Overview of current techniques for reading from the brain and writing into it. ### Reading from brain #### fMRI Functional Magnetic Resonance Imaging (fMRI) essentially works by studying differential magnetic behavior of oxygenated and deoxygenated hemoglobin in the brain. Any part of the brain that's active needs oxygen for metabolism, which is transported by hemoglobin in the blood. Hence, the extra presence of oxygenated hemoglobin over deoxygenated is a correlate of neural activity. ![[Screenshot 2021-11-23 at 11.34.39 AM.png]] (via Wikipedia) It turns out that deoxygenated hemoglobin has a magnetic field while oxygenated doesn't. An fMRI machine exploits this differential magnetic behavior by generating a strong magnetic field and studying how it gets disrupted at different locations in the brain of the patient whose head is inside the magnetic field. So, by studying how magnetic field is disrupted at different locations, researcher can deduce which parts of the brain have been more metabolically active than the others. **Advantages:** - High spatial resolution (millimeters) compared to other methods - But still very low. Each voxel of fMRI (3 mm cubed) output contains **[aggregated activity of 27 million neurons](https://skope.swiss/monitor/fmri-one-in-a-million-neurons/)** - Non-invasive - Detects activity even in deep brain **Disadvantages:** - Low temporal resolution of a few seconds (~2 seconds) - Head cannot move inside the fMRI machine so studying mobile brains is difficult #### NIRS Near Infrared Spectroscopy (NIRS, also called Optical Topography) emits light at (two) different infrared wavelengths and measures their scattering properties. Like fMRI, NIRS detects metabolic activity of brain as scattering properties of oxygenated and deoxygenated blood is different. So by measuring how emitted infrared light scatters, an estimate of metabolic activity can be made. ![[Screenshot 2021-11-23 at 11.38.26 AM.png]] *Absorption spectra for oxy-Hb and deoxy-Hb for near-infrared wavelengths* (via [Wikipedia](https://en.wikipedia.org/wiki/Functional_near-infrared_spectroscopy)) >"Two or more wavelengths are selected, with one wavelength above and one below the [isosbestic](https://en.wikipedia.org/wiki/Isosbestic_point "Isosbestic point") point of 810 nm—at which deoxy-Hb and oxy-Hb have identical absorption [coefficients](https://en.wikipedia.org/wiki/Coefficient "Coefficient")." The reason infrared wavelengths is used is because optical wavelengths don't penetrate much into the skull while infrared are able to penetrate up to 1 cm or so inside the cortex. This depth is called [optical window](https://en.wikipedia.org/wiki/Near-infrared_window_in_biological_tissue) and represents the fact that near infrared light is able to penetrate biological tissue (but is absorbed by hemoglobin) ![[Screenshot 2021-11-23 at 12.02.19 PM.png]] (via [here](https://www.researchgate.net/figure/A-Near-infrared-NIR-light-has-the-best-penetration-depth-through-soft-tissue-B_fig1_320440081)) **Advantages** - Light-weight, mobile machine. Can be made into a helmet - Head can be mobile, so people can be studied in natural environments - Since emitter and detector is placed side by side, calibration within the head is automatic once initially done - "NIRS is more robust when confronted with electrical noise and motion-based muscle activity artifacts than EEG" (via [here](https://www.nature.com/articles/sdata20183)) - Non-invasive - High temporal resolution of milliseconds **Disadvantages** - Low spatial resolution (~1 cm) - better than EEG but not as good as fMRI - Still one has to remember that each cubic miliimeter of brain contains approx 1 million neurons - Only detects activity near cortex tissue (as light intensity falls off after that) - Like fMRI, lag time exists between brain activity and metabolic activity #### EEG Electroencephalography works on an extremely simple principle. Neural activity generates an electric field that induces slight current in an electrode, which when compared to a neutral reference electrode (say attached away from brain) can indicate brain activity. **Advantages** - Extremely cheap. Basic devices available for <$100 - Non-invasive, portable and lightweight - But if electrodes change position on the head, signal meaning changes - High temporal resolution **Disadvantages** - Extremely poor spatial resolution (only aggregated and synchronous activity of centimetres of neurons is able to induce detectable currents in the electrode) - Picks only scalp electric activity which is just cortex activity - Prone to picking electrical artifacts (activity that's not generated by the brain) Because of poor spatial resolution, EEG is often used for detecting only very simple types of brain states (like calm, anxious, etc.) by classifying brain activity waves in alpha, beta or gamma bands. #### MEG Magnetoencephalography works similarly to EEG but detects magnetic fields generated from the neural activity instead of electric fields. **Advantages** - High temporal resolution - Non-invasive - Magnetic fields less distorted by tissue (so can penetrate further into brain v/s eeg that's limited to the scalp) **Disadvantages** - Low spatial resolution but better than EEG (although debatable) - Expensive. Magnetic field detectors are quite expensive. Such machines would easily cost a million dollars. - Bulky. Such detectors haven't yet been miniaturized - Requires isolation of non-brain generated magnetic fields (such as from Earth) #### Functional transcranial Doppler (fTCD) From Wikipedia: >Functional transcranial Doppler sonography (fTCD) is a neuroimaging tool for measuring cerebral blood flow velocity changes due to neural activation during cognitive tasks.[[8]](https://en.wikipedia.org/wiki/Transcranial_Doppler#cite_note-8) Functional TCD uses pulse-wave Doppler technology to record blood flow velocities in the anterior, middle, and posterior cerebral arteries. Similar to other neuroimaging techniques such as [functional magnetic resonance imaging](https://en.wikipedia.org/wiki/Functional_magnetic_resonance_imaging "Functional magnetic resonance imaging") (fMRI) and [positron emission tomography](https://en.wikipedia.org/wiki/Positron_emission_tomography "Positron emission tomography") (PET), fTCD is based on a close coupling between regional cerebral blood flow changes and neural activation. Due to a continuous monitoring of blood flow velocity, TCD offers better temporal resolution than fMRI and PET. The technique is noninvasive and easy to apply. Blood flow velocity measurements are robust against movement artifacts #### Calcium imaging Via gene therapy, florescent proteins can be introduced into neurons that emit light depending on calcium concentration. This allows for estimation of electrical activity by measuring florescent activity via a detector. [See this](https://en.wikipedia.org/wiki/Calcium_imaging). #### Direct electrodes The most direct way of reading brain activity is by inserting electrodes in the brain and reading electric activity. **Advantages** - Single cell resolution possible - High temporal resolution **Disadvantages** - Invasive. Drill holes - Potential immune system reaction to foreign object in body - Potential for infection - Small part of brain activity can be read as electrodes are localized in the brain #### Electrocorticography ECoG uses electrodes placed directly over the brain by drilling out a piece of skull and replacing it with an array of electrodes. ![[Screenshot 2021-11-24 at 2.18.12 PM.png]] **Advantages** - Much better spatial resolution than EEG as it directly touches brain - High signal-to-noise ratio **Disadvantages** - Invasive. Skull needs to be removed - Potential immune system reaction to foreign object in body - Potential for infection ### Writing into the brain #### tES Transcranial electrical stimulation is a way for zapping the brain with a small electric current at different locations above the skull. **Advantages** - Non-invasive - Cheap - No reports of adverse effects **Disadvantages** - Very broad based. Stimulates a large area of brain at once - Only works in cortex near the skull #### Two-photon optogenetic stimulation This is a complicated technique that first requires gene therapy to introduce a gene that produces proteins (opsins, specifically channel rhodopsins) that get activated by light of a certain wavelength. These proteins are channel gates that open and let in ions that depolarizes / stimulate the neuron in which they're present. So, by focusing light at particular neurons we can get them excited. <iframe width="560" height="315" src="https://www.youtube.com/embed/CZifB2aQDDM" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> Two photons from two different lasers can help focus the energy at a specific neuron in the brain. Only where the focus of two different lasers combine does the energy rise to sufficient level to activate the neuron. This enables extremely precise localization of brain activity stimulation. ![[Screenshot 2021-11-23 at 3.39.02 PM.png]] **Advantages** - Non-invasive. Infrared light is located outside the skull as it's able to penetrate the skull - Extremely precise localization **Disadvantages** - Expensive and bulky. Such machines cost upwards of $100k - Requires gene therapy (which is not yet approved for humans) - Very nascent #### transcranial Focused Ultrasound (fUS) Neurons (and other cells) have mechanosensistive channels that can get activated with mechanical shear. This mechanical energy can be delivered non-invasively via focused ultrasound. **Advantages** - Non-invasive - Higher spatial resolution than TMS/TeS - Similar spatial resolution to deep brain electrodes **Disadvantages** - Mechanisms not well understood - Not as precisely targetable as optogenetics - Increases temperature of the skull #### Sono-optogenetics Delivering energy into brain via focused ultra sound that can activate mechanolumisceint nanoparticles that emit light when get mechanically agitated via ultrasound. These nanoparticles can be injected via blood and since blood vessels span entire brain, whenever ultrasound is focused in those regions, light can get emitted and neurons can get activated. **Advantages:** - Minimally invasive. Nano particles can be injected intravenously, no intracranial surgery on scalp or skull is needed - Precise targeting. Ultrasound can be foucsed at millimeter region in brain - Fast temporal response. Ultrasound can be switched on and off at millisecond level (faster than channel opening of rodhopsin channels) - Minimal harm. Temperature is raised only <0.2 degrees where ultrasound is focused **Disadvantages** - Requires gene therapy to introduce rodhopsin gene into brain - Very recent technique Read - [Sono-optogenetics facilitated by a circulation delivered rechargeable light source for minimally invasive optogenetics](https://www.pnas.org/content/pnas/116/52/26332.full.pdf) - [Focused Ultrasound for Noninvasive, Focal Pharmacologic Neurointervention](https://www.frontiersin.org/articles/10.3389/fnins.2020.00675/full) #### Direct electrodes As the name suggests, if you have electrodes planted in the brain, you can directly stimulate neurons. **Advantages** - Extremely precise localization **Disadvantages** - Small part of brain activity can be written as electrodes are localized in the brain - Invasive, potential for infection, immune system response <iframe class="signup-iframe" src="https://invertedpassion.com/signup-collector" title="Signup collector"></iframe>