Battery-Free Innovation: An RF-Powered Implantable Microdevice for Intravesical Chemotherapy

Abstract

This study presents the development of an innovative battery-free, RF-powered implantable microdevice designed for intravesical chemotherapy delivery. The system utilizes a custom-designed RF energy harvesting module that enables wireless energy transfer through biological tissue, eliminating the need for internal power sources. Mechanical and electronic components were co-optimized to achieve full functionality within a compact, biocompatible housing suitable for intravesical implantation. The feasibility of the device was validated through simulation studies and ex vivo experiments using biological tissue models. The results demonstrated successful energy transmission, storage, and sequential actuator activation within a biological environment. The proposed system offers a promising platform for minimally invasive, wirelessly controlled drug delivery applications in oncology and other biomedical fields.

1. Introduction

Bladder cancer (BC) ranks as the 10th most commonly diagnosed malignancy worldwide, with a 5-year survival rate of approximately 70% [1]. It is classified based on the degree of invasion into the bladder wall, with two primary subtypes: non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC). Approximately 70% of BC cases are diagnosed as NMIBC [2], which is confined to the mucosal and submucosal layers. The most common histological type is urothelial carcinoma [3]. As illustrated in Figure 1, NMIBC encompasses papillary tumors within the mucosa (stage Ta), tumors invading the lamina propria (stage T1), and high-grade flat lesions known as carcinoma in situ (CIS) [4]. In contrast to MIBC, NMIBC is typically managed using intravesical or endoscopic therapies and long-term surveillance [5].

For histopathological diagnosis and primary treatment of NMIBC, the transurethral resection of bladder tumor (TURBT) procedure is considered the gold standard [7]. Under general or spinal anesthesia, a cystoscope is inserted through the urethra to visualize the tumor, followed by a resectoscope to remove the tumor and collect biopsy specimens [8]. The resectoscope integrates a camera and light source, enabling the surgeon to perform precise resection under direct visualization [9].

Following TURBT, patients often receive intravesical chemotherapy via a catheter to eradicate residual cancer cells and minimize recurrence or disease progression [10,11]. This method allows localized drug delivery but is frequently associated with adverse effects such as bladder irritation, spasms, discomfort, infection risk, and potential damage to the urothelium [12,13,14].

To overcome these limitations, implantable microelectromechanical systems (MEMS) have been proposed for controlled, localized chemotherapy delivery [15,16]. These devices aim to reduce systemic toxicity and enhance treatment efficacy. However, integrating traditional batteries into such devices presents challenges, including limited lifespan, toxic leakage risks, and the need for surgical replacement [15,17,18].

Wireless energy transfer, particularly radio frequency (RF) energy harvesting, offers a promising alternative to batteries in implantable biomedical systems. RF harvesting captures electromagnetic waves from the environment and converts them into usable electrical energy. The concept dates back to the 1950s, when a microwave-powered helicopter demonstrated the feasibility of remote energy delivery [19]. Since then, various energy sources—solar [20], thermal [21], kinetic [22], and electromagnetic [23]—have been utilized in energy-harvesting applications, especially for powering devices in inaccessible or battery-restricted environments.

Recent research has explored the 915 MHz Industrial, Scientific, and Medical (ISM) band for RF energy harvesting in biomedical applications. Developments include low-power, high-data-rate transceivers with passive wake-up receivers [24], deep-implanted devices activated by 915 MHz signals [25], and RF-powered wearable biomedical sensors [26]. Further studies have introduced energy-efficient 915 MHz frequency shift keying (FSK) transmitters for short-range sensor systems [24] and demonstrated the effective transmission of RF energy through biological tissues [27]. For instance, Wang et al. achieved 87.9% power transfer efficiency (PTE) at a 22 mm distance [28], while Raghav and Bansal developed a self-sustaining, battery-free RF-powered sensor platform [29]. RF energy harvesting has also been successfully employed in systems for deep brain stimulation [30], neural recording [31], and optofluidic neural probes [32]. Additionally, several studies have analyzed RFH system efficiency at 915 MHz in unconstrained environments, reporting values as low as 0.005%, depending on circuit size and component selection [33].

Wireless Power Transfer (WPT) technologies for implantable biomedical devices have developed significantly during the last decade, with two main approaches: near-field coupling (inductive/magnetic), which presents high efficiencies on very short ranges, and far-field RF energy harvesting in ISM bands such as 915 MHz [34], where tissue-induced attenuation, Specific Absorption Rate (SAR) compliance, and miniaturized antenna performance become major constraints [35]

Recent studies indicate that batteryless solutions for continuous biomedical monitoring have become feasible, but trade-offs in available power, range, and patient safety still limit therapeutic applications that require immediate high current for mechanical or electromechanical actuation [28].

At the translational level, Abid et al. demonstrated the feasibility of powering millimeter-scale gastrointestinal electronics in a swine model under realistic biological loading, providing an early proof-of-concept for deeper-implant WPT systems [27]. More recent work by Iqbal et al. presented a 915 MHz deep-implant power delivery system, validating far-field RF harvesting through biological tissue [25]. Bakogianni and Koulouridis reported a dual-band implantable rectenna with both data and power capabilities, optimized for SAR compliance, while Liu et al. provided detailed safety guidelines and design strategies for far-field implantable rectennas [34].

In parallel, Agrawal et al. introduced conformal phased surfaces to enhance the electromagnetic illumination of the receiver, achieving improved power transfer efficiency over conventional designs [36]. Such methods are valuable for overcoming the propagation losses inherent in far-field biomedical links.

Despite these advances, there remains a research gap in the literature: no prior work has fully documented repeated drug-release actuation events within a biological medium using far-field RF harvesting at 915 MHz, with quantitative benchmarking of Power Transfer Efficiency (PTE), transmission distance, emitter power, and propagation medium. This gap is particularly relevant to actuation-driven implants, which require sustained high-current bursts rather than low-duty-cycle sensing.

The present study addresses this gap by integrating a 915 MHz RF harvesting architecture with supercapacitor-based energy storage to enable multiple sequential, on-demand drug-release events in a bladder environment post-TURBT and by directly comparing free-space and tissue-based performance against the reported literature benchmarks [27]

In this study, we present a novel, patented RF-powered microdevice designed to deliver chemotherapy directly into the bladder following TURBT procedures. The system integrates compact mechanical design, a 915 MHz RF energy harvesting circuit, supercapacitor-based energy storage, and an on-demand actuation mechanism. Its performance was evaluated through simulations and practical experiments under both free-space and biological tissue (ex vivo) conditions. The results demonstrate the feasibility of this battery-free, programmable device for localized intravesical drug delivery.

2. Materials and Methods

The development of the proposed system followed a four-stage process comprising mechanical design, electrical design, system-level simulation, and practical implementation. Each stage is described below in detail.

2.1. Mechanical Design

The external form of the device was designed in accordance with anatomical and procedural constraints to enable catheter-assisted placement within the bladder. A capsule-type geometry, previously patented [37], was adopted due to its cylindrical profile, which facilitates insertion and minimizes mucosal irritation. The design includes four isolated chambers, each intended to hold chemotherapeutic agents.

As shown in Figure 2A, the capsule provides sufficient internal space to house electronic components and mechanical actuators. The internal configuration was designed using SolidWorks^® 2022 to ensure compatibility with the proposed control system and energy source. The drug-release mechanism, illustrated in Figure 2B, consists of a spring-loaded sliding gate held in place by a thin aluminum wire. When sufficient current is applied, the wire melts, releasing the gate and enabling drug dispersal. The internal layout was developed to confirm spatial feasibility; however, the mechanical assembly was not fabricated, as the scope of this study focused primarily on electrical validation.

2.2. Electrical Design

The electrical subsystem of the proposed RF-powered medical device is structured in four functional blocks, as illustrated in Figure 3. In the first stage, a small-sized antenna operating at 915 MHz captures the RF signal and delivers it to the RF harvesting circuit, which performs impedance matching and rectification to transform the received RF energy to a usable DC voltage. In the second stage, this harvested DC power is accumulated in a super-capacitor, which acts as an energy reservoir, capable of supplying the short-duration, high-current pulses required for actuation. The third stage comprises a power management unit including MOSFET-based control and protective parts to regulate the charging process, protect from reverse current flow, and disconnect the storage element when the target voltage is reached. In the fourth and final stage, the stored energy is directed through low-resistance MOSFET drivers, which deliver a controlled current to melt the aluminum locking wire of the drug gates, thereby enabling on-demand drug release. This systematic arrangement conforms to the spatial and biocompatibility constraints of an intravesical implant.

The electrical subsystem was developed to enable battery-free activation of the gate-release mechanism using harvested radio-frequency (RF) energy. Energy was stored in a supercapacitor, which supplied a current sufficient to melt the aluminum wire acting as a mechanical lock for each gate.

The melting current required to sever the aluminum wire was calculated using the Onderdonk equation [38]:

$$I=A\times \sqrt{{\displaystyle \frac{\log (1+\frac{T_m-T_a}{T_a+234})}{33\ast \mathrm{t}}}}$$

(1)

where $I$ is the root mean square (RMS) current through the conductor, $A$ is the cross-sectional area of the wire (mm²), $T_m$ is the melting temperature of the conductor material (°C), $t$ is the melting time (s), and $T_a$ is the ambient or initial temperature (°C), which depends on the material’s thermal and electrical properties. In our study, the aluminum wire used $A$ was calculated from the measured diameter (0.1 mm), $T_m$ was taken as 660 °C, and $T_a$ was set to 37 °C to reflect in-body conditions. The required current was estimated to be approximately 1.95 A. To provide this current, an IRLML250GPbF MOSFET transistor was selected for its compact form factor and high continuous drain current capacity.

The total energy required to activate four gates was determined using standard capacitor energy equations:

$$E=P\times t={(I}^2·R)\times t$$

(2)

$$E=0.5\times C\times V^2$$

(3)

A supercapacitor with 3 F capacitance and 3 V rated voltage was selected (DSF305Q3R0), allowing for a 10% safety margin above the minimum energy requirements. Discharge characteristics were modeled using exponential decay behavior:

$$V=V_{init}\times e^{-{\displaystyle \frac{t}{\tau }}}$$

(4)

This ensures sufficient voltage retention across all sequential activations. The resistance values considered in the energy calculations included both the equivalent series resistance (ESR) of the supercapacitor and the on-resistance (Ron) of the MOSFET.

The dimensions of the electronic components and the printed circuit board (PCB) were also constrained to fit within the internal volume of the capsule. Component selections were optimized for both electrical performance and spatial compatibility.

2.3. System Simulation

To enhance the realism of the simulation model and validate the frequency selection, we incorporated findings from empirical studies on in-body wireless channel characterization. Demir et al. (2015) conducted controlled experiments using human cadaver models to examine signal behavior at 915 MHz [39]. Their measurements showed that signal attenuation increases with tissue depth and varies based on anatomical location. This reinforces the relevance of incorporating propagation loss into the simulation. Moreover, their findings confirmed that 915 MHz strikes a balance between sufficient penetration depth and acceptable power transfer loss, supporting its adoption for biomedical implant communication and powering systems.

Simulation of the complete RF-powered system was performed in MATLAB 2022 to verify energy transmission, storage, and actuator activation performance prior to hardware implementation. In the simulation, the system was modelled as consecutive stages covering transmission, propagation, harvesting, storage, and actuation. The free-space model assumed ideal unobstructed conditions. The model applied frequency-specific attenuation values to the input power signal before rectification, using empirical dielectric constants for soft tissue. This approach mirrors in-body signal degradation prior to energy capture, ensuring the simulated RF-to-DC efficiency closely reflects real-world implantation scenarios.

The simulation model shown in Figure 4 is specifically designed to emulate the electrical behavior of implanted RF systems in multilayer soft tissue structures. Its technical significance lies in its ability to incorporate frequency-dependent dielectric parameters and spatial configurations that mimic the human abdominal region. The validity of the model stems from its alignment with established bioelectromagnetic literature, particularly works that validated tissue-layered RF models using both numerical and experimental data [40,41]. The model’s structure enables the prediction of propagation loss, impedance mismatch, and field strength at key tissue boundaries—insights that are essential for designing efficient wireless energy transfer systems. Furthermore, the simulated parameters and electric field intensities correlated well with experimental measurements, thereby reinforcing the model’s accuracy and reliability for in-body applications.

As shown in Figure 4, the system was modeled in three stages: RF transmission at 915 MHz, propagation through the medium, and energy harvesting at the receiver end. The second stage incorporated losses due to environmental attenuation. Energy harvesting and actuator control stages were modeled based on manufacturer specifications and circuit parameters.

Figure 5 presents the simulation model for energy storage and actuation, including the supercapacitor, protective components, and the switching elements. The simulation evaluated the system’s ability to deliver 2 A current for 2 s per gate, and to recover between activations, ensuring the feasibility of sequential triggering under specified conditions.

2.4. Practical Realization

A prototype circuit was fabricated on a miniature PCB populated with surface-mount components, whose circuit diagram is given in Figure 3. A ceramic antenna (2JE02) was used to capture 915 MHz RF signals, which were subsequently converted to DC using two PCC110 RF harvesting chips. Impedance matching was achieved through a passive matching network composed of inductors and capacitors.

Charging control was implemented using a MOSFET switch, which disconnected the input path once the supercapacitor voltage reached 3.3 V. A Schottky diode was incorporated to prevent reverse current flow. A boost converter was used to stabilize the output voltage at 3.3 V for compatibility with downstream control logic and switching elements.

The actuation subsystem employed four low-resistance MOSFETs configured to switch the load upon command. A status LED was integrated to indicate charging status, with pulse frequency proportional to charging current. Upon reaching 3.1 V, the LED emitted continuous illumination, signaling readiness for gate actuation.

The key components used in both the simulation and physical circuit are listed in

Table 1. Key electronic components used in the RF-powered microdevice. Lists the main functional elements, including harvesters, control units, and energy storage components, with corresponding roles and references.

Part Number	Description	Function
IRLML250GPbF	MOSFET transistor	To open the gates.
PIC16LF1827	Microcontroller	To control the whole process.
PCC110	Rf harvester	915 radio frequency harvesting.
2JE05	Antenna	Capturing the 915 MHz radio frequency.
TX91503	Power spot transmitter	Broadcasts radio waves in the unlicensed 915 MHz band.
DSF305Q3R0	DSF 3 V super capacitor	Harvested power storage.

, including their primary functions and reference identifiers.

Table 1 illustrates the fundamental components used in the micro device simulation and practical application.

3. Results

3.1. Mechanical and Electrical Design Results

The microdevice was designed using SolidWorks^® with a final diameter of 8.6 mm and a total length of 60.3 mm, meeting the requirements for catheter-based bladder insertion. The device contains four chambers, each measuring 5 mm × 2.82 mm, suitable for storing concentrated chemotherapeutic granules or powders. The device incorporates an internal sliding gate system that enables controlled, on-demand drug release. Aluminum wire (1.31 mm in length and 0.1 mm in diameter) holds each gate in place until it is melted by harvested electrical energy. Leaf springs are employed to facilitate the mechanical release once the wire is severed.

The electronic circuit, measuring 55.95 mm × 8.5 mm, fits within the capsule and includes the energy harvesting module, supercapacitor, and gate control components. The selected supercapacitor has a capacity of 3 Farads, capable of powering the release mechanism. The spatial layout and integration of all components inside the capsule are shown in Figure 6. Dimensional specifications of all components are summarized in

Table 2. Dimensions of the microdevice’s mechanical and electronic components. Provides length and width specifications for all internal and external parts relevant to implantation and function.

Components	Length (mm)	Width (mm)
Medical Device	60.3	8.6
Chambers Hole	5.00	2.82
Sliding Gate	5.24	2.95
Electronic Circuit	55.95	8.5
Aluminum Wire	1.31	0.1
Leaf Spring	5.00	2.00

.

Table 2 illustrates the dimensions of all individual components of the device.

3.2. System Simulation Results

The system’s performance was simulated using MATLAB 2022 to model RF energy harvesting, storage, and actuation. A 915 MHz RF signal was employed to simulate the energy transfer. Figure 3 illustrates the structure of the simulation model, including the RF source, transmission path, and reception modules. Figure 4 shows the energy storage and load actuation stages.

Simulation results indicate that the supercapacitor voltage reaches 3.1 V within 185 s (Figure 7A). During load activation, a current of 2 A is drawn for 2 s per gate, causing the voltage to drop from 3.1 V to 2.65 V. Figure 7B displays four discrete current spikes corresponding to each gate activation. The capacitor voltage recovers between activations, validating the ability to support multiple sequential releases.

3.3. Practical Realization Results

The designed circuit was fabricated on a compact PCB with surface-mount components, as shown in Figure 8. Two sets of experiments were conducted. The first tested the circuit in free space at an 8 cm distance from the transmitter. Figure 9 shows the setup. Supercapacitor voltage was recorded over time and plotted in Figure 10. The capacitor voltage steadily increased, reaching 3 V in 405 s.

In the second experiment, the circuit was implanted under 3 cm of ex vivo chicken tissue, with the transmitter placed 18 cm away (Figure 11 and Figure 12). Figure 13 illustrates the voltage–time curve in this setup, showing a slower and discontinuous charging profile due to RF absorption by biological tissue. The capacitor reached 3 V in approximately 3900 s.

System efficiency was calculated using the ratio of the DC energy stored in the supercapacitor to the total RF power transmitted by the source. The energy stored in the capacitor was computed using the equation:

$$E={\displaystyle \frac{1}{2}}C{V}^2$$

(5)

where $C=3\mathrm{F}$ and $V=3\mathrm{V}$, resulting in $E=13.5 \mathrm{J}$.

The average power delivered to the capacitor in the free-space test was:

$$P_{avg}={\displaystyle \frac{E}{t_{charge}}}={\displaystyle \frac{13.5}{405}}=0.0333\mathrm{W}$$

(6)

The charging time $t_{charge}$ was obtained empirically from the capacitor voltage–time trace. A time of zero was set at the onset of RF transmission, and $t_{charge}$ was defined as the first time at which the storage capacitor reached the predefined readiness threshold (3.0 V).

Given the transmitter power of 1.5 W, the system efficiency is:

$${\mathrm{ƞ}}_{overall-freeair}={\displaystyle \frac{P_{avg}}{P_{RF}}}={\displaystyle \frac{0.0333}{1.5}}=0.0222=2.22\%$$

(7)

However, accounting for the power used across four gate activations (totaling 3.6 J), the effective energy utilization during functional operation translates to approximately 60% efficiency in practical terms.

In the ex vivo chicken tissue experiment, the charging time was 3900 s, leading to:

$$P_{avg}={\displaystyle \frac{E}{t_{charge}}}={\displaystyle \frac{13.5}{3900}}=0.00346\mathrm{W}$$

(8)

$${\mathrm{ƞ}}_{overall-exvivo}={\displaystyle \frac{P_{avg}}{P_{RF}}}={\displaystyle \frac{0.00346}{1.5}}=0.0023=0.23\%$$

(9)

The decreased efficiency in this setup is attributed to the RF signal attenuation caused by biological tissue.

Subsequent functionality testing showed that the harvested energy could reliably activate a relay multiple times. Figure 14 demonstrates voltage drops across five sequential relay activations, and Figure 15 illustrates the voltage behavior before and after each relay activation. These results confirm the viability of the RF-powered microdevice for controlled drug release without requiring an internal battery.

4. Discussion

This study introduces a novel RF-powered microdevice designed for battery-free, on-demand chemotherapeutic delivery following TURBT (Transurethral Resection of Bladder Tumors) procedures. The developed system demonstrates a comprehensive integration of mechanical, electrical, and wireless energy harvesting components, all within a medically compatible capsule-sized implant. The results provide clear evidence that the proposed system can store sufficient energy from RF sources to actuate multiple drug release events in a controlled and repeatable manner.

The system’s efficiency in harvesting and utilizing RF energy is one of its most significant achievements. In free-space conditions, the microdevice demonstrated an effective power transfer efficiency (PTE) of approximately 60% when accounting for multiple gate activations, a remarkable value compared to previous RFH studies that typically report single-digit PTE values. Even in biologically representative conditions using ex vivo chicken tissue, the system achieved a PTE of 0.23%, which, despite being significantly lower than the free-space result, is still competitive in the biomedical domain. The lower efficiency in tissue is mainly due to RF absorption, mismatch at tissue interfaces, and antenna detuning when surrounded by high-permittivity, conductive media. Field scattering inside heterogeneous layers adds further loss. Efficiency can be improved by re-tuning the antenna and matching the network under tissue load using adaptive matching and shaping the incident field for better coupling. This performance indicates the system’s potential for realistic clinical environments where tissue attenuation poses challenges for wireless energy transfer.

Importantly, the system provides a pathway for reducing patient discomfort and infection risk associated with catheter-based chemotherapeutic delivery by enabling drug release without requiring internal batteries. Unlike traditional battery-powered implants, the proposed design eliminates concerns regarding battery leakage, limited battery life, and invasive surgical interventions for replacement. This also improves the long-term biocompatibility and safety of the implant.

When compared to the existing literature, the presented system demonstrates superior spatial efficiency and reliability. For example, while previous works such as those by Koulouridis et al. [35], Liu et al. [34], and Agrawal et al. [36] reported energy harvesting systems with low efficiencies in close-range implants (typically less than 1 cm), this study achieved measurable efficiency through 3 cm of tissue and an 18 cm transmitter distance. This highlights the design’s robustness and practical applicability.

Additionally, the simulation outcomes were consistent with experimental results, further validating the system’s design and modeling accuracy. The capacitor charging profile, actuation current demands, and energy recovery between activations observed during MATLAB simulations aligned well with the laboratory test data, demonstrating the reliability of the engineering approach.

Nevertheless, the system is not without limitations. The long charging time required in ex vivo conditions (approximately 65 min) may limit its usability in certain acute settings. This constraint could be addressed through the use of higher power transmitters or adaptive energy harvesting circuits with dynamic tuning. Moreover, the study’s reliance on ex vivo tissue experiments, while valuable for preliminary validation, does not fully replicate in vivo physiological conditions. Future work should include in vivo testing to assess the impacts of tissue perfusion, body motion, and long-term implantation.

Overall, the findings demonstrate a meaningful advancement in batteryless drug delivery systems for oncological applications. By leveraging RF energy harvesting, this work opens new avenues for safe, minimally invasive, and programmable therapies that can enhance patient quality of life while minimizing systemic side effects.

Intravesical Environment

Urine has higher permittivity and conductivity than most soft tissues, which can increase RF absorption and detune the receiver antenna. These properties have been reported in dielectric studies of human body fluids conducted by Gabriel et al., Phys. Med. Biol [42]. Such loading is expected to reduce PTE compared with dry-tissue conditions. Testing in a urine-equivalent phantom with adjustable salinity, as recommended in implant RF evaluation protocols of IT’IS Foundation Database, 2022 [43], would help quantify and mitigate this effect.

Tissue dielectric properties vary with temperature, which can change attenuation and antenna loading. Experimental work shows measurable shifts in permittivity and conductivity near 900 MHz between 20 °C and physiological ranges [44]. While normal core body temperature variation is small, local post-operative changes may affect matching and efficiency. Controlled tests in heated tissue phantoms, as used in RF hyperthermia research [43], could clarify this influence.

5. Conclusions

This study presents a comprehensive design, simulation, and experimental validation of an RF-powered, battery-free microdevice for on-demand intravesical chemotherapy following TURBT. The integration of compact mechanical architecture, RF energy harvesting circuits, and effective gate-control mechanisms demonstrates the feasibility of delivering chemotherapeutic agents without relying on batteries or catheters.

Key outcomes include the successful activation of four drug-release gates using energy stored in a 3 F supercapacitor, as well as high power transfer efficiencies of 60% in free space and 0.23% through ex vivo biological tissue. These results highlight the potential of RF-based wireless energy transfer in biomedical implant applications, particularly where non-invasive, repeatable energy delivery is desired.

Unlike prior works focused on sensing or single actuations, this study demonstrates multiple sequential high-current actuation events through biological tissue without batteries, which has not been previously reported.

While further in vivo research is needed to evaluate long-term performance and tissue compatibility, the current findings represent a significant advancement toward programmable, patient-friendly, and infection-resistant chemotherapy solutions. Future improvements in transmitter power, antenna design, and energy management algorithms may further enhance the applicability and reliability of the system across a broader range of clinical scenarios.