[Paper Review] On the use of impedance detuning for gastrointestinal segment tracking of ingestible capsules

Erdem Cil, Ícaro V. Soares+5 more|arXiv (Cornell University)|Jul 16, 2022

Wireless Body Area Networks40 references21 citations

TL;DR

This paper proposes using impedance detuning of meandered dipole antennas in ingestible capsules to track gastrointestinal (GI) segment location—stomach, small intestine, or large intestine—by measuring phase and magnitude shifts in the reflection coefficient (S11) due to varying tissue electromagnetic (EM) properties. Experiments in tissue-mimicking liquids and ex vivo porcine models show a minimum phase difference of ~10° between tissues, confirming that phase changes provide sufficient information for real-time, non-invasive capsule localization using off-the-shelf RF components.

ABSTRACT

During their travel through the gastrointestinal tract, ingestible antennas encounter detuning in their impedance response due to varying electromagnetic properties of the surrounding tissues. This paper investigates the possibility of using this impedance detuning to detect in which segment of the gastrointestinal tract - stomach, small intestine, or large intestine - the capsule is located. Meandered dipole antennas operating in the 433 MHz Industrial, Scientific, and Medical Band are designed for this purpose. The antennas conform to the inner surface of 3D-printed polylactic-acid capsules with a shell thickness of 0.6 or 0.4 mm. The impedance response is first optimized numerically in a homogeneous cylindrical phantom with time-averaged electromagnetic properties. The magnitude and the phase of the reflection coefficient are then obtained in different tissues and compared with simulations and measurements. The experimental demonstration is carried out first using tissue-mimicking liquids and then in a recently deceased ex vivo porcine model. The minimum change in the phase between different gastrointestinal tissues was determined to be around 10 degrees in the porcine model, indicating that the changes in the impedance response, particularly the changes in the phase, provide sufficient information to follow the position of the capsule in the gastrointestinal tract.

Motivation & Objective

• To investigate whether impedance detuning of ingestible antennas can be used to determine the GI tract segment (stomach, small intestine, large intestine) where a capsule is located.
• To evaluate the feasibility of using the phase and magnitude of the reflection coefficient (S11) as indicators of tissue-specific electromagnetic (EM) properties.
• To validate the method experimentally using tissue-mimicking liquids and ex vivo porcine models with realistic tissue heterogeneity.
• To demonstrate that phase shifts in S11 are sufficiently distinct between GI tissues to enable reliable localization without additional sensors.

Proposed method

• Design of meandered dipole antennas on a 100-µm Rogers CLTE-MW substrate, conformally integrated into 3D-printed polylactic acid (PLA) capsules with shell thicknesses of 0.6 mm or 0.4 mm.
• Numerical optimization of antenna trace length (16.9 mm for t=0.6 mm, 14.5 mm for t=0.4 mm) in a homogeneous cylindrical phantom with time-averaged GI tissue EM properties (εr=63, σ=1.02 S/m at 433 MHz).
• Simulation of S11 magnitude and phase in numerical phantoms representing stomach, small intestine, and large intestine using tissue-specific EM properties from Table I.
• Experimental validation using tissue-mimicking liquids with EM properties matching GI tissues, followed by ex vivo measurements in a recently deceased porcine model.
• Measurement of S11 magnitude and phase at 434 MHz across five locations: stomach (M1), duodenum (M2), ileum (M3, M4), and colon (M5), with capsules inserted 10 cm deep.
• Use of offset feed (6.3 mm from center) to achieve 50 Ω impedance matching in the in-body environment, improving performance over center-fed designs.

Experimental results

Research questions

• RQ1 Can impedance detuning of ingestible antennas be used to distinguish between different GI tract segments based on changes in S11 magnitude and phase?
• RQ2 What is the minimum measurable phase shift between different GI tissues that could enable reliable localization?
• RQ3 How do realistic tissue heterogeneities and air gaps in ex vivo models affect the detectability of impedance changes compared to homogeneous phantoms or liquids?
• RQ4 Can off-the-shelf, low-power RF components reliably measure the required phase shifts for real-time localization?

Key findings

• The minimum measurable phase difference between different GI tissues in the ex vivo porcine model was approximately 10°, indicating sufficient sensitivity for segment-level tracking.
• Phase shifts in S11 were more distinguishable in ex vivo porcine tissues than in tissue-mimicking liquids, due to tissue heterogeneity and air gaps enhancing impedance variations.
• For both shell thicknesses (0.6 mm and 0.4 mm), the S11 phase exhibited consistent trends: decreasing from M1 (stomach) to M2 (duodenum), and increasing from M4 (ileum) to M5 (colon), confirming measurement stability.
• The magnitude of S11 showed a minimum difference of 0.9 dB between tissues at 434 MHz for t=0.6 mm and 1.1 dB for t=0.4 mm, indicating measurable but less distinct changes than phase.
• The phase response was more sensitive and reliable than magnitude for tissue differentiation, supporting its use as the primary tracking metric.
• The results confirm that phase-based impedance sensing using standard RF components is feasible for real-time, fully wireless capsule localization in the GI tract.