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.

REVERBERATION ARTEFACTS IN LUNG US, WFUMB POSITION PAPER

Abstract
The analysis of vertical reverberation artefacts is an essential component of the differential diagnosis in pulmonary ultrasound. Traditionally, they are often, but not exclusively, called B-line artefacts (BLA) and/or comet tail artefacts (CTA), but this view is misleading.
In this position paper we clarify the terminology and relation of the two lung reverberation artefacts BLA and CTA to specifc clinical scenarios. BLA are defned by a normal pleura line and are a typical hallmark of cardiogenic pulmonary edema after exclusion of certain pathologies including pneumonia or lung contusion, whereas CTAs show an irregular pleura line representing a variety of parenchymal lung diseases. The dual approach using low frequency transducers to determine BLA and high frequency transducer to determine the pleural surface is recommended.

Keywords: lung ultrasound; artefact; B-lines; comet tails; guidelines; misdiagnosis

Suggested approach:

The transducer should be positioned such that the emenating ultrasound beam perpendicularly intersects
the surface of the lung to maximize likelihood of seeing all BLA and CLA as well as A line artifacts (fig 1).
A recent study highlighted the potentially detrimental effects of placing the focal zone below the pleural line,using spatial compounding, higher frequency and tissue harmonics [14]. Once machine settings and transducer orientation have been optimized, we suggest that two most important and distinct vertical lung artefacts should be differentiated: BLA and CTA. While true BLA (fig 2) originate from a smooth pleural reflex due to cardiogenic pulmonary edema and present in a diffuse pattern, CTAs are seen in many lung disorders with irregular and fragmented pleural reflexes and can be focal or diffuse (fig 3).
Hence, the initial step should be to determine if there is evidence for diffuse pulmonary disease or defned focal or localized pathology. Focal lung pathologies by defnition should display vertical artifacts that are consistent with CTAs (fig 4).
Diffusely distributed vertical reverberation artefacts can be divided into two groups: with or without detectable pleural line irregularities and with stable or distally widening width:

1. The reverberation artefact (evaluated by low frequency transducer <5 MHz without interfering presets) is called BLA if arising from a smooth pleural line (evaluated by high frequency transducer ≥10 MHz). The BLA arises from edema within the interstitium, is well defned with stable width, hyperechoic and extending indefnitely (the entire depth, at least 10 cm), erasing A-lines and moving with lung sliding. It is important to realize that many modern ultrasound machines have post-processing and other features which will eliminate not only BLA but essentially all discernable image detail near the bottom of the screen at greater depths (fig 5).

2. The reverberation artefact is called CTA if arising from an irregular (or fragmented) pleural line (evaluated by high frequency transducer ≥10 MHz), changes in width (such as e a comet with narrow head and wide tail), is well defned, hyperechoic, and extending defnitely (<10 cm in depth) (evaluated by low frequency transducer <5 MHz without interfering presets). It is important to make sure image compounding is turned off to make sure the CTA is not distorted farther field [14].
The differentiation of BLA from CTA is also dependent on the technical adjustments of several external factors, including the type of ultrasound machine, transducers and probe frequencies [6].

In conclusion, the correct diagnosis of pulmonary edema (the etiology of which may be decided upon
through integration of ultrasound data with clinical presentation) in the emergency setting is crucial for the correct management of the patient. The differentiation between ultrasonographic BLA and CTA, using two types (high and low frequency) of transducers allows accurate differentiation between pulmonary edema and other cause of diffuse pulmonary pathology. Both can lead to acute respiratory failure but may require different clinical management. Localized pulmonary diseases representing with CTA are distinguished. Mixed forms of diffuse,but also diffuse and focal lung diseases have to be considered.