Korean, Edit

Chapter 5. Brain engineering

Higher category : 【Biology】 Brain Science Index 

1. The definition of brain engineering and the application

⑴ Definition : The technology which connects the principle of the brain function and the application to the artificial system.

⑵ Subject : Artificial intelligence, BCI, brain imaging technology, computer imaging, pattern recognition, neuron-on-a-chip, biomimetics, virtual reality, nano-bioengineering, neuroinformatics, neural modeling, neural microsystem, optogenetics, etc.

2. The basics of the brain engineering

⑴ Neural encoding

① Tuning curve : The graph of the strength of the nerve stimulus relative to the direction of the muscle.

⑵ Aliasing and Nyquist sampling theory : The sampling frequency should be 2, 3, ··· times of the original frequency.

⑶ Neuromorphic system

① Definition : The electronic system functioning perception, recognition, and control, imitating the biological nerve mechanism of the brain

② Aim : To output high efficiency with lower cost imitating the function of the brain

⑷ STDP(spike time dependent plasticity)

① The definition of memory : The change of the neural synaptic strength ωij which connects the j th pre-synaptic neuron and the i th post-synaptic neuron

② The change of the synaptic strength is the function of the difference of the transmission time tjpre of the pre-synaptic neuron and the reception time tipost of the post-synaptic neuron.

⑸ Memristor(memory resistor)

① It is one of the 4 main electronic elements.

cf. Resistor, inductor, capacitor, etc.

② Mathematization

③ It memories the value of the current.

⑹ The artificial neural network : The deep learning algorithm(see here)

① The hypothesis of the McCulloch-Pitts model

○ Neurons follow the law of “all-or-none”.

○ To activate any neuron, the fixed number (>2) of synapses should be activated within the determined time.

○ All the delay comes from the delay within synapses.

○ Inhibitory synapses absolutely stop the activation of the post-synaptic neuron.

○ Unlike human, the structure of the neural network doesn’t change by the time.

② Machine learning

○ Supervised learning : To find out the rule of a data set with labeled samples

○ Unsupervised learning : To find out the rule of a data set with unlabeled samples

○ Reinforcement learning : To use the reward-behavior system

3. Processing the brain signal

⑴ EEG(electroencephalogram)

① Definition : To measure and record the electrical signal produced by neurons externally

② History

○ 1875 : Coton recorded the electrical behavior of brains of animals.

○ 1925 : Nemunski researched EEGs of dogs.

○ 1929 : Hans Berger discovered α wave and β wave in EEG.

○ 1933 : Adrin showed the clinical application.

③ The basics of EEG

○ Time domain : About the lapse of EEG

○ Frequency domain : About the spectrum of EEG

④ The decomposition of EEG

○ Filtering : It eliminates undesired signals(e.g., noise).

○ Fourier analysis : It separate EEG into various sine waves.

○ Wavelet analysis : It treats information from both the time domain and the frequency domain. By the uncertainty principle, there is a trade-off between time and frequency.

⑤ The activity of each class of EEG

○ EEG δ wave : Deep sleep(Steriade, 1993), cortical integration in a large scale(Bruns & Eckhorn, 2004), the network of neocortex and thalamus(Steriade, 1999)

○ EEG θ wave : Hippocampus(Kahana et al., 1999), thalamic nuclear(Hughes et al., 2004), thalamic cortex loop(Talk et al., 1999), working memory function

○ EEG α wave : Berger(1924), PLI(phase-locking index), phase consistency

○ EEG β wave

○ EEG γ wave

⑵ X-ray analysis

① Röntgen discovered X-ray by chance in 1895.

② The transmittance of X-ray depends on the molecular weight and the density of a sample.

○ X-ray analysis studies the amount of the penetrated wave from a sample or a patient.

○ X-ray rarely penetrates materials of high molecular weight and high density such as metals and bones.

○ Transmittance : Lung(air) > fats > water > white matter > gray matter > bones

○ The higher transmittance the tissue has, the darker the image is.

③ If injecting contrast media, the morphology of blood tube can be observed.

④ By exposure of radiation, X-ray analysis is harmful to patients if used frequently.

⑶ MEG(magnetoencephalography)

⑷ MRI(magnetic resonance imaging)(see here)

① 1^st. A MRI equipment looks like a solenoid, encompassed by a shim coil and a gradient coil.

② 2^nd. If strong magnetic field is applied in MRI, the direction of atomic magnetic moment of 1H becomes parallel to the magnetic field.

③ 3^rd. RF power amplifier transmits RF(radio frequency) wave of high amplitude.

④ 4^th. NMR(nuclear magnetic resonance) : 1H resonates some of RF wave so that it absorbs the energy and becomes excited state from bottom state.

⑤ 5^th. The direction of the magnetic moment of the excited 1H is changed.

⑥ 6^th. The excited 1H becomes the bottom state emitting the energy it absorbed(④), and the direction of the magnetic moment returns.

○ The returning time varies in tissues considerably.

⑦ 7^th. The change of the magnetic moment generates induced currents and voltages in the RF receiver.

⑧ 8^th. ADC converts RF signals into digital signals by Fourier transformation.

⑨ 9th. By reiterating processes of 3^rd-8^th, the variable returning time of 1H can be used in imaging and MRS(magnetic resonance spectroscopy). 

⑩ The lateral view of the brain

⑪ The medial view of the brain

⑫ The superior view of the brain

⑬ The coronal view of the brain

⑸ fMRI(functional MRI)

① It studies BOLD(blood oxygenation level dependent) signals.

② OxyHb : There is no signal loss by OxyHb because it is diamagnetic and doesn’t distort the ambient magnetic field.

③ DeoxyHb : There is signal loss by DeoxyHb because it is paramagnetic and distort the ambient magnetic field.

④ The fMRI image of the primary motor cortex

⑤ The fMRI image of the motor speech area

⑥ The fMRI image of the sensory speech area

⑦ The fMRI image of the visual cortex

⑹ NIRS(near infrared spectroscopy) : It uses the Beer-Lambert law

⑺ PET(positron emission tomography)

① It studies the radiation of the isotope of the glucose.

⑻ CT(computed tomography) : 3D computer analysis

① Allan Cormack, Godfrey Hounsfield won a Novel prize in 1979 by inventing CT technology.

② As a source of radiation, CT uses X-ray or PET, and X-ray is preferred. 

③ X-rays are located three-dimensionally and the transmittance are quantized digitally so that it can create the image of the cross section of body.

④ Rather than simple X-ray, it clarify the various tissues.

4. The control of the brain function

⑴ The invasive implement : It gives brain considerable burden.

① EpCS

② DBS(deep brain stimulation)

③ Ultrasonic wave

④ Penumoencephalogram : It blows air into the brain ventricle so that the brain image is created.

⑵ Noninvasive implement : Restricted information

① TMS(transcranial magnetic stimulation)

② tDCS

⑶ Optogenetics

⑷ BBI(brain-to-brain interface)

5. The instances of the brain engineering

⑴ Lie detector

⑵ Dream decoder

Input: 2018.09.19 00:01

1. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
2. Source: http://serc.carleton.edu/research_education/geochemsheets/techniques/CT.html [본문으로]
3. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
4. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
5. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
6. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
7. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
8. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
9. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
10. Source : Seoul Nat’l University Medical College Prof. Hwang Yeong Il, from Curriculum Handout [본문으로]
11. See https://jb243.github.io/pages/447 [본문으로]
12. Source : Seoul Nat’l University Medical College Prof. Park Seon Won, from Curriculum Handout [본문으로]
13. Source: http://www.noliemri.com/ [본문으로]
14. Source: http://www.atelierth.net/bbs_view.php?term_id=6677&id=21933 [본문으로]