In this paper, we present and discuss a novel reliability metric to quantify the extent a ground truth, generated in multi-rater settings, as a reliable basis for the training and validation of machine learning predictive models. To define this metric, three dimensions are taken into account: agreement (that is, how much a group of raters mutually agree on a single case); confidence (that is, how much a rater is certain of each rating expressed); and competence (that is, how accurate a rater is). Therefore, this metric produces a reliability score weighted for the raters’ confidence and competence, but it only requires the former information to be actually collected, as the latter can be obtained by the ratings themselves, if no further information is available. We found that our proposal was both more conservative and robust to known paradoxes than other existing agreement measures, by virtue of a more articulated notion of the agreement due to chance, which was based on an empirical estimation of the reliability of the single raters involved. We discuss the above metric within a realistic annotation task that involved 13 expert radiologists in labeling the MRNet dataset. We also provide a nomogram by which to assess the actual accuracy of a classification model, given the reliability of its ground truth. In this respect, we also make the point that theoretical estimates of model performance are consistently overestimated if ground truth reliability is not properly taken into account.

Magnetic resonance imaging (MRI) is the gold-standard technique for evaluating muscle fatty infiltration and muscle atrophy due to its high contrast resolution. It can differentiate muscular from adipose tissue accurately. MRI can also quantify the adipose content within muscle bellies with several sequences such as T₁-mapping, T₂-mapping, spectroscopy, Dixon, intra-voxel incoherent motion, and diffusion tensor imaging. The main fields of interest in musculoskeletal radiology for a quantitative MRI evaluation of muscular fatty infiltration include neuro-muscular disorders such as myopathies, and dystrophies. Sarcopenia is another important field in which the evaluation of the degree of muscular fat infiltration or muscular hypotrophy is required for a correct diagnosis. This review highlights several MRI techniques and sequences focusing on quantitative methods of assessing adipose tissue and muscle atrophy.

Groin pain syndrome (GPS) is one of the most frequent injuries in competitive sports. Stresses generated in the lower limbs by quick turns and accelerations, such as in soccer, basketball or hockey, can produce localized regions of increased forces, resulting in anatomical lesions. The differential diagnoses are numerous and comprise articular, extra-articular, muscular, tendinous and visceral clinical conditions and a correct diagnosis is crucial if treatment is to be efficient. MRI is the gold standard of diagnostic techniques, especially when an alternative pathology needs to be excluded and/or other imaging techniques such as ultrasound or radiography do not lead to a diagnosis. This paper, based on the current literature, gives a comprehensive review of the anatomy of the pubic region and of the typical MRI findings in those affected by GPS. Many clinical conditions causing GPS can be investigated by MRI within appropriate protocols. However, MRI shows limits in reliability in the investigation of inguinal and femoral hernias and therefore is not the imaging technique of choice for studying these clinical conditions.