Journal highlights

CT is routinely used to evaluate liver metastases, a critical task that impacts patient outcomes. But conventional energy-integrating detector (EID) CT has limitations, especially in detecting small liver lesions (<1.5 cm). More precise imaging methods are needed to help assess resectability and monitor therapeutic response.

Photon-counting CT (PCCT) represents a breakthrough in CT imaging technology. PCCT improves spatial resolution, reduces image noise and requires lower radiation doses and less iodinated contrast material than EID CT. However, clinical validation often lags behind technological advancements.

A recent Radiology study compared PCCT with EID CT using a virtual imaging framework. Researchers led by Benjamin Wildman-Tobriner, MD, Duke University Hospital in Durham, NC, analyzed 50 phantoms with 183 simulated liver lesions scanned at routine and low-dose levels.

The results showed that PCCT yielded better sensitivity in detecting small liver lesions. Radiologists also reported greater confidence and better image quality.

“As PCCT becomes more widely available, radiologists will be able to take full advantage of its potential for dose reductions across a variety of imaging indications,” the authors write.

Read the full article, “Photon-Counting CT Effects on Sensitivity for Liver Lesion Detection: A Reader Study Using Virtual Imaging,” and the related editorial, “Photon-Counting CT: Virtual Study, Real Benefit.”

Follow the Radiology editor on X @RadiologyEditor.

For patients with neurologic conditions, MRI is commonly performed using 1.5T and 3T systems. Lower-field-strength MRI systems, common in the early 1980s, were phased out due to lower image quality, longer imaging times and reduced efficiency. However, recent software and hardware advances, along with a push for greater accessibility to MRI, have sparked renewed interest in MRI systems with field strengths ranging from 0.064 T to 1.0 T.

In a new RadioGraphics article, Lauren J. Kelsey, BS, and colleagues from the University of Michigan Medical School, Ann Arbor, explore the potential of a modern, whole-body, mid[1]field-strength 0.55T MRI system for neurologic imaging. Although image resolution may be lower compared with conventional 1.5T and 3T systems, the authors highlight significant advantages that include cost savings, enhanced patient comfort and opportunities for use in nontraditional settings such as emergency departments.

“Modern midfield systems have the potential to improve access to MRI as they require fewer resources to acquire and maintain,” the authors conclude.

Read the full article, “Routine and Advanced Neurologic Imaging at 0.55-T MRI: Opportunities and Challenges,” and invited commentary, “Routine and Advanced Neurologic Imaging at 0.55-T MRI: Opportunities and Challenges.”

This article is also available for CME on EdCentral. Follow the RadioGraphics editor on X @RadG_Editor.

Congenital diaphragmatic hernia (CDH) is a rare disease where a defect in the diaphragm allows abdominal organs to move into the thorax. It often hinders lung development and causes breathing problems at birth.

Advanced imaging techniques like fetal MRI are used to measure lung volumes and predict outcomes, with low lung volume indicating a poor prognosis and possible need for treatments like tracheal occlusion.

Recent studies are exploring the use of deep learning algorithms to improve segmentation of fetal MRI scans for more accurate assessments, but further evaluation is needed for their application in fetuses with CDH.

In a Radiology Advances article, Leon M. Bischoff, MD, University Hospital Bonn, in Germany, and colleagues sought to determine if DL segmentation of total fetal body volume (TFBV) and total fetal lung volume (TFLV) in 208 fetuses with CDH demonstrated comparable performance to manual segmentation.

“DL for body and lung segmentation in fetuses with congenital diaphragmatic hernia allows reliable and rapid calculations of the observed/expected ratio and equally predicts prognostic outcome,” the authors conclude.

Read the full article, “Fetal MRI deep learning segmentation of body and lung in congenital diaphragmatic hernia.”