Evaluating artificial intelligence for a focal nodular hyperplasia diagnosis using magnetic resonance imaging: preliminary findings - Diagnostic and Interventional Radiology
Background/Objectives: The prevalence of developmental hip dysplasia is estimated to be 0.1–2 per 1000 infants. Hip imaging by ultrasonography is considered to be the gold standard method for screening and detecting developmental dysplasia of the hip (DDH), as per the Graf categorization. The classification of hip differentiation into type 1 and type 2 is determined by the alpha angle, as assessed by the Graf classification. Type 1 hips are defined as those with an alpha angle exceeding 60 degrees, whilst type 2 hips are defined as those with measurements falling within the range of 50 to 59 degrees. Methods: The computerized patient card in our institution had a compilation of 208 hip photographs taken from 110 patients, with 98 of them being bilateral. The acquisition of these photos occurred from January 2020 to December 2020. A retrospective review was conducted on the ultrasound (US) scans, with a specific emphasis on the outcomes related to type 1 and type 2 hips. Results: There were 108 high-resolution US photos in the type 1 hip group and 100 high-resolution US images in the type 2 hip group. In terms of unilateral or bilateral cases, gender, or age, no statistically significant differences were seen between the two groups (p > 0.05). The FMD model exhibited a sensitivity of 86% and specificity of 70% in effectively predicting the presence of type 1 mature hips when the values surpassed 2.9 mm. The AUC (area under the curve) value achieved was 0.628. Conclusions: The process of diagnostic categorization may occasionally encounter challenges in accurately differentiating between type 1 and type 2 hip separation subsequent to a hip ultrasound examination. The findings of our analysis indicate that the assessment of the FMD is a highly successful method, demonstrating both high specificity and sensitivity in differentiating between various scenarios.
Background: Fenestration is a segmental duplication of an artery and is a new entity described in coronary arteries, distinct from dissection. This study aimed to show CT angiography (CTA) findings of fenestrated coronary arteries, along with their possible associations or any accompanying findings. Methods: This was a multicenter study and included examinations from four different institutions. Between the years 2022 and 2025, 5554 patients who underwent coronary CTA without prior coronary catheter angiography and intervention were included in this study. Coronary CTA was acquired using four different multidetector CT scanners. Computed tomography angiography (CTA) examinations were reevaluated for fenestration. Fenestrations were analyzed for morphological type, length, and diameter. Results: In total, 26 of 5554 patients (0.47%) had coronary artery fenestration. Fenestration was observed in 13 left circumflex (Cx) arteries, 8 left anterior descending (LAD) arteries, and 5 right coronary arteries (RCA). Two of the 26 patients underwent catheter angiography. Single-loop-type fenestration was observed in 12 patients, and two patients had multiple-loop fenestrations. There were 12 patients with parallel-type fenestration. Among the patients with fenestrated coronary arteries, myocardial bridging was observed in 5 patients as an associated anomaly. No significant difference was found in the length or diameter of the fenestrated segments. There were no symptoms related to fenestration in our patients. Conclusions: Fenestration of a coronary artery can be demonstrated by CTA and may be associated with myocardial bridging. It is most common in the Cx artery. Fenestration should be differentiated from dissection for clinical implications. Further studies are warranted to understand the importance of this observation.
14 July 2026 in Archives of Basic and Clinical Research