Classic ultrasound (US) is the most widely used imaging method for the diagnosis of fatty liver filtration due to its high availability, its user-friendliness, its tolerance by the examinees and its low cost. Typical findings are based on the ultrasonographic comparison to the right renal parenchyma, the peripheral attenuation and the presence of areas with focal amplification of the ultrasound beam. Ref. [20] The degree of infiltration can be subjectively classified as mild, moderate or severe, or as suggested by some studies, can be classified using absolute ultrasound criteria. Ref. [21,22] During a meta-analysis (2815 patients with suspected or confirmed liver disease), the exact sensitivity and specificity of the ultrasound to distinguish moderate to severe fat infiltration from no filtration, as characterized after biopsy, was 85% (80–90%) and 93% (87–97%), respectively. This study involved adult patients [23]. In children, liver US for fat detection presented a sensitivity of 70% to 85% and specificity of 50% to 60% [10,24]. Despite the diagnostic limitations for people with co-occurring kidney disease and obese patients found in the aforementioned study [23], the European directives for the management of NAFLD recommend the use of ultrasound as the first-choice imaging for people with suspected NAFLD/NASH [25].

A large pediatric cohort study showed a positive correlation between ultrasound degree of fatty infiltration and histologically established severity of the disease [26]. In the same study, a particularly interesting finding was the lack of a positive correlation between transaminase values with either ultrasound or histological findings, suggesting a lower diagnostic value of these markers in pediatric patients. Ultrasound was considered an effective tool for the staging and screening of these patients [26]. Interestingly, there has been a higher incidence of NAFLD/NASH in obese children with steatosis upon biopsy, which makes it a possible risk factor for the disease. Therefore, it is important to note that the combination of fatty liver on ultrasound with a high serum ALT value increased the detection of NAFLD in children in these highly suspected groups [27]. However, there was a remarkable percentage with conspicuous fatty infiltration on ultrasound and unaffected value of ALT in serum. In addition, it was found that children with fatty infiltration and a normal ALT value did not usually show other manifestations of metabolic syndrome. These studies rendered the ultrasound an effective tool for screening NAFLD/NASH in the suspected pediatric population, in contrast to the measurement of serum aminotransferases, which in the current literature seem to be insufficient [28]. A study in Egypt showed 100% sensitivity and 10% specificity in the detection of histologically confirmed NAFLD via ultrasound examination. Therefore, it was concluded that the absence of ultrasound or the presence of grade “A” hepatic infiltration could safely rule out histology and prevent biopsy [29].

To support the above, the use of ultrasound as a screening tool for pNAFLD/NASH showed high sensitivity but low specificity for fat infiltration in highly obese children [30]. To aid the diagnosis, MRE (magnetic resonance elastography) was proposed to detect the presence or absence of fibrosis. In an attempt to quantify the fatty infiltration by ultrasound (>5% of infiltration), there was a significant overlap with the findings of the MRE (classification: absence, mild, moderate and severe infiltration) [30]. In another group of children with histologically confirmed NAFLD, the ultrasound findings of fatty infiltration appeared to correlate well with the corresponding clinical parameters and liver pathology. In addition, the degree of hepatic fat infiltration was correlated with the fibrosis stage, and it was associated with higher BMI (body mass index), waist circumference, hematocrit and insulin resistance [31,32]. Regarding NASH, ultrasonography can provide a valid outcome, provided the use of US contrast matter, as a significant decrease has been found on the uptake of the contrast in NASH patients [33]. In addition to that, a new US fatty liver indicator score, based on the intensity of liver contrast, has been found to successfully evaluate the pathologic criteria for NASH, proving the usefulness of US for staging the disease in pediatric patients [21].

There are some practical limitations that undoubtedly reduce the diagnostic sensitivity of ultrasound in obese children, who are usually candidates for developing pNAFLD. The diagnostic sensitivity of the method is reduced either when the liver parenchyma contains percentages of fat below 30% or in people with a body mass index of 40 or more. In addition, ultrasound cannot detect the presence of steatohepatitis or fibrosis. In general, for the adult population, the sensitivity of ultrasound for the diagnosis of NASH varies from 60% to 90% with a specificity of 84% to 100%. The differences among NASH and NAFLD are not conspicuous via radiological diagnostic tools [34]. The limitations of traditional US diagnosis for pediatric patients could be eliminated by novel technologies such as controlled attenuation parameters that could detect steatosis [10].

Some opponents of the use of US argue that the low sensitivity and specificity of the method for the diagnosis and determination of fatty infiltration makes it unsuitable and inaccurate for the application of population screening for NAFLD/NASH in children. Despite the universally accepted fact that ultrasound is widely available and can rule out the presence of liver masses, cysts or bile pathology, a normal liver ultrasound cannot rule out the existence of NAFLD with absolute certainty and therefore, it is not preferred [35]. Ultrasound was also used to diagnose NAFLD with a positive prognostic value of 47% and 62%. No stable correlation was found between ultrasonic fatty infiltration and reference measurements. That does not clearly lead to the exclusion of ultrasound as a screening medium for NAFLD, but it rejects it for the staging of fat infiltration in children [36]. The same applies to the diagnosis of NASH. Several studies observed the limitations of US in NASH diagnosis, especially for obese patients and patients with hepatosteatosis less than 30% [37].

Thankfully, new technological extensions of the classic ultrasound have emerged to satisfactorily cover the need for staging of NAFLD/NASH not only in children, but also in the adult population.

Ultrasound seems to be used quite effectively in its classic form for the evaluation of interventions in children with NAFLD after specific periods in which a change in the metabolic profile is anticipated. A 12-month program that included diet and physical activity resulted in a significant reduction in BMI and fasting glucose, insulin, lipid and liver enzyme levels as well as hepatic ultrasound findings on conventional ultrasound [10]. Any degree of reduction in excess weight was associated with a significant reduction in the impact of NAFLD. Even a minimal reduction in excess weight led to a significant improvement in the ultrasound image, due to the reduction of fatty filtration [38].