dataset and detailed results are publicly available online, while the platform will remain open for new submissions.

Our results indicate that retrospective techniques have the potential to produce satisfactory results much of the time, but that visual inspection is necessary to guard against large errors.

This radiogenomics strategy for identifying imaging biomarkers may enable a more rapid evaluation of novel imaging modalities, thereby accelerating their translation to personalized medicine.

Radiomics describes a broad set of computational methods that extract quantitative features from radiographic images. The resulting features can be used to inform imaging diagnosis, prognosis, and therapy response in oncology. However, major challenges remain for methodological developments to optimize feature extraction and provide rapid information flow in clinical settings. Equally important, to be clinically useful, predictive radiomic properties must be clearly linked to meaningful biological characteristics as well as qualitative imaging properties familiar to radiologist. Here we use a cross-disciplinary approach to highlight studies in radiomics. We review brain tumor radiological studies (e.g., imaging interpretation) through computational models (e.g., computer vision and machine learning) that provide novel clinical insights. In particular, we outline current quantitative image feature extraction and prediction strategies with different level of available clinical classes for supporting clinical decision making. We further discuss machine learning challenges and data opportunities to advance radiomic studies.

Measuring the properties of the white matter pathways from retina to cortex in the living human brain will have many uses for understanding visual performance and guiding clinical treatment. For example, identifying the Meyer’s loop portion of the optic radiation (OR) has clinical significance because of the large number of temporal lobe resections. We use diffusion tensor imaging and fiber tractography (DTI-FT) to identify the most likely pathway between the lateral geniculate nucleus (LGN) and the calcarine sulcus in sixteen hemispheres of eight healthy volunteers. Quantitative population comparisons between DTI-FT estimates and published postmortem dissections match with a spatial precision of about 1 mm. The OR can be divided into three bundles that are segmented based on the direction of the fibers as they leave the LGN: Meyer’s loop, central, and direct. The longitudinal and radial diffusivities of the three bundles do not differ within the measurement noise; there is a small difference in the radial diffusivity between the right and left hemispheres. We find that the anterior tip of Meyer’s loop is 28 ± 3 mm posterior to the temporal pole, and the population range is 1 cm. Hence, it is important to identify the location of this bundle in individual subjects or patients.

Diagnostic radiology requires accurate interpretation of complex signals in medical images. Content-based image retrieval (CBIR) techniques could be valuable to radiologists in assessing medical images by identifying similar images in large archives that could assist with decision support. Many advances have occurred in CBIR, and a variety of systems have appeared in nonmedical domains; however, permeation of these methods into radiology has been limited. Our goal in this review is to survey CBIR methods and systems from the perspective of application to radiology and to identify approaches developed in nonmedical applications that could be translated to radiology. Radiology images pose specific challenges compared with images in the consumer domain; they contain varied, rich, and often subtle features that need to be recognized in assessing image similarity. Radiology images also provide rich opportunities for CBIR: rich metadata about image semantics are provided by radiologists, and this information is not yet being used to its fullest advantage in CBIR systems. By integrating pixel-based and metadata-based image feature analysis, substantial advances of CBIR in medicine could ensue, with CBIR systems becoming an important tool in radiology practice.

Magnetic resonance diffusion-weighted imaging coupled with fiber tractography (DFT) is the only non-invasive method for measuring white matter pathways in the living human brain. DFT is often used to discover new pathways. But there are also many applications, particularly in visual neuroscience, in which we are confident that two brain regions are connected, and we wish to find the most likely pathway forming the connection. In several cases, current DFT algorithms fail to find these candidate pathways. To overcome this limitation, we have developed a probabilistic DFT algorithm (ConTrack) that identifies the most likely pathways between two regions. We introduce the algorithm in three parts: a sampler to generate a large set of potential pathways, a scoring algorithm that measures the likelihood of a pathway, and an inferential step to identify the most likely pathways connecting two regions. In a series of experiments using human data, we show that ConTrack estimates known pathways at positions that are consistent with those found using a high quality deterministic algorithm. Further we show that separating sampling and scoring enables ConTrack to identify valid pathways, known to exist, that are missed by other deterministic and probabilistic DFT algorithms.