G adoxetic acid-enhanced MRI is used to depict and help characterize focal liver nodules (1,2) in patients with chronic liver diseases (CLDs) (3,4), including nonalcoholic steatohepatitis (5) and chronic hepatitis C (5,6). Gadoxetic acid-enhanced MRI has been shown to help predict both liver failure after subtotal hepatectomy and graft survival after liver transplant (7-9).As laboratory and clinical estimators of liver disease severity, the albumin-bilirubin index, the Model for End-Stage Liver Disease, and the Child-Turcotte-Pugh score correlate well with gadoxetic acid uptake in the liver in the hepatobiliary phase (ie, 20 minutes after contrast agent administration of gadoxetic acid) (10,11). Previously described methods to assess hepatobiliary phase uptake include the relative liver enhancement, the hepatic uptake index, the contrast enhancement index, and T1 values (12). These methods all require complex computations and have vendor, field-strength, and sequence dependencies that complicate their clinical application.Recently, Bastati et al ( 13) introduced the functional liver imaging score (FLIS), derived from the three hepatobiliary phase features of gadoxetic acid-enhanced MRI and each scored on an ordinal 0-2 scale. The three features included in the FLIS semiquantitatively assess the enhancement

Multiparametric prostate magnetic resonance imaging (mpMRI) is widely used as a triage test for men at a risk of prostate cancer. However, the traditional role of mpMRI was confined to prostate cancer staging. Radiomics is the quantitative extraction and analysis of minable data from medical images; it is emerging as a promising tool to detect and categorize prostate lesions. In this paper we review the role of radiomics applied to prostate mpMRI in detection and localization of prostate cancer, prediction of Gleason score and PI-RADS classification, prediction of extracapsular extension and of biochemical recurrence. We also provide a future perspective of artificial intelligence (machine learning and deep learning) applied to the field of prostate cancer.

Laparoscopic cholecystectomy (LC) is the standard technique for treatment of gallbladder disease. In case of acute cholecystitis we can identify preoperative factors associated with an increased risk of conversion and intraoperative complications. The aim of our study was to detect preoperative laboratory and radiological findings predictive of difficult LC with potential advantages for both the surgeons and patients in terms of options for management. We designed a retrospective case–control study to compare preoperative predictive factors of difficult LC in patients treated in emergency setting between January 2015 and December 2019. We included in the difficult LC group the surgeries with operative time > 2 h, need for conversion to open, significant bleeding and/or use of synthetic hemostats, vascular and/or biliary injuries and additional operative procedures. We collected 86 patients with inclusion criteria and difficult LC. In the control group, we selected 86 patients with inclusion criteria, but with no operative signs of difficult LC. The analysis of the collected data showed that there was a statistically significant association between WBC count and fibrinogen level and difficult LC. No association were seen with ALP, ALT and bilirubin values. Regarding radiological findings significant differences were noted among the two groups for irregular or absent wall, pericholecystic fluid, fat hyperdensity, thickening of wall > 4 mm and hydrops. The preoperative identification of difficult laparoscopic cholecystectomy provides an important advantage not only for the surgeon who has to perform the surgery, but also for the organization of the operating block and technical resources. In patients with clinical and laboratory parameters of acute cholecystitis, therefore, it would be advisable to carry out a preoperative abdominal CT scan with evaluation of features that can be easily assessed also by the surgeon.

Magnetic Resonance Imaging-based prostate segmentation is an essential task for adaptive radiotherapy and for radiomics studies whose purpose is to identify associations between imaging features and patient outcomes. Because manual delineation is a time-consuming task, we present three deep-learning (DL) approaches, namely UNet, efficient neural network (ENet), and efficient residual factorized convNet (ERFNet), whose aim is to tackle the fully-automated, real-time, and 3D delineation process of the prostate gland on T2-weighted MRI. While UNet is used in many biomedical image delineation applications, ENet and ERFNet are mainly applied in self-driving cars to compensate for limited hardware availability while still achieving accurate segmentation. We apply these models to a limited set of 85 manual prostate segmentations using the k-fold validation strategy and the Tversky loss function and we compare their results. We find that ENet and UNet are more accurate than ERFNet, with ENet much faster than UNet. Specifically, ENet obtains a dice similarity coefficient of 90.89% and a segmentation time of about 6 s using central processing unit (CPU) hardware to simulate real clinical conditions where graphics processing unit (GPU) is not always available. In conclusion, ENet could be efficiently applied for prostate delineation even in small image training datasets with potential benefit for patient management personalization.

Newer biologic drugs and immunomodulatory agents, as well as more tolerated and effective radiation therapy schemes, have reduced treatment toxicity in oncology patients. However, although imaging assessment of tumor response is adapting to atypical responses like tumor flare, expected changes and complications of chemo/radiotherapy are still routinely encountered in post-treatment imaging examinations. Radiologists must be aware of old and newer therapeutic options and related side effects or complications to avoid a misinterpretation of imaging findings. Further, advancements in oncology research have increased life expectancy of patients as well as the frequency of long-term therapy-related side effects that once could not be observed. This pictorial will help radiologists tasked to detect therapy-related complications and to differentiate expected changes of normal tissues from tumor relapse.

No abstract

No abstract

In healthcare industry 4.0, a big role is played by radiomics. Radiomics concerns the extraction and analysis of quantitative information not visible to the naked eye, even by expert operators, from biomedical images. Radiomics involves the management of digital images as data matrices, with the aim of extracting a number of morphological and predictive variables, named features, using automatic or semi‐automatic methods. Multidisciplinary methods as machine learning and deep learning are fully involved in this field. However, the large number of features requires efficient and effective core methods for their selection, in order to avoid bias or misinterpretations problems. In this work, the authors propose a novel method for feature selection in radiomics. The proposed method is based on an original combination of descriptive and inferential statistics. Its validity is illustrated through a case study on prostate cancer analysis, conducted at the university hospital of Palermo, Italy.