Non-invasive video capsule endoscopes (VCE) are widely used for gastrointestinal (GI) imaging. Usually, excretion of VCE occurs within 2 weeks. However, excretion is not always observed by the patient and rare cases of capsule retention for years have been reported.  Magnet Resonance Imaging (MRI) with incorporated VCE is considered a contraindication due to possible adverse interaction with the ferromagnetic components of VCE. While single case reports with incidental MRI have documented no clinically evident harm due to VCE, systematic experimental data are lacking. This study aims to quantify magnetically induced displacement forces and torque on VCE within the electromagnetic environment of a 3 Tesla (T) MRI scanner according to internationally recognized ASTM standards.

This experimental study used a 3T whole-body MRI scanner (Magnetom Vida, Siemens Healthcare, Erlangen, Germany). Eleven commercially available VCEs (PillCamSB3, PillCamColon2, OMOM HD, OMOM RC, OMOM CC, NaviCam, NaviCam Colon, CapsoCam Plus, MiroCam 1200, MiroCam 2000, MiroCam Navi) designed for imaging of the small bowel, upper GI tract, or colon were tested. For displacement force measurement, each capsule was suspended by a string near the MRI bore entrance at the position of maximum spatial magnetic field gradient (ASTM F2052-21). The deflection angle relative to the vertical axis was measured using a protractor setup. A deflection angle greater than 45° was defined as exceeding gravitational force, thus failing the MRI safety criterion.Torque was measured with the pulley method (ASTM F2213-17). Each capsule was mounted in a test apparatus allowing free rotation and connected to a calibrated force gauge. The maximum tensile force and lever arm radius were recorded to calculate torque. For VCE with excessive displacement, qualitative assessments and torque estimations based on displacement data were performed. Data were compared with MRI safety limits described in ISO/TS 10974:2018(E). Length of the paramagnetic interior parts of each VCE was calculated from the measured length and the ratio of internal ferromagnetic parts to the outer surface on X-ray images. 

All VCE except NaviCam Colon demonstrated deflection angles greater than 45°, indicating displacement forces exceeding gravitational forces within the 3 T static magnetic field. The cutoff for MRI safety was therefore not met in 10 of 11 tested devices. Capsules containing small permanent magnets allowing gastric navigation by external magnets, exhibited significantly higher displacement forces compared to VCE without magnet (2338 mN vs. 168.1 mN; p < 0.0001). In contrast, mean gravity was 36.4 mN.Torque measurements were successfully performed for eight VCE models, ranging from undetectable to 54.0 mNm. All but one capsule (NaviCam Colon) exceeded the individual gravity-induced torque safety limit (mean 1.04 mNm). For the three magnet-containing capsules torque measurement was not possible due to strong displacement forces. Instead, estimated torque values based on displacement measures were calculated, ranging from 271.95 to 293.90 mNm. Mean X-ray assisted length estimation of internal paramagnetic parts was 21.3 mm (range 18.2 – 25.4 mm). 

This study provides the first systematic experimental evaluation of magnetic displacement force, and torque acting on VCE in a 3T MRI environment. Induced forces on all but one tested capsules exceed safety thresholds defined for passive medical implants, especially for VCE containing integrated permanent magnets. Estimated length of paramagnetic VCE parts was within or only slightly above the limit of 24 mm for dynamic thermal effects in a 3T MR environment. Furthermore, VCE include air and a protective plastic cover likely protecting from thermal damage. Hence no direct thermal measurements were performed.Although the measured and calculated forces remain below those reported to cause bowel perforation, the findings highlight a potential safety hazard in clinical practice. Interdisciplinary awareness of these risks is critical for clinicians and MRI staff. Future research should include in-vivo and clinical safety assessments, as well as manufacturer-specific labeling, consideration of MR compatibility by engineers, and improved workflow for device identification before MRI procedures.