Search

The use of artificial intelligence (AI) algorithms in the field of radiology is becoming more common. Several studies have demonstrated the potential utility of machine learning (ML) and deep learning (DL) techniques as aids for radiologists to solve specific radiological challenges. The decision-making process, the establishment of specific clinical or radiological targets, the profile of the different professionals involved in the development of AI solutions, and the relation with partnerships and stakeholders are only some of the main issues that have to be faced and solved prior to starting the development of radiological AI solutions. Among all the players in this multidisciplinary team, the communication between radiologists and data scientists is essential for a successful collaborative work. There are specific skills that are inherent to radiological and medical training that are critical for identifying anatomical or clinical targets as well as for segmenting or labelling lesions. These skills would then have to be transferred, explained, and taught to the data science experts to facilitate their comprehension and integration into ML or DL algorithms. On the other hand, there is a wide range of complex software packages, deep neural-network architectures, and data transfer processes for which radiologists need the expertise of software engineers and data scientists in order to select the optimal manner to analyse and post-process this amount of data. This paper offers a summary of the top five challenges faced by radiologists and data scientists including tips and tricks to build a successful AI team., (Copyright © 2020 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.)

Background and Purpose: While third ventricular colloid cysts may present as an incidental finding, they also harbor the potential to cause ventricular obstruction and sudden death. Herein we analyze the relationship between imaging appearance and the risk of obstructive ventriculomegaly., Materials and Methods: This is a retrospective review of the MR imaging appearance of 64 patients with colloid cysts, 46 of whom also had a CT scan, obtained by a tertiary hospital imaging report data base search over a 10-year period. Cysts were categorized by appearance on T2-FLAIR and correlated with patient age, cyst size, and the risk of obstructive ventriculomegaly. Histopathologic correlation was available for 28 cases., Results: The 64-patient cohort was 52% female, median age 50 years (range 10 to 99 years). Cysts hyperintense on T2-FLAIR (53.1%) were larger ( P <.001), occurred in younger patients ( P = .01), and had a higher risk of obstructive ventriculomegaly than homogeneously hypointense cysts (relative risk 6.18, 95% CI [2.04, 18.67]). Three patterns of T2 hyperintensity were identified: homogeneously hyperintense, hyperintense rim, and cysts with "dot sign." Although "dot sign" cysts were larger ( P < .001), there was no significant difference in patient age or the risk of ventricular obstruction among T2 hyperintense cysts. Cyst wall histopathology did not vary with imaging appearance., Conclusions: Hyperintensity on T2-FLAIR, whether homogeneous, rim, or "dot sign," is associated with larger cyst size and younger patient age, and is an imaging risk factor for obstructive ventriculomegaly. The hyperintense rim does not represent a thickened cyst wall., (© 2020 by American Journal of Neuroradiology.)

Background and Purpose: Clinical and imaging manifestations of idiopathic intracranial hypertension should prompt early diagnosis and treatment to avoid complications. Multiple diagnostic imaging criteria are reported to suggest the diagnosis of idiopathic intracranial hypertension with questionable sensitivity and/or specificity. Increased intracranial pressure results in dilation of the perineural cisternal spaces such as the optic nerve sheaths and the Meckel cave. It may also cause protrusion of cisternal structures of the Meckel cave through the skull base foramina, which could result in indentation or a bilobed appearance of the Meckel cave. We investigated the changes in the Meckel cave in patients with proved idiopathic intracranial hypertension versus healthy controls., Materials and Methods: We studied 75 patients with a diagnosis of idiopathic intracranial hypertension and 75 age-and sex-matched healthy controls. The transverse diameter of Meckel cave was measured in the axial and coronal planes of T2-weighted MR imaging sequences, and comparison was made between the 2 groups., Results: The mean diameters of the Meckel cave on the coronal T2 plane in patients with idiopathic intracranial hypertension were 5.21  ± 1.22 mm on the right side and 5.16  ± 0.90 mm on the left side, while in the control group, they measured 3.89  ± 0.62 mm and 4.09  ± 0.68 mm, respectively ( P value < .001). Of 75 patients with an approved diagnosis of idiopathic intracranial hypertension, 57 (76%) showed an indented Meckel cave as opposed to 21 (28%) in the control group., Conclusions: Our results confirm for the first time that the shape and size of the Meckel cave can be used as sensitive and specific diagnostic imaging markers for the diagnosis of idiopathic intracranial hypertension., (© 2020 by American Journal of Neuroradiology.)