Dong-Hwi Choi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Dong-Hwi Choi has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Minkyu Je
Korea Advanced Institute of Science and Technology (KAIST)
3 shared publications - Sohmyung Ha
New York University Abu Dhabi
3 shared publications - Dong-Woo Jee
Ajou University
2 shared publications - Byeongseon Choi
Korea Advanced Institute of Science and Technology (KAIST)
1 shared publication - Christopher Santos
School of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea
1 shared publication - Dongyoon Lee
School of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea
1 shared publication
Latest Publications
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 20 (5) , 720-731
A Bias-Electrode-Free Multichannel ExG Readout IC With Time-Multiplexed PAM-4 Body Channel Communication
We propose a multi-channel ExG recording system integrated with body channel communication (BCC), targeting wearable multimodal human-machine interfaces, such as extended reality (XR) and brain-computer interfaces (BCIs). To suppress noise from the BCC transmitter that appears as common-mode interference (CMI) at the ExG analog front end (AFE), we use time-multiplexed acquisition scheme that allocates separate time slots for ExG recording and BCC transmission. This temporal separation preserves ultra-low-noise ExG acquisition, enabling reliable recording of EEG and EOG in addition to EMG and ECG. Because the BCC transmitter must transmit multi-channel ExG data within a limited slot, it requires a high data rate. Using PAM-4 signaling, the proposed system achieves a 10-Mbps data rate. This modulation scheme is enabled by a bias-electrode-free ExG AFE, which increases the BCC signal amplitude by approximately 2.13×, thereby providing sufficient amplitude margin for reliable PAM-4 level discrimination. In addition, a charge-pump-based CMI cancellation loop incorporating a multi-channel least-mean-square (LMS) filter, together with a DC servo loop, mitigates inter-channel mismatch and suppresses differential-mode artifacts. Fabricated in 180-nm CMOS, the chips achieve a total common-mode rejection ratio (CMRR) of 101.1 dB and suppress electrode DC offsets up to 500 mV, while maintaining stable operation under 18-Vpp common-mode interference and 15% electrode mismatch.
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 20 (5) , 720-731
A Bias-Electrode-Free Multichannel ExG Readout IC With Time-Multiplexed PAM-4 Body Channel Communication
We propose a multi-channel ExG recording system integrated with body channel communication (BCC), targeting wearable multimodal human-machine interfaces, such as extended reality (XR) and brain-computer interfaces (BCIs). To suppress noise from the BCC transmitter that appears as common-mode interference (CMI) at the ExG analog front end (AFE), we use time-multiplexed acquisition scheme that allocates separate time slots for ExG recording and BCC transmission. This temporal separation preserves ultra-low-noise ExG acquisition, enabling reliable recording of EEG and EOG in addition to EMG and ECG. Because the BCC transmitter must transmit multi-channel ExG data within a limited slot, it requires a high data rate. Using PAM-4 signaling, the proposed system achieves a 10-Mbps data rate. This modulation scheme is enabled by a bias-electrode-free ExG AFE, which increases the BCC signal amplitude by approximately 2.13×, thereby providing sufficient amplitude margin for reliable PAM-4 level discrimination. In addition, a charge-pump-based CMI cancellation loop incorporating a multi-channel least-mean-square (LMS) filter, together with a DC servo loop, mitigates inter-channel mismatch and suppresses differential-mode artifacts. Fabricated in 180-nm CMOS, the chips achieve a total common-mode rejection ratio (CMRR) of 101.1 dB and suppress electrode DC offsets up to 500 mV, while maintaining stable operation under 18-Vpp common-mode interference and 15% electrode mismatch.
- Open Access
- 10 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Solid-State Circuits Society
- Vol. 5, 324-334
An Implantable Peripheral Nerve Interface System With Simultaneous Multisite Data Transmission via Galvanically Coupled Body-Channel Communication for Next-Generation Prosthetic Limbs
This paper presents a neural implant system for prosthetic limbs that enables simultaneous multi-site data transmission via galvanically coupled body-channel communication (GC-BCC) for fine, low-latency control. Ultrasound and RF uplinks typically require multiple-access schemes to avoid collisions across implants. Capacitive BCC can support simultaneous links, but the small, motion-variant return-path capacitance reduces channel gain and undermines robustness. In contrast, GC-BCC forms a closed-loop current path for each transmitter.receiver pair via a dedicated termination, enabling simultaneous uplinks without multiple-access scheme. Implemented in 180-nm CMOS, the system integrates a 16-channel neural recording analog front end, a GC-BCC transceiver, and low-dropout regulators. Through 15-mm porcine tissue, the link sustains a 20-Mbps data rate with a bit-error rate below 1.3×105. Two GC-BCC transceivers operate simultaneously with 100-Ω termination with negligible crosstalk at ≥1.6-cm spacing. The measured energy efficiencies are 20.35 pJ/bit for the transmitter and 467.4 pJ/bit for the receiver.
- 10 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Solid-State Circuits Society
- Vol. 5, 324-334
Open AccessAn Implantable Peripheral Nerve Interface System With Simultaneous Multisite Data Transmission via Galvanically Coupled Body-Channel Communication for Next-Generation Prosthetic Limbs
This paper presents a neural implant system for prosthetic limbs that enables simultaneous multi-site data transmission via galvanically coupled body-channel communication (GC-BCC) for fine, low-latency control. Ultrasound and RF uplinks typically require multiple-access schemes to avoid collisions across implants. Capacitive BCC can support simultaneous links, but the small, motion-variant return-path capacitance reduces channel gain and undermines robustness. In contrast, GC-BCC forms a closed-loop current path for each transmitter.receiver pair via a dedicated termination, enabling simultaneous uplinks without multiple-access scheme. Implemented in 180-nm CMOS, the system integrates a 16-channel neural recording analog front end, a GC-BCC transceiver, and low-dropout regulators. Through 15-mm porcine tissue, the link sustains a 20-Mbps data rate with a bit-error rate below 1.3×105. Two GC-BCC transceivers operate simultaneously with 100-Ω termination with negligible crosstalk at ≥1.6-cm spacing. The measured energy efficiencies are 20.35 pJ/bit for the transmitter and 467.4 pJ/bit for the receiver.
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 20 (2) , 169-179
A Neural Recording IC for 64-Channel Time-Multiplexed MEA with 3.3-G$\Omega$ Total Input Impedance Using Dual Positive Feedback Loop $Z_{\textit{IN}}$-Boosting
This paper presents a 64-channel time-domain multiplexed (TDM) neural recording IC that achieves a high total input impedance (T-$Z_{\textit{IN}}$) for direct interfacing with a time-multiplexed microelectrode array (MEA). Unlike conventional IC-side multiplexing implementations, the proposed system performs multiplexing at the electrode side, creating a shared external parasitic path across channels and allows the dual positive feedback loop (DPFL) to use shared feedback capacitors and a single calibration code. The DPFL cancels both internal and external parasitics, thereby boosting T-$Z_{\textit{IN}}$. Thus, the proposed scheme eliminates parasitic mismatch and improves scalability and T-$Z_{\textit{IN}}$ than prior works. Fabricated in a 180 nm CMOS process, the system implements 8 to 1 multiplexing per analog front end, achieves 3.3 G$\Omega$ T-$Z_{\textit{IN}}$ at 10 Hz with 3 pF added external capacitance, and demonstrates saline-based spike recording with 6.66 $\mu$V${}_{\mathrm{RMS}}$ input referred noise over 1 Hz to 10 kHz, while consuming 8.87 $\mu$W per channel and 0.0619 mm${}^{2}$ per channel.
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 20 (2) , 169-179
A Neural Recording IC for 64-Channel Time-Multiplexed MEA with 3.3-G$\Omega$ Total Input Impedance Using Dual Positive Feedback Loop $Z_{\textit{IN}}$-Boosting
This paper presents a 64-channel time-domain multiplexed (TDM) neural recording IC that achieves a high total input impedance (T-$Z_{\textit{IN}}$) for direct interfacing with a time-multiplexed microelectrode array (MEA). Unlike conventional IC-side multiplexing implementations, the proposed system performs multiplexing at the electrode side, creating a shared external parasitic path across channels and allows the dual positive feedback loop (DPFL) to use shared feedback capacitors and a single calibration code. The DPFL cancels both internal and external parasitics, thereby boosting T-$Z_{\textit{IN}}$. Thus, the proposed scheme eliminates parasitic mismatch and improves scalability and T-$Z_{\textit{IN}}$ than prior works. Fabricated in a 180 nm CMOS process, the system implements 8 to 1 multiplexing per analog front end, achieves 3.3 G$\Omega$ T-$Z_{\textit{IN}}$ at 10 Hz with 3 pF added external capacitance, and demonstrates saline-based spike recording with 6.66 $\mu$V${}_{\mathrm{RMS}}$ input referred noise over 1 Hz to 10 kHz, while consuming 8.87 $\mu$W per channel and 0.0619 mm${}^{2}$ per channel.
- 21 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 58 (10) , 2757-2766
A 1984-Pixels, 1.26 nW/Pixel Retinal Prosthesis Chip With Time-Domain In-Pixel Image Processing and Bipolar Stimulating Electrode Sharing
This article presents a 1984-pixel retinal prosthesis (RP) chip with time-domain in-pixel image processing. The proposed time-domain image processing (TDIP) algorithm efficiently generates the edge-extracted stimulation and significantly reduces the stimulus power consumption. For the area-and power-efficient time-domain processing, a leakage current-based pulsewidth comparator (PWC) and its control scheme are also proposed. The proposed RP chip adopts the bipolar stimulation-based local return that minimizes image dispersion and a sequence control scheme that produces 1984 stimulation points with 1024 electrodes. The RP chip was implemented in a 180 nm CMOS process and achieved a power consumption of 1.26 nW/pixel, which is $\times$ 44.7 better than the previous state-of-the-art image processing RP chip.
- 21 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 58 (10) , 2757-2766
A 1984-Pixels, 1.26 nW/Pixel Retinal Prosthesis Chip With Time-Domain In-Pixel Image Processing and Bipolar Stimulating Electrode Sharing
This article presents a 1984-pixel retinal prosthesis (RP) chip with time-domain in-pixel image processing. The proposed time-domain image processing (TDIP) algorithm efficiently generates the edge-extracted stimulation and significantly reduces the stimulus power consumption. For the area-and power-efficient time-domain processing, a leakage current-based pulsewidth comparator (PWC) and its control scheme are also proposed. The proposed RP chip adopts the bipolar stimulation-based local return that minimizes image dispersion and a sequence control scheme that produces 1984 stimulation points with 1024 electrodes. The RP chip was implemented in a 180 nm CMOS process and achieved a power consumption of 1.26 nW/pixel, which is $\times$ 44.7 better than the previous state-of-the-art image processing RP chip.
- 16 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 15 (6) , 1140-1148
A 4.49nW/Pixel Light-to-Stimulus Duration Converter-Based Retinal Prosthesis Chip
This paper presents a 288-pixel retinal prosthesis (RP) chip implemented in a 0.18 μm CMOS process. The proposed light-to-stimulus duration converter (LSDC) and biphasic stimulator generate a wide range of retinal stimuli proportional to the incident light intensity at a low supply voltage of 1V. The implemented chip shows 25.5 dB dynamic stimulation range and the state-of-the art low power consumption of 4.49 nW/pixel. Ex-vivo experiments were performed with a mouse retina and patch-clamp recording. The electrical artifact recorded by the patch electrode demonstrates that the proposed chip can generate electrical stimuli that have different pulse durations depending on the light intensity. Correspondingly, the spike counts in a retinal ganglion cell (RGC) were successfully modulated by the brightness of the light stimuli.
- 16 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 15 (6) , 1140-1148
A 4.49nW/Pixel Light-to-Stimulus Duration Converter-Based Retinal Prosthesis Chip
This paper presents a 288-pixel retinal prosthesis (RP) chip implemented in a 0.18 μm CMOS process. The proposed light-to-stimulus duration converter (LSDC) and biphasic stimulator generate a wide range of retinal stimuli proportional to the incident light intensity at a low supply voltage of 1V. The implemented chip shows 25.5 dB dynamic stimulation range and the state-of-the art low power consumption of 4.49 nW/pixel. Ex-vivo experiments were performed with a mouse retina and patch-clamp recording. The electrical artifact recorded by the patch electrode demonstrates that the proposed chip can generate electrical stimuli that have different pulse durations depending on the light intensity. Correspondingly, the spike counts in a retinal ganglion cell (RGC) were successfully modulated by the brightness of the light stimuli.