TL;DR
This paper introduces "NeoNet," a novel aluminum thin-film EEG net designed to minimize artifacts in MRI and CT scans of pediatric patients. By leveraging ultra-high-aspect-ratio aluminum nanostructures, NeoNet significantly reduces image distortions and potential safety hazards, enabling concurrent EEG monitoring during critical imaging procedures without compromising signal quality.
What Problem Does This Paper Address?
Continuous electroencephalogram (EEG) monitoring is crucial for diagnosing and managing neurological conditions like seizures in neonates and children. However, traditional EEG electrodes, often made of conductive metals, create significant artifacts and distortions in Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) scans. These artifacts can obscure important diagnostic information, leading to misdiagnosis or delayed treatment. Furthermore, the metallic nature of conventional EEG leads poses a safety risk in the strong magnetic fields of MRI scanners. To obtain diagnostic images, these EEG nets typically need to be removed before scanning and reapplied afterward, which is time-consuming, labor-intensive, and interrupts continuous monitoring, especially critical for fragile pediatric patients. This paper addresses the need for an EEG net that is compatible with MRI and CT, minimizes artifacts, and ensures patient safety, thereby facilitating simultaneous EEG and imaging acquisition.
Methods
The core innovation is the development of the "NeoNet," an EEG net utilizing ultra-high-aspect-ratio aluminum thin film nanostructures. These traces were fabricated with dimensions as small as 30 nm in thickness and up to 50.8 cm in length, achieving an aspect ratio of 17,000:1. This design choice was driven by two key properties of aluminum: its low density, which reduces CT artifacts, and its relatively low conductivity (compared to copper), which helps mitigate MRI artifacts by reducing induced currents.
The fabrication process involved photolithography, etching, and deposition of aluminum thin films onto a flexible polymer substrate. Electrodes were then assembled and waterproofed. The researchers employed numerical simulations using the Finite-Difference Time-Domain (FDTD) method to assess the impact of NeoNet on the B₁ transmit field in a 3 Tesla (T) MRI environment. They compared NeoNet with a standard copper-based EEG net (CuNet), both with and without current-limiting resistors.
To validate their findings, the study included:
- MRI Safety Testing: A child head phantom was used to measure temperature rise during high-power MRI scanning.
- Image Quality Assessment: MRI images from adult and pediatric subjects were acquired with and without NeoNet. These images were then qualitatively assessed by a pediatric neuroradiologist and quantitatively evaluated by two board-certified neuroradiologists using a Likert scale, blinded to the presence or absence of the NeoNet.
- CT Artifact Evaluation: CT images of a head phantom were acquired to assess artifact reduction.
- EEG Recording: EEG data was collected from a pediatric patient with epilepsy using both NeoNet and a commercial HD-EEG net to compare signal quality.
Datasets
The study did not use datasets for training machine learning models. Instead, it relied on:
- Numerical Simulations: A 29-month-old male whole-body voxel model was used for FDTD simulations of MRI B₁ field distortion.
- Phantom Studies: A child head-sized agar phantom was used for MRI safety testing, and a head phantom was used for CT artifact evaluation.
- Human Subject Data: MRI and EEG data were collected from adult and pediatric volunteers and patients. Specifically, MRI data was acquired from three subjects (one adult, two children) and EEG data from one pediatric patient. The study also referenced CT images from the Massachusetts General Hospital (MGH) Picture Archiving and Communicating System (PACS).
Key Insights and Ideas
- Ultra-thin aluminum as an "artifact cloaking" material: The extremely thin (30 nm) aluminum traces, combined with their specific resistivity, act to "cloak" the EEG leads from MRI and CT scanners, minimizing their impact on image quality.
- Low density is key for CT: Aluminum's inherent low density is crucial for reducing streak artifacts in CT scans, a problem often caused by denser metallic implants.
- Resistivity, not just conductivity, matters for MRI: While copper is highly conductive, its high conductivity can lead to significant RF-induced currents and B₁ field distortions. NeoNet's controlled resistivity, achieved through thin-film fabrication, offers a better balance for MRI compatibility.
- Current limiting resistors are less effective than intrinsic material properties: Numerical simulations showed that even with ideal current-limiting resistors, copper traces still induced significant B₁ field distortions. The intrinsic properties of the thin-film aluminum were more effective at reducing these artifacts.
- Simultaneous EEG-MRI is feasible and safe: The study demonstrates that NeoNet allows for high-quality EEG recordings during MRI scans without compromising image quality or posing a significant safety risk (minimal temperature rise).
Key Results
- MRI Simulations: NeoNet predicted a 65% lower B₁ transmit field distortion compared to a copper-based EEG net with ideal current-limiting resistors, and a significantly larger reduction compared to copper without resistors.
- MRI Image Quality: Blinded neuroradiologists found no significant difference in image quality (using a 5-point Likert scale) between scans performed with NeoNet and scans without any EEG net. Both conditions allowed for clear identification of clinical findings like atrophy and malformations.
- CT Artifacts: CT images acquired with NeoNet showed no significant artifacts, unlike those with conventional EEG electrodes which exhibited streak artifacts.
- MRI Safety: The maximum temperature rise during a 30-minute high-power MRI scan with NeoNet was 0.84 °C, well within FDA acceptance criteria for 1-hour normal operation.
- EEG Quality: EEG recordings with NeoNet showed comparable quality to commercial HD-EEG nets, successfully capturing physiological signals and pathological events like interictal epileptiform discharges.
Limitations and Future Work
- Manufacturing Complexity: Reaching the target resistance required designing five loops per trace, indicating a need for further optimization in trace design or material selection. The authors mention the possibility of using other metals like titanium but faced manufacturing constraints.
- Numerical Simulation Scope: The high-frequency lumped circuit model of resistors (including series inductors and parallel capacitors) was not included in the simulations. Additionally, the effect of different EEG trace trajectories was not numerically simulated or validated.
- Electrode Instability: Early prototypes experienced electrical instability with carbon conductive glue, which was later addressed by coating the aluminum layer with chromium and gold.
- Future Work: The paper suggests that the NeoNet has potential for other applications, such as electrical impedance spectroscopy (EIS) in children. Further validation of the numerical simulations with experimental data for different trace trajectories would be beneficial. The authors also highlight the need for further research into the safety of EEG-fMRI in infants and toddlers, building upon their findings.
Summary updated