Special training not needed for COVID lung ultrasound scoring

By Amerigo Allegretto, AuntMinnie.com staff writer

May 26, 2021 -- Although there were some variations in performance, it may not be necessary for users of ultrasound scanners to receive special training before scoring lung exams of patients with COVID-19, according to research published May 21 in Nature.

A group of researchers led by Dr. Markus Lerchbaumer from Charité Institute of Radiology in Berlin, Germany found that while different types of observers had "fair to moderate" interobserver agreement in interpreting specific lung findings in patients with COVID-19, a little background in lung ultrasound goes a long way.

"As long as observers have some experience in lung ultrasound, no specific clinical background is needed for scoring the findings, even though specific expertise is often reported as a requirement," the authors wrote.

Examining the lungs in patients is usually performed with nonenhanced CT scans, but point-of-care ultrasound (POCUS) is being looked at as a safer method since patients infected with the SARS-CoV-2 virus would not need to be transferred, and risk of exposure for medical staff would decrease.

The researchers said lung ultrasound may have a big advantage for COVID-19 due to its widespread availability and cost-effectiveness.

"Additionally, lung ultrasound has emerged over the last two decades as a noninvasive tool for the fast differential diagnosis of pulmonary diseases and is now used in different settings in intensive care," they wrote.

In the current study, 10 observers from three different medical specialties participated in rating 100 lung ultrasound images from 13 patients. These included observers specializing in intensive care medicine, emergency medicine, and physiology.

Images were acquired by a radiologist using a hand-held POCUS system (Viamo sv7, Canon Medical Systems) performed at the bedside. Ultrasound presets were optimized for lung ultrasound.

Through an online tool, observers could use multiple-choice options with predefined answers for rating the scans. Options included typical COVID-19-associated lung ultrasound findings; these included the following:

• Pleural thickening and fragmentation
• Presence of B-lines subclassified in single or confluent, subpleural consolidations
• Positive air bronchogram

Selecting none of these pathologies was also an option.

The team found that observers in the intensive care unit tended to interpret B-lines more accurately, while physiology researchers and emergency physicians more often categorized B-lines as confluent rather than single. This tendency became even stronger over the course of viewing instances, probably explaining the poorer than expected overall inter- and intraobserver agreement.

"We assume that ICU observers have greater clinical experience with patients with severe ARDS or cardiogenic edema and their corresponding lung ultrasound findings, especially compared to scientists whose experience relies on lung ultrasound in rodents," they wrote.

ICU observers, on the other hand, differed from the latter two groups regarding the identification of pleural thickening.

Meanwhile, agreement was highest for more distinct lung ultrasound findings such as air bronchograms and subpleural consolidations, as well as more severe lung ultrasound scores.

The researchers wrote that training lung ultrasound users may improve agreement and clinical feasibility. They also suggest that training material used for lung ultrasound in POCUS should pay better attention to areas such as B-line quantification and differentiation of intermediate scores, which revealed only "mediocre" agreement in the study.

US AI model can help evaluate chronic kidney disease

By Erik L. Ridley, AuntMinnie.com staff writer

May 24, 2021 -- Artificial intelligence (AI)-based analysis of kidney ultrasound studies could serve as a first-line method for evaluating patients with chronic kidney disease, according to research published online May 24 in JAMA Network Open.

A team of researchers led by Dr. Ambarish Athavale of Cook County Health in Chicago developed a deep-learning algorithm that yielded approximately 90% accuracy on a test set for quantifying interstitial fibrosis and tubular atrophy (IFTA).

"This article provides proof-of-principle that a [deep-learning] system can be used to noninvasively, accurately, and independently predict IFTA grade in patients with kidney disease," the authors wrote. "Although the system in its current form may not be an alternative to kidney biopsy, after robust external validation, a [deep learning]-based, noninvasive assessment of IFTA has the potential to significantly enhance clinical decision-making and prognostication in patients with CKD."

A strong indicator for decline in kidney function, interstitial fibrosis and tubular atrophy is currently measured using histopathological assessment of a kidney biopsy core. There currently isn't a noninvasive test for IFTA, according to the researchers.

The authors utilized AI to test their hypothesis that subtle signs of IFTA are ingrained within kidney ultrasound images and could be quantitatively extracted and analyzed. A deep-learning algorithm was trained and tested to segment the kidney and classify IFTA using 6,135 consecutive Crimmins-filtered kidney ultrasound images acquired at their institution between January 1, 2014, and December 31, 2018. The longitudinal images were obtained from both kidneys and were acquired between six months before and two weeks after kidney biopsy.

Of the total image dataset, 5,122 were used for training and 401 were used for validation. The researchers then tested the model on 612 images. The algorithm was 91% accurate for segmenting the kidney ultrasound images.

Performance of AI algorithm for quantifying IFTA on kidney ultrasound
	Image level	Patient level
Precision	0.893	0.900
Recall	0.804	0.842
Accuracy	0.868	0.896
F1 score	0.839	0.864

In other results, the researchers noted that the algorithm's accuracy remained high irrespective of the timing of the ultrasound studies and the biopsy diagnosis. Also, adding baseline clinical characteristics into the model's analysis didn't significantly improve its performance.

"From a clinical standpoint, it is foreseeable that a [deep-learning] system such as the one developed in this study has the potential to act as a gatekeeper for rationalizing the decision to conduct a kidney biopsy in patients with CKD," the authors wrote. "We anticipate that because of the ability of this system to provide [a] probabilistic estimate of IFTA in real-time, the system is likely to be acceptable (because it is unlikely to put any time burden on the technicians) and can also reduce the costs associated with kidney biopsy."

The researchers acknowledged that more work is needed to improve the accuracy of the model before it's ready for clinical use. Furthermore, the algorithm needs to be validated on external datasets to assess its performance across varying clinical settings.