Human gastric magnet-controlled capsule endoscopy conducted in a standing position: the phase 1 study

From October to December 2018, 31 healthy volunteers, including 11 men and 20 women (age, range 21–65 years), were recruited to the study. They were all absent of any concerning abdominal symptoms. None had taken any medications in the prior 3 months or undergone any abdominal surgery in their lives. Written informed consent was obtained from each participant. The study protocol (registration NO.CHiCTR1800018824) was approved by Shenzhen Sixth People’s Hospital ethics committee. The study wan performed in accordance with the Declaration of Helsinki.

As shown in Fig. 1, our newly developed MCE system consists of four main subsystems: an ingestible capsule endoscope; a guidance magnet robot; a data recorder; and a computer workstation with software for real-time viewing and control (all components from FUZI Technologies Co. LTD, Shenzhen, China). The capsule endoscope is a modification of a standard FUZI Dasen capsule that was approved for clinical examination of the small bowel. The magnetic control system controls the movement of the large magnet in the magnetron control equipment through the magnetron workstation equipped with control software. The large magnet can move in the plane perpendicular to the ground plane at 500x500mm, and can change its attitude at any angle through two rotating axes. Capsule endoscopy contains magnetic dipole. The magnetic control system controls the capsule endoscopy through the traction force of the large magnet to the dipole. The dipole inside the capsule interacts with the magnet outside to control the movement of the capsule, such as translation and rolling. When the big magnet moves, the capsule endoscope can move to the predetermined position together (unless the capsule is blocked by the gastric wall, then try to change the path of the magnet motion so that the capsule can bypass the obstruction). When the big magnet rotates along two rotating axes, the capsule also rotates with the rotation, taking full-angle photographs of the gastric wall in front of the lens. The capsule, which is 27 mm × 12 mm in size, has a single CMOS sensor and a permanent magnet inside its dome [Fig. 1]. (The view depth is 0-50 mm and the view angle is 136⁰ The image resolution is 480 × 480.) The capsule is activated by infrared photoelectricity rather than magnetically, as the original model had been. Its capture rate is four frames per second with a 30~40-min power supply. Under the guidance of a magnetic navigation robot, the capsule can perform six basic movements, including forward, backward, ascending, descending, rotating and flipping motions.

A data recorder with an antenna is integrated on the magnetic navigation robot. Image signals from the capsule are transmitted directly to the antenna of the data recorder, obviating the need for commonly used wire-linked skin contact sensors. The images are displayed and stored in real time on the workstation monitor, simultaneously.

The magnetic navigation robot is shaped like the backplane of the human chest X-ray machine standing upright. The manipulating magnet, which is installed behind the flat plate, can move horizontally, vertically, or in rotary motion under the control of a mechanical manipulating arm. The maximum intensity of the magnetic field generated by the magnetic robot is 200 + 50mT. The capsule can be controlled by the magnetic robot even at a distance of 300 mm.

The computer workstation provides a real-time display and control platform. Capsule movement is manipulated by mode click and a joystick. The operator conducts the examination by clicking the needed action icon. Briefly, the basic motion direction icons are presented in the computer control interface, including raise, lower, move left, move right, and rotate 180° and 360°. An icon click moves the capsule within a 3-cm range. In addition, the control interface has fine tuning icons that provide smaller movements, that is, 15° adjustments and 1 cm movements.

The subjects arrived at the hospital between 8:00 am and 10:00 am after fasting overnight (> 8 h) and drinking 500 ml of warm water shortly after waking in the morning. About an hour before capsule ingestion, 10 ml of simethicone emulsion solution was administrated orally with 200 ml of clear warm water, followed by 80 ml of mucus scavenger (SZ 2000 U and 1.0 g sodium bicarbonate dissolved in 80 ml of water at 30–40 °C) about 15 min before ingestion. At the beginning of each examination, the subject was asked to stand upright with his or her chest and abdomen close to the magnetic control plate, and then to swallow the infrared activated magnet-control capsule endoscope followed by 500 ml of drinking water within 3 min [Fig. 2].

After the capsule reached the stomach, the investigator sat in front of the workstation and directed the capsule navigation by clicking on action icons. First, the investigator identified the gastric anglus, which is used as the orienting point for navigation; the capsule would be guided back to the anglus if it went off course. Second, the subjected was instructed to turn with his or her left hand on his or her head to orient the left side of the body close to the magnetic control plate, at which point the capsule was guided to turn and cruise through the cardia, fundus, greater curvature, and lesser curvature of the gastric body, and then return to the anglus. Third, the subject was instructed to return to the original position facing the magnetic control plate, after which the capsule was guided through the greater curvature and lesser curvature of the antrum and the pylorus, and then to return to the anglus. After completion of the gastric navigation course, the capsule was guided into the duodenum. If gastric distension was unsatisfactory at any time during the examination, the subject was asked to drink additional water (200 ml/time).

The study end points were safety, gastric preparation quality, and feasibility. Evaluations of these indexes were completed by two investigators with substantial endoscopy experience, and the lower index value of the two was selected for further analysis.

Safety was assessed according to the occurrence of adverse events and the tolerability of subjects, as assessed by questionnaire.

Gastric preparation quality included degree of cleanliness and distention. Cleanliness was graded as good (transparent fluid and < 5% of the gastric mucosa obscured), moderate (slightly opaque fluid and/or 5–10% of the gastric mucosa obscured), or poor (opaque fluid and/or > 10% of gastric mucosa obscured. Distention was classified as good (satisfactory gastric distention with a small amount of collapsed gastric folds), moderate (considerable collapse of gastric folds obscuring limited parts of the stomach) or poor (noninflated gastric cavity was not inflated making examination of areas of interest impossible).

Feasibility was evaluated in terms of three end points: maneuverability, visualization, and time. MCE maneuverability was graded as good (followed control signals reliably and moved to targeted anatomical landmarks precisely), moderate (followed control signals and moved towards the target but did not reach it precisely), or poor (failure to follow controls). Visualization of primary anatomical landmarks and the gastric mucosa were assessed. The former was expressed as the discovery rate of the cardia, fundus, body, angulus, antrum, and pylorus of stomach. The latter was graded as good (> 75% of mucosa observed), moderate (50–75% observed), or poor (< 50% observed). The total examination time recorded was the time from the swallowing of the capsule to the end of the examination with the capsule entering the descending portion of the duodenum.

Descriptive statistics were used to describe the demographic characteristics of the participants, quality of gastric preparation, feasibility of the MCE examination, and lesion findings.

All 31 healthy volunteers completed the examination, tolerating it well and accepting it without complaint. There were no reports of discomfort in the preparation process, nor throughout the entire examination. All 31 subjects affirmed that they would be very willing to accept the examination again if necessary. All capsules were discharged without incident within 72 h of the examination. In the week after the examination, follow-up revealed no complaints of abdominal pain, diarrhea, or other abdominal symptoms.

Gastric cleanliness was good in 24 (77.4%) and moderate in 7 (22.6%) of the 31 subjects; there were no cases of poor cleanliness. All 31 subjects (100%) had good distention.