Higher spatial resolution and thinner slice thickness are essential in ultrasound imaging because they reduce out-of-plane contamination and improve micro-lesion detection. These principles also apply to endoscopic ultrasound (EUS), where image quality directly affects diagnostic accuracy and the safety of EUS-guided procedures. However, spatial resolution and slice thickness have rarely been evaluated systematically or quantitatively across EUS devices. This study aimed to quantitatively assess axial and lateral spatial resolution and slice thickness across multiple EUS platforms, depending on imaging frequency and measurement depth.

We used the following endoscope– processor pairings: GF-UCT260 with EU-ME2/EU-ME3 (Olympus, Japan) and Aplio i800 (Canon, Japan); and EG-580UT/EG-740UT with SU-1 (Fujifilm, Japan). Spatial resolution and cyst model delineation were evaluated using the phantom (Kyoto Kagaku N-365 multipurpose phantom, Japan), and slice thickness was evaluated using a custom 45° inclined phantom (film: OHM laminate film 100 µm [LAM-FA41003]).

[Spatial resolution] axial/lateral resolution were evaluated using wires (diameter 0.05 mm, spacing 0.5, 1, 2, 3, and 4 mm) at measurement depths of 15, 30, and 50 mm.

[Cyst delineation] Cyst models (1, 2, 3, and 4 mm) were evaluated at measurement depths of 10, 30, and 50 mm.

[Slice thickness] Ultrasound beams were vertically incident, and the longitudinal width (X) of the hyperechoic band generated in the inclined direction on the image was measured. Because the inclination was 45°, X was defined as the slice thickness (Y mm). Measurements were performed for each frequency and measurement depth (10–60 mm).

[Spatial resolution]Axial resolution at measurement depths of 15 and 30 mm was 0.5 mm, while lateral resolution at the same depths was 1–2 mm. At a depth of 50 mm, axial resolution was 1 mm, and lateral resolution was 4 mm or unmeasurable. In all devices and under all conditions, axial resolution was superior to lateral resolution. The performance was nearly equivalent across all endoscopes and processor.

[Cyst delineation]Cyst models with diameters of 1, 2, 3, and 4 mm were detectable at a measurement depth of 10 mm, and those with diameters of 2, 3, and 4 mm were detectable at 30 mm. At a depth of 50 mm, all cysts were non- detectable. The performance was nearly equivalent across all endoscopes and processor.

[Slice thickness]

Slice thickness reached its minimum at a scope-specific focal depth. For the Olympus GF-UCT260 scope, the focal depth was 30 mm, with minimum values of 1.55 mm (12 MHz, EU-ME2), 1.52 mm (THE-R, EU-ME3), and 1.66 mm (d11, Aplio i800). In contrast, for the Fujifilm SU-1 processor, the focal depth was 20 mm, with minimum slice thicknesses of 1.63 mm (THE-8.0, EG-580UT) and 1.87 mm (12 MHz, EG-740UT). These findings indicate that the depth at which slice thickness is minimized differs among devices. Across all systems, slice thickness increased at both shallower and deeper depths relative to the focal depth. At 10 mm, slice thickness ranged from 4.33–4.86 mm (EU-ME2), 4.97–5.23 mm (EU-ME3), and 3.65–4.25 mm (Aplio i800) for GF-UCT260, and 1.98–3.74 mm (EG-580UT), 2.22–3.34 mm (EG-740UT) on the SU-1 processor. At 60 mm, values ranged from 3.90–4.60 mm (EU-ME2), 3.05–4.55 mm (EU-ME3), and 3.70–3.95 mm (Aplio i800) for GF-UCT260, and 3.72–5.22 mm (EG-580UT), 6.47–6.99 mm (EG-740UT) on the SU-1 processor. 

In EUS, spatial resolution is maintained at depths ≤30 mm but declines at 50 mm. Slice thickness is lowest at a scope-specific focal depth and increases at shallower and deeper levels. Thus, targets away from the focal depth may become partially embedded within the slice, raising the risk of off-target sampling and bleeding. Aligning the lesion with the focal depth and adjusting frequency, scope choice, and approach angle to device-specific characteristics may improve procedural safety, tissue acquisition, and diagnostic accuracy. These findings also enhance endoscopy service quality by providing objective metrics to optimize EUS imaging settings and standardize EUS-guided procedures.