Milana Flusberg

LI-RADS® ancillary features on CT and MRI

The Liver Imaging Reporting and Data System (LI-RADS) uses an algorithm to assign categories that reflect the probability of hepatocellular carcinoma (HCC), non-HCC malignancy, or benignity. Unlike other imaging algorithms, LI-RADS utilizes ancillary features (AFs) to refine the final category. AFs in LI-RADS v2017 are divided into those favoring malignancy in general, those favoring HCC specifically, and those favoring benignity. Additionally, LI-RADS v2017 provides new rules regarding application of AFs. The purpose of this review is to discuss ancillary features included in LI-RADS v2017, the rationale for their use, potential pitfalls encountered in their interpretation, and tips on their application.

Transient arterial phase respiratory motion-related artifact in MR imaging of the liver: an analysis of four different gadolinium-based contrast agents

PURPOSE The purpose was to compare hepatic arterial phase (HAP) respiratory motion artifact (RMA) between gadoxetate, gadobutrol, gadopentetate, and gadobenate. MATERIALS/METHODS Two hundred cases of each gadolinium agent were included. RMA was assigned using 5-point Likert scale (1=no motion, 5=extreme motion) on precontrast and HAP. RMA increase (increase ≥1 on HAP from precontrast) was the outcome in logistic regression. RESULTS Odds of RMA increase for gadoxetate were 5.5 (P<.001), 3.6 (P=.034), and 9.5 (P<.001) times higher than gadobutrol, gadopentetate, and gadobenate, respectively. Gadolinium volume and dose were not independent predictors of RMA increase. CONCLUSION Gadoxetate has increased odds of RMA compared with other gadolinium agents; tight contrast bolus is not a contributor.

Pancreatic cysts: What imaging characteristics are associated with development of pancreatic ductal adenocarcinoma?

PURPOSE To assess relationship between pancreatic cysts (PC) and pancreatic ductal adenocarcinoma (PDAC) and to compare imaging features of PC in subjects who develop PDAC and those who do not. MATERIAL AND METHODS This case-control IRB-approved HIPAA-compliant study included patients with PDAC (cases) and lung cancer (controls), diagnosed between 1/1/05-1/1/14. The most recent abdominal CT/MR of each patient done >6months prior to cancer diagnosis date was reviewed by radiologist blinded to case/control status. Presence of PC, number and size of largest PC, presence of main pancreatic duct (MPD) dilatation (MPD >3mm) were recorded. Simple PC was defined as lack of calcifications, septations and enhancement. Logistic regression models with binary outcome of PDAC were constructed. RESULTS There were 88 cases and 273 controls with mean ages of 69.6 years (±10.3) and 69.8 years (±12.0), respectively (p=0.864). PC were present in 21.6% (19/88) cases and in 9.2% (25/273) controls (p=0.002). The OR of PC for development of PDAC was 2.83, adjusting for age, sex and race (p=0.001). Mean PC size was 14.8mm (±8.7) in cases and 7.6mm (±8.0) in controls (p=0.007). PC were solitary in 6 (31.6%) of 19 cases and 21 (84.0%) of 25 controls (p=0.001). There was no significant difference in proportions of simple cysts or MPD dilatation between cases and controls. Multiple PC had 8.2 times increased odds of PDAC compared with solitary PC, adjusting for cyst size (p=0.007). CONCLUSION Multiple PC are associated with 8.2 times higher odds of PDAC compared with solitary PC.

Practical guide to dynamic pelvic floor MRI

Pelvic floor dysfunction encompasses a spectrum of functional disorders that result from impairment of the ligaments, fasciae, and muscles supporting the pelvic organs. It is a prevalent disorder that carries a lifetime risk over 10% for undergoing a surgical repair. Pelvic floor weakness presents as a wide range of symptoms, including pain, pelvic pressure or bulging, urinary and fecal incontinence, constipation, and sexual dysfunction. A correct diagnosis by clinical examination alone can be challenging, particularly in cases involving multiple compartments. Magnetic resonance imaging (MRI) allows noninvasive, radiation‐free, high soft‐tissue resolution evaluation of all three pelvic compartments, and has proved a reliable technique for accurate diagnosis of pelvic floor dysfunction. MR defecography with steady‐state sequences allows detailed anatomic and functional evaluation of the pelvic floor. This article provides an overview of normal anatomy and function of the pelvic floor and discusses a practical approach to the evaluation of imaging findings of pelvic floor relaxation, pelvic organ prolapse, fecal incontinence, and obstructed defecation.

Accuracy of prostate measurements on MRI with and without an endorectal coil.

PURPOSE To assess accuracy of prostate measurements with and without endorectal coil (ERC). MATERIALS AND METHODS Anteroposterior (AP), transverse (TX) and craniocaudal (CC) measurements were recorded from 49 prostate magnetic resonance images (MRIs) done both with and without ERC. Prostate weight was calculated as follows: AP*TX*CC*π/6. Prostate dimensions and weight were obtained from radical prostatectomy pathology report. RESULTS After ERC placement, AP decreased by 0.71 cm [95% confidence interval (CI) 0.80-0.61], TX and CC increased by 0.26 cm (95% CI 0.18-0.33) and 0.25 cm (95% CI 0.16-0.35), respectively. Agreement between weight on pathology and MR was excellent: intraclass correlation coefficient (ICC) without ERC=0.96, ICC with ERC=0.90. CONCLUSION Although ERC distorts measurements and despite a tendency to underestimate the prostate weight, absolute agreement between prostate weight on pathology and MRI is excellent, both with and without ERC.

Patients 65 years and older with incidental pancreatic cysts: Is there a relationship between all-cause mortality and imaging follow-up?

PURPOSE To assess the relationship between imaging follow-up and all-cause mortality in subjects ≥65 years with and without incidental pancreatic cysts (IPC). METHODS AND MATERIALS Patients ≥65 years with abdominal CT/MR 11/1/01-11/1/11 were included. IPC group included subjects with IPC on CT/MR report; No-IPC group was 3:1 frequency-matched on age decade, imaging modality and year of initial study from the pool without reported IPC. Demographics, date of last encounter, date of death, Charlson scores within 3 months before initial CT/MR and number of abdominal CTs and MRs performed after initial study were recorded. Logistic regression models with binary outcomes of death and having post-index imaging were constructed. Models were adjusted for age, race, sex, Charlson score and follow-up time. Subgroups were created based on interactions between variables. RESULTS There were 1320 subjects in IPC group and 3805 in No-IPC group, with mean ages 79.1 (±8.0) and 78.8 (±8.0) years, respectively (p=0.293), and median follow-up times of 3.1 (IQR 0.74-5.26) and 3.0 (0.36-5.23) years, respectively (p=0.009). Adjusted odds ratios of post-index imaging for IPC were 2.18 (p<0.001) in subgroup<84years and follow-up <4years, 3.37 (p<0.001) in subgroup <84 years and follow-up ≥4 years, and 1.20 (p=0.201) in subgroup ≥84 years. Number of follow-up CTs and MRs was not independently associated with decreased odds of death in any subgroup. CONCLUSION Older subjects with IPC are more likely to undergo imaging follow-up compared to subjects without IPC, yet increasing number of follow-up studies does not decrease the odds of death.

MR Imaging of Acute Abdomen and Pelvis

Magnetic resonance imaging (MRI), although not usually the first diagnostic study in assessment of acute abdominal pain, can offer a radiation-free high tissue-contrast resolution alternative to CT. MRI can be particularly useful in assessment of young or pregnant patients, or patients with renal insufficiency precluding administration of intravenous contrast. In this article, we review MRI appearance of a number of acute abdominal and pelvic processes, including hemorrhage and various ischemic, inflammatory, and infectious conditions.

Outside the Inside: A Review of Soft-Tissue Abnormalities Seen on Thoracoabdominal Computed Tomography

In this review, we illustrate the computed tomographic features of thoracoabdominal soft-tissue abnormalities, which may be easily overlooked and often can provide important information regarding systemic processes. Examples include necrotizing fasciitis, heterotopic ossification, fat necrosis, benign and malignant neoplasms, endometriosis, and collagen vascular disease as well as systemic and congenital pathology.