Comparative accuracy and reliability of handheld versus cart-based ultrasound in assessing inflammatory bowel disease activity

Abstract

BACKGROUND: Bowel wall thickness (BWT) and vascularity are key parameters for assessing inflammatory bowel disease (IBD) activity. Handheld ultrasound devices (HHUS) provide a portable and affordable alternative for incorporating intestinal ultrasound (IUS) into IBD management. However, their diagnostic accuracy and reliability compared to cart-based ultrasound (CBUS) remain uncertain. This study assesses the diagnostic accuracy, reproducibility, and practitioner confidence with HHUS in comparison to CBUS for assessing BWT and vascularity. METHODS: In this prospective, comparative study, 12 patients with IBD, with body mass index (BMI) <23 kg/m2, were assessed. Each patient underwent ultrasound scanning of 2-3 pre-specified bowel segments using five devices: one CBUS device (SamsungTM) and four HHUS devices (ClariusTM L15 HD3, GETM Vscan Air CL, MindrayTM TE Air, and PhilipsTM Lumify L12-4). Linear probes were used for all devices except MindrayTM, which employs a phased array probe optimised for cardiac imaging. Scans were conducted by two to three gastroenterologists accredited by the Gastroenterology Network of Intestinal Ultrasound (GENIUS), who were blinded to each other's findings to reduce bias. BWT and Doppler vascularity, as well as other standard IUS parameters, were assessed for each segment. Diagnostic accuracy and inter-/intra-expert agreement were analysed using statistical models. RESULTS: Among the 12 patients (42% male; median age: 26 years, IQR: 23-36) with CD (n=6) and UC (n=6), a total of 29 bowel segments were assessed with all devices (Table 1). BWT measurements showed high concordance across devices for active segments (mean BWT: 4.3-4.6 mm, p=0.73). However, significant discrepancies were noted in inactive segments (p<0.001), particularly with the MindrayTM device showing larger deviations from the CBUS measurements. Intraclass Correlation Coefficients (ICC) for intra-expert BWT measurements were good to excellent across devices (ICC 0.79-0.97), while inter-expert agreement was good for all devices (ICC 0.70-0.77) except for MindrayTM (ICC 0.36) (Figure 1B). Significant confidence was noted for BWT, vascularity, stratification, mesenteric fat, and inflamed bowel length across devices, highlighting consistent measurement reliability for these parameters (Table 1). CONCLUSIONS: The majority of HHUS devices used in this study demonstrate high diagnostic accuracy and reproducibility in assessing IBD activity and are comparable to CBUS for identifying and assessing active disease. This study supports the integration of HHUS into IBD management as a portable and accessible tool while highlighting the importance of recognising device-specific limitations when selecting devices for clinical practice.