Diagnostic Pitfalls of High‐Frequency Ultrasound in Pediatric Peripheral Neurogenic Tumors: An Analysis of 44 Challenging Cases
Tao Qi, Shen Qi, Zhou Suzhi, Ye HaipengABSTRACT
Objective
This study systematically analyzed misdiagnosis classification, incidence and root causes of pediatric peripheral neurogenic tumors on high‐frequency ultrasound among diagnostically ambiguous cases, and summarized core diagnostic pitfalls.
Methods
We retrospectively collected clinical and ultrasonographic data from 44 children with superficial soft tissue masses treated between January 2022 and June 2026. Final diagnoses were confirmed via surgical pathology, pathogenic genetic testing combined with typical multimodal imaging, or multidisciplinary consensus after a minimum 12‐month clinical and MRI follow‐up. Only atypical lesions with initial ultrasound suspicion of neurogenic origin were enrolled; all cases with classic definitive sonographic signs were excluded. Using final pathological or follow‐up diagnosis as the gold standard, we quantified misdiagnosis proportions and categorized misdiagnosis patterns.
Results
Of the 44 enrolled children, 31 (70.5%) were confirmed to have peripheral neurogenic tumors, and 13 (29.5%) carried non‐neurogenic lesions. A total of 15 misdiagnoses occurred (34.1% of the cohort). Thirteen misdiagnoses represented source misjudgment, in which all non‐neurogenic lesions were incorrectly categorized as neurogenic; these predominantly included five lipogenic tumors, four fibrogenic lesions, and four other miscellaneous soft tissue masses. Two misdiagnoses were subtype misjudgments of confirmed neurogenic tumors (6.5% of all neurogenic cases). Importantly, this cohort only included atypical confusing lesions without classic imaging features; the 0% specificity of the combined ultrasound criteria only reflects ultrasound performance in difficult cases and cannot be extended to routine pediatric superficial masses. The rat‐tail sign yielded 100% specificity (95% confidence interval (CI): 75.3%–100%) and 100% positive predictive value (95% CI: 73.5%–100%) for identifying neurogenic lesions, while its sensitivity was only 38.7% (12/31; 95% CI: 21.8%–57.8%).
Conclusions
The rat‐tail sign achieves 100% specificity and positive predictive value for neurogenic lesions within this atypical cohort, yet shows only 38.7% sensitivity. Major pitfalls include pseudo‐rat‐tail artifacts and mimicking lesions. Integrating patient age, growth velocity, and dynamic multiplanar tracing into a standardized stepwise workflow can reduce false‐positive misdiagnoses in daily practice.