Searcharxiv⌕ Search

arXiv subjects

Mali Halac

Publications and source records attributed to Mali Halac.

3 recordsLinked to original sources

HDR-DTCP: High-Dynamic-Range Differential Tissue-Coupled Wireless Powering for Tunable Biphasic Stimulation in Injectable Implants

Next-generation implantable neurostimulators require efficient wireless power delivery in smaller, less invasive form factors, yet variations in tissue coupling, implantation conditions, and stimulation loads can alter received power (or requirements) and degrade energy-harvesting performance. This work presents high-dynamic-range (HDR) differential tissue-coupled powering (DTCP), termed HDR-DTCP, a body-coupled architecture in which noninvasive differential electrodes establish an AC electric field in tissue and a thread-like implant extracts power from the longitudinal potential gradient for on-chip harvesting and stimulation. Distributed coupling along the implant length, rather than receiver cross-section, enables minimally invasive injectable geometries, with harvested power scaling with implant length. A 65-nm complementary metal-oxide-semiconductor (CMOS) prototype integrates a reconfigurable four-stage energy harvester with automatic stage selection and a tunable constant-current biphasic stimulator, achieving 63.5\% peak power-conversion efficiency (PCE) at $-16.3$ dBm and $>50\%$ PCE over a 9-dB input-power range. Preliminary in-vivo validation of DTCP was done with a second application-specific integrated circuit (ASIC) on an injectable flexible substrate, capable of being delivered through a 14-gauge needle adjacent to the sciatic nerve of Sprague-Dawley rats (n=4). External copper-pad DTCP powering evoked repeatable hind-limb twitches and electromyography (EMG) response; ultrasound confirmed sub-millimeter nerve proximity, video-based paw-tip tracking captured the response, and 8-day post-implant dissection verified placement along the sciatic-nerve trajectory. Overall, this work provides an integrated wireless powering, energy-harvesting, and stimulation platform toward fully battery-free injectable neuromodulation implants.

eess.SP↗

Multiscale Voxel Based Decoding For Enhanced Natural Image Reconstruction From Brain Activity

Reconstructing perceived images from human brain activity monitored by functional magnetic resonance imaging (fMRI) is hard, especially for natural images. Existing methods often result in blurry and unintelligible reconstructions with low fidelity. In this study, we present a novel approach for enhanced image reconstruction, in which existing methods for object decoding and image reconstruction are merged together. This is achieved by conditioning the reconstructed image to its decoded image category using a class-conditional generative adversarial network and neural style transfer. The results indicate that our approach improves the semantic similarity of the reconstructed images and can be used as a general framework for enhanced image reconstruction.

cs.CV↗

Functional Protein Structure Annotation Using a Deep Convolutional Generative Adversarial Network

Identifying novel functional protein structures is at the heart of molecular engineering and molecular biology, requiring an often computationally exhaustive search. We introduce the use of a Deep Convolutional Generative Adversarial Network (DCGAN) to classify protein structures based on their functionality by encoding each sample in a grid object structure using three features in each object: the generic atom type, the position atom type, and its occupancy relative to a given atom. We train DCGAN on 3-dimensional (3D) decoy and native protein structures in order to generate and discriminate 3D protein structures. At the end of our training, loss converges to a local minimum and our DCGAN can annotate functional proteins robustly against adversarial protein samples. In the future we hope to extend the novel structures we found from the generator in our DCGAN with more samples to explore more granular functionality with varying functions. We hope that our effort will advance the field of protein structure prediction.

q-bio.BM↗