Research Ethics and EEG Principles

The 50Hz noise in EEG recordings arises from the alternating current (AC) used in power lines. The oscillating voltage in power lines creates a magnetic field that induces a small oscillating voltage in nearby conductors, like EEG electrodes, leading to noise.

A Faraday cage is a conductive enclosure that shields its contents from external electromagnetic fields. It is a common method for reducing electrical noise in EEG recordings.

The Tuskegee Syphilis Study was a unethical medical experiment conducted in the United States from 1932 to 1972. African American men with syphilis were misled and denied treatment to observe the natural progression of the disease, causing significant suffering and deaths.

The Tuskegee Syphilis Study highlighted the importance of informed consent, patient rights, and equitable treatment in medical research. It revealed a major flaw in ethical conduct and underscored the need for ethical guidelines in all research involving human subjects.

The Tuskegee Syphilis Study was a significant ethical violation in medical research. It violated the rights of the study participants, causing significant harm and leading to long-term consequences. It exemplifies the importance of ethical conduct and patient autonomy in medical research.

DNA data is unique to each individual and permanent, making it a major concern for data breaches.

The possibility of unauthorized access to sensitive information stored in a large database, such as those held by companies like 23andMe.

A situation where confidential data about individuals is exposed to unauthorized parties, potentially leading to harm.

The process of making research results and data publicly available to enhance scientific progress and credibility.

The potential for misuse of shared data, including re-identification of individuals from anonymized datasets, posing a privacy risk.

Balancing the need for transparency in research with the protection of individual privacy and security.

Government initiatives requiring data sharing for funded research projects, promoting scientific advancement and collaboration.

The challenge of making data accessible to researchers while ensuring the safety and privacy of the individuals involved.

A method used in research to reduce the number of animals needed for experiments by refining the experimental procedures to be more efficient and less invasive.

Replacing animals with non-animal techniques in research, such as using cell cultures or computer simulations.

A statistical method used to determine the smallest sample size needed for a study to detect a statistically significant result.

Using the same control group across multiple experiments to reduce the number of animals needed.

This involves using improved technology and methods that reduce the number of animals needed to obtain meaningful data.

Sharing tissues or samples from research animals with other researchers to maximize data collected from each animal.

Techniques used to mitigate pain, distress, and discomfort in research animals throughout all stages of the experiment.

Techniques focused on replacing the use of animals with alternative methods in research.

A 7T MRI scanner used to collect image data from 36 female Wistar rats.

The type of MRI sequence used to generate images for the deep learning model.

A 3D software used for manually segmenting the rodent brain into different regions.

The brain region segmentation followed a detailed map based on known brain structures in rats.

A total of 26 brain images were used to train the model, 5 for testing its accuracy, and 5 for adjusting its parameters.

Image resizing to 256x256 pixels and normalization to improve consistency across the dataset.

A type of convolutional neural network commonly used for image segmentation.

A process that applies filters to extract features by sliding across the image and learning local patterns.

Open data refers to data that is freely available for anyone to access, use, modify, and share. It promotes transparency, collaboration, and innovation by making information readily accessible.

The ADNI (Alzheimer's Disease Neuroimaging Initiative) database is a large collection of clinical, neuroimaging, genetic, and biomarker data from individuals with and without Alzheimer's disease. It aims to understand the progression of the disease and identify potential treatments.

Structural MRI (Magnetic Resonance Imaging) provides detailed images of the brain's anatomy and structure. It measures the density and distribution of different tissues, allowing researchers to identify and measure brain volume, degeneration, and lesions.

Functional MRI (fMRI) measures brain activity by detecting changes in blood flow. It shows which areas of the brain are active during specific tasks or mental processes, providing insights into brain function.

PET (Positron Emission Tomography) scans use radioactive tracers to visualize metabolic processes in the brain. They measure the distribution and uptake of these tracers, providing information about brain function and metabolism.

EEG (Electroencephalography) measures electrical activity in the brain through electrodes placed on the scalp. It records brain waves, revealing brain states like sleep, wakefulness, and cognitive processes.

Longitudinal data refers to data collected over time from the same individuals or subjects. It allows researchers to track changes, progression, and trends in a population over time, providing insights into disease progression and treatment effectiveness.

De-identifying sensitive data, such as brain MRI images, involves removing any identifiable features that could link the data back to a specific individual. This preserves privacy and allows for safe and ethical data sharing.

The fundamental problem in protein structure prediction is trying to figure out the unique 3D shape a protein folds into based on its linear sequence of amino acids.

A protein's amino acid sequence holds the blueprint for its final 3D structure. This means that the order of amino acids dictates how the protein will fold, but the process involves complex interactions between the amino acids.

AlphaFold2 uses evolutionary data to predict protein structures by analyzing related protein sequences. This helps identify conserved structural features and understand how evolution has shaped protein structures.

AlphaFold2 takes into account geometric constraints, such as bond angles and distances between atoms, to build a realistic 3D structure. These constraints help limit the number of possible configurations.

AlphaFold2 uses a powerful neural network called Evoformer to process multiple sequence alignments (MSAs). This network identifies patterns of evolutionary conservation and co-evolution, revealing important relationships between amino acids.

The Structure Module in AlphaFold2 is responsible for building the 3D coordinates of the protein. It iteratively refines the structure based on geometric reasoning, predicting angles, distances, and amino acid positions.

AlphaFold2 can leverage existing protein structures that are similar to the target protein. These structures can guide the prediction process and help refine the final model.

AlphaFold2 uses deep learning, a type of artificial intelligence, to learn complex relationships between amino acid sequences and protein structures. This allows it to make highly accurate predictions even for proteins with no known structure.