Magnetic resonance imaging-induced forces on video capsule endoscopes: First systematic in vitro assessment

Abstract

BACKGROUND

Magnetic resonance imaging (MRI) is contraindicated in patients with an ingested video capsule endoscopy (VCE) due to the theoretical risk of mechanical and thermal harm. However, isolated cases of inadvertent scanning have not shown clear harm. Manufacturers advise against performing VCE in patients with an MRI scheduled within the next 14 days, yet asymptomatic retention for up to 12 years has been reported. With the increasing use of VCEs, the likelihood of incidental MRI exposure is rising. To date, no systematic data have been published on the forces exerted on VCEs in an MRI environment.

AIM

To quantify magnetically induced forces on VCEs in a 3-Tesla (3-T) scanner to assess clinical safety risks.

METHODS

Displacement forces on 11 commercially available VCE models were measured in a 3-T MRI scanner according to the American Society of Testing Materials (ASTM) F2052-21. Following ASTM F2213-17, torque was measured by the pulley method for eight VCEs and calculated for three magnet-containing gastric capsules with high displacement forces. The lengths of internal metallic components of the VCEs, as indicators of potential thermal effects, were determined using X-ray images.

RESULTS

VCEs containing magnets showed significantly higher median displacement forces than those without magnets (2366 mN vs 293.92 mN, P = 0.012). Magnetically induced displacement forces were 50.7-fold higher than the individual gravitational force for magnet-containing devices (2366 mN vs 46.66 mN), compared to a 9.7-fold excess for capsules without a magnet (293.92 mN vs 30.20 mN). Median torque was 19.2 mNm for capsules without integrated magnets vs 289.02 mNm with magnets (P = 0.017). For all but one VCE, the forces exceeded the safety threshold defined by the capsule’s gravitational force. The mean length of internal metallic components was 21.3 mm (range: 18.2-25.4 mm); only two capsules slightly exceeded the 24-mm threshold for possible thermal effects.

CONCLUSION

In a 3-T MRI, all but one VCE, particularly magnet-containing capsules, exceeded accepted safety thresholds for displacement force and torque but remained below perforation-relevant levels.

Key Words: Video capsule endoscopy; Magnetic resonance imaging; Capsule retention; Displacement force; Torque; Magnetic resonance imaging safety; Inadvertent imaging; Medical device safety

Core Tip: With the increasing use of video capsule endoscopes (VCEs), accidental magnetic resonance imaging (MRI) in patients with an incorporated VCE is a growing risk. To date, no systematic data have been published on MRI-induced forces on VCEs. This in vitro study is the first to systematically quantify displacement forces and torque on 11 commercially available VCEs in a 3-Tesla MRI scanner. All but one VCE exceeded accepted safety thresholds, with magnet-containing capsules showing forces up to 50-fold above gravitational force. These findings highlight a potential safety concern relevant to both clinicians and engineers.

INTRODUCTION

Video capsule endoscopy (VCE) is well established as a non-invasive endoscopic technique for visualization of the small bowel, particularly in patients with suspected small bowel bleeding, inflammatory bowel disease, complicated celiac disease, certain polyposis syndromes, and suspected small bowel tumors[1].

The technique has been extended to colon imaging, primarily as an alternative to flexible colonoscopy in regions with limited resources. More recently, VCE has been adapted for gastric evaluation. This procedure requires a small permanent magnet inside the capsule and either an external handheld permanent magnet for steering within the large gastric lumen or an external robotic arm with a variable electromagnetic field. These external devices typically produce magnetic fields of approximately 0.2-Tesla (T), which are not considered clinically hazardous.

Contraindications for VCE are few and include known or suspected gastrointestinal stenosis without proven patency, swallowing disorders without the option of endoscopic capsule placement, and pregnancy due to a lack of data on safety. Implanted cardiac devices, such as pacemakers and defibrillators, are not considered relevant contraindications, as multiple applications have been performed without any detectable influence of the radiofrequency used for transmitting data to the external recorder and vice versa[2].

Magnetic resonance imaging (MRI) with an ingested video capsule is considered a contraindication due to the theoretical possibility of inducing mechanical or thermal damage to the human body[3]. Manufacturers advise against performing VCE in patients with an MRI scheduled within the next 14 days, which by consensus is considered the time limit for normal excretion[4]. Notably, many patients do not observe excretion of the capsule[5], and asymptomatic retention for up to 12 years has been reported[6], raising the likelihood of incidental MRI exposure. Several case reports describe inadvertent MRI in patients with retained capsules without clinical harm[7-10].

To date, no systematic data have been published on MRI-induced forces on video capsules. This study provides the first systematic assessment of magnetically induced displacement forces and torque in a range of commercially available VCEs in a 3-T MRI environment. X-ray imaging was used to quantify the length of the internal metallic components and thereby evaluate potential thermal risks. These measurements aim to improve the understanding of patient risks but do not constitute formal safety evaluations, which must be performed independently by each manufacturer.

MATERIALS AND METHODS

MR scanner

A Magnetom Vida (Siemens Healthcare, Erlangen, Germany), 3T, horizontal static magnetic field, actively shielded, was used.

Magnetically induced displacement force

The magnetically induced displacement force was measured according to the American Society of Testing Materials (ASTM) F2052-21 Standard Test Method (“Standard Test Method for Measurement of Magnetically Induced Displacement Force on Medical Devices in the Magnetic Resonance Environment”)[11]. The medical device is suspended by a string in an MR system at a location near the entrance of the bore and on the axis of the bore. To increase the measurement sensitivity, this location should be chosen to ensure that the spatial gradient of the field strength, ∇B = dB/dz, is within 20% of the maximum value of the spatial gradient on the axis of the bore (Figure 1). In the present study, this point was located 87 cm from the isocenter along the longitudinal axis (Z). The angular deflection of the string from the vertical was measured three times and averaged (Figure 2). If the deflection was less than 45° (Figure 2A), the magnetic force induced by the MRI system was considered lower than the gravitational force [gravitational force = 𝑚 × 𝑔 (mN). 𝑚: Mass of device (kg). 𝑔: Acceleration due to gravity (m/s²)]. For larger angles, additional counterweights were applied.

Figure 1 Test fixture for magnetically induced displacement force measurement. Static magnetic field B₀ at the isocenter, spatial magnetic gradient ∇B = dB/dz of magnetic field dB between coordinates on the Z-axis (dz). MR: Magnetic resonance.

Figure 2 Test setup for calculation of displacement force. A: Deflection angle < 45°; B: Deflection angle > 45° even with additional weights.

For paramagnetic materials (e.g., nitinol, CoCrMo alloys, titanium and its alloys, 316 L stainless steel) and unsaturated ferromagnetic materials, the magnetically induced deflection force is proportional to the product of the static magnetic field and its spatial gradient. For devices composed of these materials, the location of maximum deflection is the point where |B| |∇B| is maximal. For saturated ferromagnetic materials, the maximum deflection occurs at the location where ∇B is maximal.

Test preparation: B was determined by using a 3-axis Hall magnetometer (THM1176, Metrolab Technology SA, Plan-les-Ouates, Geneva, Switzerland; accuracy = ± 1%). The spatial magnetic gradient was calculated from B-values. The coordinate system is located with its origin at the isocenter. The magnitudes of the static magnetic field B and the spatial gradient |B| are provided for the testing location (x/y/z = 0/0/87 cm). Before the test, the fan was switched off, and the protractor’s 0° line was aligned to the vertical. The test fixture was positioned in front of the magnetic bore.

Test procedure: The test object was positioned with its center of mass at the location of the maximum calculated spatial gradient along the Z-axis, B_max = dB/dz = 4.51 T/m with a magnetic field strength of B = 1.78 T (position 1: Z = 87 cm). At that location, the magnetically induced force, F_m, was horizontal, and both B and ∇B acted in the Z-direction. The test was performed by holding the string in the vertical 0° position and then releasing it. Afterwards, the angle α (between the string and 0° position) was recorded to the nearest 0.5° marking. This procedure was performed three times, and the mean value was used to calculate the displacement force [displacement force (𝐹𝑚) = 𝑚 × 𝑔 × tan 𝛼. 𝑚: Mass of device (kg) including additional counterweights, where necessary. 𝑔: Acceleration due to gravity (m/s²). 𝛼: Angular deflection of string measured with a protractor (°)].

Non-magnetic (plastic) counterweights were attached to the test object for compensation in cases of deflection angles close to 90° (Figure 2B), as these angles allow no reasonable resolution for determining magnetic force.

Weights used: Small tape roll = 14.642 g ± 0.001 g, large tape roll = 37.429 g ± 0.001 g, total weight (both rolls) = 52.071 g ± 0.001 g. An additional test was performed at a position where the magnetic field gradient was negligible. In this test, the string holding the test object remained almost perfectly vertical (0°), indicating that almost no force acted on the object.

Magnetically induced torque (pulley standard method)

Test method: The pulley standard quantitatively measures the magnetically induced torque of the test device using the tensile force along a defined radius. In this study, the test apparatus was fixed centrally on a holder (H) to prevent movement; the holder was rotatably mounted on a holding fixture. The holding fixture was placed on the patient table and positioned in the isocenter of the MR system using the laser indicator. A string (S) was rolled onto the roll (R), which was attached to the holder (H). The loose end of the string (S) was connected to a force gauge (F). The tensile force was recorded by pulling the string (S) with the force gauge (F) along the defined radius (r) of the roll (R). The torque was calculated as the product of the measured maximum tensile force and the lever arm (radius of the roll) [τ = F × r. τ = torque (mNm). F = tensile force along the roll (N). r = radius of the roll (mm)]. The maximum magnetically induced torque was compared with the worst-case torque due to gravity, conservatively defined as the product of the device length and its weight. This method was in accordance with the ASTM F2213-17 Standard Test Method[12].

Test procedure: The test object was fixed approximately centrally with its magnetic center and secured in a non-twistable manner onto the holder (H) of the test apparatus with tape. The test apparatus was placed on the patient table and positioned in the isocenter using the laser indicator (Figure 3). The test was performed by pulling with the digital force gauge (F) on the lightweight string, which was rolled onto the roll (R) (Figures 4 and 5). After a 360° rotation of the test object, the maximum tensile force was recorded. This process was repeated three times, and the maximal value was used for consideration. The maximum magnetically induced torque was calculated as the product of the measured maximum tensile force and the radius of the roll (R). This procedure was performed for all principal axes of the device in the vertical direction. The fan within the magnet bore was switched off during the test.

Figure 3 Magnetic resonance system with test apparatus relative to the isocenter. MR: Magnetic resonance.

Figure 4 Schematic drawing of test apparatus for torque with force gauge.

Figure 5 Test apparatus with force gauge for torque.

Magnetically induced torque (calculated based on displacement force)

Test method: The test was performed according to ASTM F2213-17, “Standard Test Method for Measurement of Magnetically Induced Torque on Medical Devices in the Magnetic Resonance Environment”[12]. This test method provides an upper bound for the magnetically induced torque, which is determined by the magnetic saturation and the measured displacement force according to ASTM F2052 [𝜏𝑚𝑎𝑥 = 𝑀𝑠 × 𝐹𝑚/4 × (∇𝐵). 𝜏𝑚𝑎𝑥: Maximum magnetically induced torque (Nm). Ms: The magnetic saturation of the test object (T). 𝐹𝑚: The measured magnetic force according to Test Method F2052 (N). ∇B: The spatial gradient magnetic field at the location of the measured magnetic force (T/m)].

Test procedure: Prior to the test, the displacement force of the test object was measured according to ASTM F2052. The upper bound for the magnetically induced torque was then calculated based on the displacement force and the magnetic saturation of the test object. If the test object consisted of several materials, the material with the higher magnetic saturation was used. The magnetic saturation of the test object was not specified by the manufacturers; therefore, the conservative magnetic saturation value of iron (2.2 T) was used.

Quality assessment

Calibrations of all measuring instruments (magnetometer, force gauge, length measurer, scale, protractor) were performed by external testing laboratories certified for this specific requirement and meeting the standards of the German accreditation body (Deutsche Akkreditierungsstelle GmbH; DAkkS, Berlin, Germany).

All measurements were performed three times. Additionally, reference measurements were taken before each test to ensure that the initial conditions were identical for all test measurements.

Measurement uncertainties were calculated separately for each test following the guide to the expression of uncertainty in measurement guidelines[13]. Basically, calculations were based on measures of examiner’s reading accuracy, length measurement for the position in the MRI environment, magnetometer readings for magnetic flux density, scale readings for determining weights, and angles measurements for reading the deflection angle (for displacement forces only), or force meter (for torque only).

Video capsules

Commercially available video capsules were tested (Figure 6): PillCam SB3, PillCam Colon2 (Medtronic, Dublin, Ireland); OMOM HD, OMOM RC, OMOM CC (Jinshan Science and Technology Co. Ltd., Chongqing, China); NaviCam, NaviCam Colon (Ankon, Wuhan, Hubei Province, China/AnX Robotica, Plano, TX, United States); CapsoCam Plus (CapsoVision Inc., Saratoga, CA, United States); MiroCam 1200, MiroCam 2000, MiroCam Navi (IntroMedic Co. Ltd., Seoul, Korea).

Figure 6 Video capsule endoscopy tested. From left to right: PillCam SB3, OMOM HD, MiroCam 1200, CapsoCam Plus, MiroCam 2000, PillCam Colon2, OMOM CC, NaviCam Colon, NaviCam, OMOM RC, MiroCam Navi. Note the magnetic attraction of the three capsules to the right.

Length, width, and weight were measured, and plain X-ray films were taken for all capsules. The digital images obtained were analyzed with a digital screen ruler (Microsoft PowerToys Digital Screen Ruler; Redmond, WA, United States) to obtain the quotient of the relative length of the metal-containing paramagnetic interior and outer surface length for each capsule (Figure 7). The product of this quotient and the directly measured length of each capsule’s outer surface were used to determine the absolute length of the paramagnetic interior.

Figure 7 Examples of the radiographs taken from all video capsule endoscopes. The relative distances between yellow lines on the left represent the outer length (L_o). The relative distances between yellow lines on the right side of each capsule mark the length of the potentially magnetic inner parts (L_i). The relation of these relative distances (measured as pixels) together with physically measured length (mm) (Table 1) is used to calculate the absolute length of the solid inner part for each capsule.

Statistical analysis

Given the exploratory nature of this first systematic in vitro assessment and the small number of devices, the analysis was descriptive and hypothesis-generating rather than confirmatory. Continuous outcomes (gravitational force, magnetically induced displacement force, and torque) are reported separately for capsules with and without an internal permanent magnet as the median and interquartile range (IQR). Groups were compared using the exact Mann-Whitney U test (Wilcoxon rank-sum test) based on the exact conditional permutation distribution, which accommodates tied values without recourse to a normal approximation. Exact P values are reported as descriptive measures alongside the rank-biserial correlation as a distribution-free effect size and the Hodges-Lehmann estimate of the median between-group difference with its 95% confidence interval; no threshold for statistical significance was applied and no adjustment for multiplicity was made, in keeping with the exploratory design. For magnetically induced torque, measured values were used for capsules without magnets, and the calculated upper-bound torque was used for magnet-containing capsules; one capsule without a magnet (NaviCam Colon) was below the measurement limit and was thus excluded from the torque comparison. All analyses were performed in R version 4.5.1, The R Foundation for Statistical Computing, Vienna, Austria.

RESULTS

The intended use, number of cameras, and physical dimensions of the tested VCEs are outlined in Table 1. Additionally, the length of the internal metal-carrying parts, as calculated from their relation to the outer surface length on X-rays, and the absolute length of these inner parts are shown.

Table 1 Characteristics of the tested video capsule endoscopes.

Video capsule endoscope	Target organ	Cameras	Permanent magnet	Weight (g)	Length (mm)	Length of internal metal parts (mm)	Width (mm)
PillCam SB3	Small bowel	1	No	2.97	26.0	18.22	11.4
OMOM HD	Small bowel	1	No	3.188	26.1	19.53	11.5
MiroCam 1200	Small bowel	1	No	3.351	24.2	18.82	11.3
CapsoCam Plus	Small bowel	4	No	3.727	30.3	25.44	11.4
MiroCam 2000	Small bowel	2	No	3.634	30.0	20.23	11.2
PillCam Colon2	Colon	2	No	2.944	32.2	21.54	11.6
OMOM CC	Colon	2	No	2.956	33.3	23.42	11.7
NaviCam Colon	Colon	2	No	2.821	31.0	24.17	11.5
NaviCam	Stomach, small bowel	1	Yes	4.653	27.0	18.15	11.5
OMOM RC	Stomach	1	Yes	5.84	30.5	23.73	11.5
MiroCam Navi	Stomach	1	Yes	4.756	25.6	20.45	11.2

Length of internal metallic parts was measured by X-ray imaging.

The length of all capsules (range 24.2-33.3 mm) was below the threshold of 50 mm for thermal safety in a 1.5 T MRI scanner but above the 24-mm threshold for a 3-T MRI scanner. Radiographically assisted length assessment of internal metal-containing parts was, slightly above the 24-mm threshold for a 3-T MRI scanner in only two devices (24.2 mm and 25.4 mm).

Displacement forces on the 11 different VCEs, as measured according to the ASTM F2052-21 Standard Test Method[11], are shown in Table 2. Measurement uncertainties derived from the displacement force observations ranged from 15%-16%. No statistically significant deviations were observed between the repeated tests (individual repeat measurements for each VCE are provided in Supplementary Table 1). All but one capsule (NaviCam Colon) showed a deflection angle > 45°, indicating displacement forces above the gravitational forces in a 3-T MR scanner and, thus not fulfilling the safety threshold.

Table 2 Calculated gravitational force and magnetically induced displacement force on the tested video capsule endoscopes.

Video capsule endoscope	Gravitational force (mN)	Magnetically induced displacement force (mN)	F mag < F grav
PillCam SB3	29.14	288.27	Not fulfilled
OMOM HD	31.27	47.91	Not fulfilled
MiroCam 1200	32.87	571.96	Not fulfilled
CapsoCam Plus	36.56	672.66	Not fulfilled
MiroCam 2000	35.65	575.92	Not fulfilled
PillCam Colon2	28.88	299.57	Not fulfilled
OMOM CC	29.00	127.5	Not fulfilled
NaviCam Colon	27.67	21.96	Fulfilled
NaviCam	45.65	2232	Not fulfilled
OMOM RC	57.29	2366	Not fulfilled
MiroCam Navi	46.66	2415	Not fulfilled

Measurements were obtained at position 1 with the maximal gradient on the Z-axis (Z = 87 cm; |∇B| = 4.51 T/m; B = 1.78 T; B₀ × |∇B| = 8.03 T²/m). “Not fulfilled” indicates that the magnetically induced displacement force exceeded the safety threshold defined by the capsule’s gravitational force. F mag: Magnetically induced displacement force; F grav: Gravitational force.

As expected, capsules intended for gastric investigation with an integrated small permanent magnet showed substantially higher median displacement forces than capsules without a magnet: 2366 mN (IQR: 2299-2391) vs 293.92 mN (IQR: 107.6-573.0), respectively (exact P = 0.012; Hodges-Lehmann median difference 2072 mN, 95% confidence interval: 1660-2344; rank-biserial r = 1.00), respectively (Figure 8). For comparison, gravitational forces were 46.66 mN vs 30.20 mN, respectively.

Figure 8 Box plots of magnetically induced displacement force and gravitational force on video capsule endoscopes. A: Magnetically induced displacement force on video capsule endoscopes (VCEs) without (left) and with (right) an integrated permanent magnet; B: Gravity-induced force on VCEs without (left) and with (right) an integrated permanent magnet. The lengths of the boxes mark the 25^th and 75^th percentiles, the yellow lines indicate the 50^th percentile (median), and the whiskers represent the 1.5-fold interquartile range. Note the different scales of the Y-axis in Figure 8A and B.

Magnetically induced displacement forces were 50.7-fold higher than the individual gravitational force for magnet-containing devices (2366 mN vs 46.66 mN), compared to a 9.7-fold excess for capsules without a magnet (293.92 mN vs 30.20 mN). All test objects except one (NaviCam Colon) exhibited strong dynamic and fast static torque (Table 3 and Figure 9) with a mean 24.0 mNm for the eight VCEs without a magnet, compared to a calculated worst-case torque due to gravity of 0.92 mNm.

Figure 9 Box plots of magnetically induced torque and gravity-induced torque on video capsule endoscopes. A: Magnetically induced torque on video capsule endoscopes (VCEs) without an integrated permanent magnet (left; measured) and with an integrated permanent magnet (right; calculated from displacement force); B: Worst-case torque due to gravity on VCEs without (left) and with (right) an integrated permanent magnet. The lengths of the boxes mark the 25^th and 75^th percentiles, the yellow lines indicate the 50^th percentile (median), and the whiskers represent the 1.5-fold interquartile range. Note the different scales of the Y-axis in Figure 9A and B.

Table 3 Magnetically induced torque on the tested video capsule endoscopes.

Video capsule endoscope	Max magnetically induced torque (mNm)	Calculated upper bound of magnetically induced torque (mNm)	Worst-case torque due to gravity (mNm)	T mag < T grav
PillCam SB3	19.2	Not determined	Approximately 0.76	Not fulfilled
OMOM HD	5.1	Not determined	Approximately 0.82	Not fulfilled
MiroCam 1200	41.7	Not determined	Approximately 0.80	Not fulfilled
CapsoCam Plus	51.0	Not determined	Approximately 1.11	Not fulfilled
MiroCam 2000	54.0	Not determined	Approximately 1.07	Not fulfilled
PillCam Colon2	15.6	Not determined	Approximately 0.93	Not fulfilled
OMOM CC	5.1	Not determined	Approximately 0.97	Not fulfilled
NaviCam Colon	Below measuring limit	Not determined	Approximately 0.86	Fulfilled
NaviCam	Not determined	271.95	Approximately 1.23	Not fulfilled
OMOM RC	Not determined	289.02	Approximately 1.75	Not fulfilled
MiroCam Navi	Not determined	293.90	Approximately 1.19	Not fulfilled

For magnet-containing capsules, the upper bound of magnetically induced torque was calculated based on the measured displacement force. Worst-case torque due to gravity is shown for comparison. T mag: Magnetically induced torque; T grav: Gravitational torque.

Due to the strong displacement force and the resulting high torque, the three capsules with an integrated small permanent magnet (NaviCam, OMOM RC, and MiroCam Navi) could not be measured using the pulley method, due to the risk of damage to the VCEs and the test apparatus. These three test objects were measured qualitatively, and the calculated value based on the displacement force method was used for illustration purposes (Table 3).

The median torque was 19.2 mNm (IQR: 10.4-46.4) without a magnet vs 289.0 mNm (IQR: 280.5-291.5) with a magnet, respectively; (exact P = 0.017; Hodges-Lehmann difference 256.3 mNm, 95% confidence interval: 220.9-288.8; rank-biserial r = 1.00). One capsule without a magnet (NaviCam Colon) was below the measurement limit and was thus excluded from this comparison. The median worst-case torque due to gravity was 0.90 mNm (IQR: 0.81-1.00) without a magnet vs 1.23 mNm (IQR: 1.21-1.49) with a magnet, respectively (exact P = 0.012). Determined measurement uncertainty for torque ranged from 15%-20%. No statistically significant deviations were detected between the measurements.

DISCUSSION

Main findings

This study presents the first systematic evaluation of magnetically induced forces on VCEs in a 3-T MRI environment. X-ray-assisted size measures of metal-containing components were performed to indirectly assess the risk of thermal hazards. Potential patient hazards due to active implantable medical devices in an electromagnetic environment, according to ISO/TS 10974: 2018(E)[14], are gradient field-induced device heating or vibration, radio frequency-induced heating, static magnetic field (B₀)-induced force, and torque. Unintended stimulation and device malfunction can be caused by the above mechanisms or their combination.

In this in vitro study, the displacement force and torque due to the permanent static magnetic field of the MRI scanner were tested at the location of the maximal gradient. For all but one capsule (NaviCam Colon), the calculated magnetically induced displacement force and torque exceeded gravity. Hence, the cutoff for considering this constellation MR-safe according to the ASTM F2052-21 and ASTM F2213-17 standards was not fulfilled. These forces were significantly higher for the three gastric capsules with an integrated permanent magnet necessary for external magnetic control.

Clinical context

Our experimental in vitro findings support the existing clinical contraindication to MRI in patients with incorporated VCEs. However, failing to meet conservative ASTM MR safety criteria does not necessarily cause harm to the patient. Four case reports have been published showing no adverse events. In three, the small bowel capsule endoscope had already reached the colon, but an MRI was conducted before excretion of the capsule. Reasons included acute lumbar radiculopathy in one case (study terminated after detection of VCE at scout scan)[7] and medical professionals’ unawareness of possible interactions when ordering both tests for work-up of Crohn’s disease in two others[8,10]. In a fourth patient with stricturing Crohn’s disease, asymptomatic retention of a VCE for two years had not been considered[9]. Double balloon enteroscopy four weeks later for retrieval of the retained capsule and dilation of the stricture reported no damage from MRI forces. In another patient, immediate computed tomography after MRI visualized the capsule in the right colon but showed no tissue alteration[10].

Mechanical damage to the bowel wall with possible perforation represents a major theoretical concern. Although all but one capsule exceeded the safety thresholds defined by their gravitational force, the measured forces (maximum displacement force 2.4 N, maximum torque 0.3 Nm) appear to be substantially lower than those observed during colonoscopy or reported to cause bowel perforation in experimental settings.

For comparison, colon perforations were observed in human cadaveric colons, at mean forces of 53.4 N, ranging from 26.7 N to 89 N[15]. In a Hoken silica colonoscopy training model, the maximum force was 12.73 N[16]. Forces in an ex vivo porcine colonoscopy model exceeded 5 N for an expert endoscopist[17]. In 21 colonoscopies, an externally attached device measured a peak pushing force of 43.2 N and a peak torque of 1.0 Nm[18]. For the small bowel, mean forces of 10.3 ± 2.9 N were required for perforation by a laparoscopic grasper[19].

However, these comparisons should be interpreted with caution as the effect of magnetic forces on a mobile intraluminal VCE may be modified by mucosal contact, peristaltic movements, and luminal friction, thus differing from the forces caused by a colonoscope.

Thermal effects and vibration

Thermal risks of MRI were not directly measured given the small size of the metallic components of the VCEs. Radiographically assessed length was, in all cases, shorter than the 50 mm threshold for 1.5-T scanners; only two VCEs slightly exceeded the 24 mm threshold for 3-T scanners[20]. Furthermore, VCEs have a protective plastic cover, and the metallic components (mainly batteries, printed circuits, and in some cases small permanent magnets) are partially surrounded by air within the capsule[21].

After heating of adjacent thermosensitive tissue (e.g., brain, eyes, neural tissue, testes, and ovaries) by > 2° within 15 minutes, a cooling phase after 1 hour of normal MRI scanning operation should be considered[22]. On the other hand, animal studies in rodents suggest that severe small bowel injury occurs at temperatures between 41.5°C and 44.5°C sustained for 15-60 minutes[23-25].

Possible vibration due to dynamic MRI sequences was not investigated in this study. However, ingestible capsules that intentionally generate mechanical vibration within the bowel have previously been applied as an experimental treatment for constipation in humans and were well tolerated[26,27].

Device malfunction

Device malfunction is less relevant, as incidental MRI scanning of patients with external metallic recording equipment during the maximal recording time for a VCE of 20 hours is very unlikely. This is only a concern for capsules with on-board storage of devices that must be retrieved after excretion. After transmission or image acquisition ends, the capsule becomes an inert passive device.

Strengths and limitations

This study is the first systematic in vitro evaluation of magnetically induced displacement forces and torque on VCEs in a 3-T MRI environment. Tests based on ASTM standards were conducted in a specialized institute for MRI safety and showed good reproducibility. A further strength is the wide coverage of different VCE types with single, double or panoramic cameras, also including small bowel, colon and magnetically controlled gastric capsules.

The major limitation of this study is the in vitro design. Assessment restricted to the effect of static magnetic fields in a single 3-T scanner may not be generalizable to other constellations. Torque on magnet-containing capsules could only be calculated from displacement forces to avoid damage. Magnitude of measured magnetic forces on VCEs is not directly transferrable to possible interaction with human tissue in clinical scenarios. Sequence-dependent radiofrequency-induced effects such as heating or vibration were not measured. Instead, the improbability of possible thermal effects was deducted from the limited length of metallic components of VCEs. Possible vibrations were not assessed.

Implications and future research

Our findings provide a quantitative baseline for the mechanical forces that incorporated VCEs may experience in a 3-T MRI environment. Studies involving different scanners, inclusion of dynamic sequences, and ex vivo models are required. A clinical survey on the outcomes of incidental MRI scans in patients with ingested VCEs is being planned. Given the increasing use of VCEs as first line diagnostic devices for small bowel endoscopy, and as screening tools for colorectal or gastric cancer, future development of video capsules should consider the risks of interference in an electromagnetic environment. Clinicians and MR staff should be aware of the possibility of incorporated video capsules during multimodal testing including VCE and MRI. Furthermore, capsules may be retained proximal to strictures without symptoms even years after ingestion. This is of special importance, as 7-T MR scanners are already available in study settings.

CONCLUSION

In a 3-T MRI environment, all but one VCE experienced displacement forces and torque above accepted safety limits as defined by ASTM standards, although levels remained below those previously associated with bowel perforation, or assessed during colonoscopy. Nevertheless, these findings, along with the growing use of VCEs, should alert engineers and clinicians to this potential safety concern.