Lectures 8-11 (Rasmus Petersen)

Question 1

What are LMNs? Where are they found?

Answer 1

Lower Motor Neurons (alpha motor neurons) , they directly innervate the muscle

They are found in the ventral horn of the spinal cord and in certain nuclei of the brainstem

Question 2

What are the inputs to LMNs?

Answer 2

Sensory Neurons

Local circuit neurons in spinal cord

Premotor neurons (Upper motor neurons - UMM - in the brain)

Question 3

Briefly outline the study into decerebrated cats study into locomotion.

Answer 3

Decerebrated cat (cut at level of brainstem) can still walk - although less agile.

Therefore, walking comes from spinal cord circuitry.

Question 4

What are the components of the serial model of motor control? Describe them in order.

What is key to remember about this model?

Answer 4

COMPETITION:
- Internal needs vs External stimuli
>
GOAL:
- what we want to achieve + sub goals of movement to perform required action.
>
MOTOR CORTEX:
- Plan how to achieve goal
>
BRAINSTEM/SPINAL CORD:
- Turns plan into muscle commands (activity of motor neurons)

KEY:
- It is not correct, it is a old model of how the system works

Question 5

What was the Bjursten et al (1976) study into lesions in cats? What does it suggest about the serial model of motor control?

Answer 5

METHODS:
- Took young cats and removed as much of the cortex as possible (until only brainstem + cerebellum was left)
- Cats were then left to recover.
- Serial model suggest cats would be absolutely crippled.

RESULTS:
- The cats were clumsy but could: walk, run, jump, drink, avoid obstacles.
- Was remarkable how much of the cats behaviour was preserved.

Serial model cannot explain how the brain works - clearly it is wrong.

Question 6

What did Dobzhansky (1973) famously say?

(Doubt it’s testable but good to know)

Answer 6

Nothing in biology makes sense except in the light of evolution.

Question 7

What is the significance of the Lamprey?

(A lamprey is an ugly ass fish)

Answer 7

It is called the ‘basal vertebrate’ - it’s believed to have the brains of our vertebrate ancestors.

This is because it is comprised of mainly brainstem and yet it can still move.

Therefore, cortex is not needed to survive and function.

Question 8

What is the relationship between cortex and brainstem evolutionarily speaking?

Answer 8

The brainstem is enough to allow a vertebrate to survive, so the cortex grew as an extra parallel component that allowed us to add to our basal functions.

Question 9

What is the parallel model of motor control? (simplified model of what is actually occurring)

Answer 9

GOAL
→
Cerebral cortex ←→ brain stem
→
Spinal cord

This model outlines the fact that there is a pathway from both the cerebral cortex AND brainstem.

And that there are TWO sets of premotor neurons: from the cerebral cortex and brainstem.

(Look at slide 11 of Rasmataz first lecture for picture).

Question 10

What are the two main types of descending pathways in the spinal cord?

Answer 10

Lateral pathway

Ventromedial pathway.

Both coming from higher parts to the lower parts of the spinal cord.

Question 11

Briefly describe the lateral pathway.

Answer 11

Descending pathway (Lateral part of spinal cord)

Controls distal muscles

Main origin is cortex (e.g., fine motor movements of fingers)

Question 12

Briefly describe the ventromedial pathway.

Answer 12

Descending pathway (Front-middle part of spinal cord)

Controls proximal muscles of the limbs and axial muscles of the trunk.

Main origin is in brainstem.

Question 13

What are the major cortical pathways? How are tracts named?

Answer 13

Corticospinal tract (cortex to spinal cord)

Corticobulbar tract (cortex to brainstem)

Named by: origin-destination

Question 14

Highlight the key information about the corticospinal tract.

Answer 14

AKA pyramidal tract

Most important lateral tract in humans.

Decussates in medulla

Premotor neurons in cortex: posterior frontal lobe and parietal lobe.

Question 15

What are the 2 major brainstem pathways, their subsections and the location of their premotor neurons?

Answer 15

Ventromedial tracts:

• Vestibulospinal tract: Vestibular nuclei
• Reticulospinal tract: Reticular formation
• Tectospinal tract: Superior colliculus (AKA tectum)

Lateral tract:
- Rubrospinal tract: Red nucleus

Question 16

Describe the Lawrence and Kuypers (1968) study about tracts. (Part 1)

(hint: monkeys)

Answer 16

CONTROL: intact brain
- Tested the monkeys motor control when it comes to grasping and reaching.
- Monkey able to this very skilfully with high dexterity.
- Sophisticated use of fingers

METHODS:
- Performed a bilateral pyramidotomy (cut the corticospinal tract on both sides).
- Therefore, connection of the cerebral cortex and spinal cord had been cut.
- Given 5 months to recover

RESULTS:
- Despite cutting descending pathways, shows that motor skills are intact.
- Can still grab and reach for apple but loss of dexterity in fingers (only whole hand grasping).
- Summary: major function intact but impairment in fingers dexterity.

Question 17

How do Corticospinal tract (CST) neurons innervate the spinal cord?

Answer 17

Indirect connections to LMN’s :
- Synapse interneurons.

Direct connections to LMN’s
- Synapse the LMN’s

Question 18

What is the importance/relevance of direct connections of spinal neurons to MNs? What animal(s) have this?

Answer 18

In primates/humans - due to evolution.

Allows direct cortical control of our hands due to being able to connect with the motor neurons.

This has allowed us to fabricate things with our hands/use tools - key for our dominant position on the world.

Question 19

Describe the Lawrence and Kuypers (1968) study about tracts. (Part 2)

Answer 19

Following on from the bilateral section of corticospinal tract…

METHODS:
- Sectioned either the lateral or ventromedial brainstem tract in the monkeys.

RESULTS:
Lateral Cut:
- Intact posture and locomotion
- Impaired whole-hand grasping (struggled to grab the apple at all)
Ventromedial Cut:
- Intact whole-hand grasping
- Impaired posture and locomotion.

Results are in line with what was expected.

Sections effect the parts that they control (obviously).

Question 20

What are the pathways of the parallel control systems, as proved by Lawrence and Kuypers? (3)

Answer 20

Ventromedial brainstem pathways:
- Basic system for movement control, posture and locomotion.

Lateral brainstem pathways:
- Control extremities, especially the hand

Corticospinal pathways:
- Mediate control similar to brainstem pathways.
- Add capacity for independent finger movement.

Question 21

Briefly describe the experiment into unilateral pyramidal lesions in monkeys and what its clinical relevance is.

Answer 21

METHODS:
- Incomplete lesion of the right pyramid in monkeys.
- Allowed to recover for 2 months.

RESULTS:
- Impacts the left hand ONLY
- Still had skilled food retrieval but not as dexterous
- Relatively intact independent finger movements.
- Remarkable recover from extensive lesions.

CLINICAL RELEVANCE:
- Shows how much we can recover from LoF in motor systems.
- Helped aid research and methods for patients who have had extensive damage.

Question 22

What are the major functions controlled by the brainstem in the trunk and limbs?

Answer 22

• Posture
• Locomotion
• Orienting to salient objects
• Reaching and grasping

Question 23

What are orienting behaviours?

Answer 23

Coordinated movements (walking, flying, swimming, etc.) that occur in response to an external stimulus.

Question 24

Why do we orient?

Answer 24

This occurs as our brains are selective, so if we sense a stimulus we need to direct our attention towards it.

Question 25

What are the four major behaviours animals use to orient?

Answer 25

foveation moving our head and neck moving the pinnae of the ear to locate the sound whisking - orienting whiskers to touch and sense surrounding objects.

Question 26

Briefly summarise the basis of how orienting eyes occurs.

Answer 26

The cone receptors are dense on the fovea, therefore, the orienting of eyes requires saccades which is controlled by the superior colliculus.

Question 27

Outline the circuit basis for the orienting of your eyes, and briefly describe the study that proves this. | Not including the actual process of motor control

Answer 27

The Superior Colliculus (SC) is a key structure in this process. It receives input from the retina to the visual layer (most superficial layers) in retinotopic organisation (each neuron has a spatial preference). Then the SC outputs to the brainstem from the motor layer. STUDY: - If you record from an area of the SC (e.g., area 7) and then flash a light in the region of the retina it is assigned to, a receptor potential will fire. - When an animal is looking straight ahead, microstimulation of that area will make it saccade to the area that the neuron is assinged to.

Question 28

Outline the circuit basis for the orienting of your eyes in reference to bringing about the movement.

Answer 28

The Brainstem contains premotor neurons called "Gaze centres" which control eye movement which are in the Paramedian Pontine Reticular Formation (PPRF). The PPRF is innervated by the superior colliculus and the frontal eye fields. It then projects to the LMN's that control eye movement - cranial nerves III, IV, VI for eye muscles.

Question 29

Outline the study by Hoy et al (2019) into prey capture. Why did they do it?

Answer 29

Lot's of experiments into SC's role in orieniting behaviour are done in artifical settings, so they wanted to do it in a more realistic setting. METHODS: - Recorded a control of a mouse orienting towards a cricket in a cage with them - did it almost immediately with whole body orienting. - Then, expressed iDREADDS in the NF of the SC. - Then they activated them by injecting an agonist CNO to stimulate the artificial inhibitory receptors. - This should decrease firing rate of NF if working correctly - which they did. RESULTS: - Shows intact running speed - suggests SC is not general motor controller. - WHEN given cricket, the orienting behaviour has gone, the mouse is much slower to get to the cricket. - Striking difference compared to the control group. - SHOWS significant role in orienting behaviour in hunting behaviour. iDREADS: molecular biology technique which allows you to express artificial receptors in different neuron types.

Question 30

What does iDREADDS stand for and what are they?

Answer 30

Inhibitory Designer Receptors Exclusively Activated by Designer Drugs. A molecular biology technique which allows you to express artificial inhibitory receptors in different neuron types.

Question 31

Give examples of eye movements (2) and orofacial movements (6) that the brainstem controls. | (in terms of eye movements and orofacial behaviours) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 31

* *Eye movements* * Vestibular reflexes * Orienting * *Orofacial behaviours:* * Breathing * Swallowing * Chewing * Sucking * Licking * Whisking

Question 32

What are two possibilities in how rhythms are generated ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 32

Sensory feedback Central Pattern Generator (CPG)

Question 33

What is Central Pattern Generation (CPG) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 33

A small network of neurons whose activity can generate specific movements with correct timing and sequences in the absence of sensory feedback

Question 34

What is whisking controlled by ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 34

A CPG (central pattern generator)

Question 35

Which 2 methods are used to measure whisking? | (2 ways) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 35

via EMG (electromyography) or via camera

Question 36

What effect does sensory feedback have on Whisking behaviours and how do we know this ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 36

* Whisking exists even if you cut the sensory nerve showing that the whisking rhythmn isnt generated by sensory feedback patterns. * sensory patterns are involved but their abolition isnt enough to stop whisking in mice

Question 37

What effect does removing /lesioning the frontal cortex have on whisking and what does this suggest? ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 37

* little effect on whisking - provides evidence whisking isn't dependent of cerebral cortex * suggests central pattern generator in the brainstem is controlling whisking behaviour (this is backed by analogous results for other orofacial rhythms)

Question 38

What is an orofacial rhythm ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 38

rhythmic movements involving the mouth and face.

Question 39

How would one go about finding the orofacial CPG (central pattern generator) circuits ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 39

* by finding the relevant motor neurons for orofacial rhythms * the expectation is that premotor neurons are the circuit behind rhthym generation - we find these via retrograde tracing

Question 40

# In reference to finding orofacial CPG circuits What is the process of retrograde tracing ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 40

* Inject retrograde tracer into target structure (it'll be taken up by axon terminals and transported till it reaches cell body) * If brain is remived and processed appropriately, should see cell bodies of the target structure (in this case premotor neurons)

Question 41

What is a candidate CPG for whisking behaviour ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 41

Vibrissa Intermediate reticular formation (vIRt)

Question 42

Moore et al (2013) conducted an experiment in order to investigate the candidacy of vIRt as the Central pattern generator for whisking behaviour in mice. What three things did Moore address in order to succeed in his research and how did he address them ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 42

(1) Do neurons in vIRt fire during whisking? Put microelectrodes into vIRt and conducted electrophysiological recording → Can clearly see a rhythm (2) Is activity in vIRt sufficient to generate whisking? Injected glutamate agonist (kainate) into vIRt - this should in theory induce whisking rhythm → It does - if you activate the vIRt you can then get the whisking rhythm, thus vIRt is the central pattern generator (3) Does lesioning vIRt abolish whisking behaviour? Conducted unilateral lesion of vIRt via electrolytic → Intact whisking on unaffected side, no whisking on lesioned side Collectively provide Strong evidence that whisking central pattern generator is in the vIRt brainstem nucleus

Question 43

What does the Central sulcus do structurally? ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 43

Seperate the frontal lobe from the parietal lobe

Question 44

What is another work for the primary motor cortex ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 44

Precentral gyrus

Question 45

Where are the primary and premotor cortex ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 45

The frontal lobe

Question 46

What did Fritsch and Hitzig (1870) discover in their research involving a dog ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 46

Used electrical stimulation of frontal lobe in dog Discovery of ‘motor cortex’ - easiest to elicit muscle contraction

Question 47

# As a follow on from Fritsch and Hitzigs electrical stimulation study What did Leyton & Sherrington (1917) discover | (hint:map) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 47

* Used much more precise electrical stimulation (in comparison to Fritsch and Hitzig) * Electrical stimulation of cortical surface of area 4 of an anaesthetised monkey * Saw specific body movements when electrically stimulating diff parts of area 4 * Discovered a topographic map of the body on the surface of cortical area

Question 48

What is the Classical definition of the primary motor cortex ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 48

frontal area where electrical stimulation elicits somatotopically organised movements at low stimulation intensities

Question 49

Outline the experiment By Penfield & Boldrey involving cortical stimulation in epileptic patients (M1) | (general outline) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 49

**Method** * Cortical stimulation in awake humans: * Tumour removal * Identification of epileptic focus - was trying to identify diseased tissue, if you electrical stimulate the affected region, it results in an epileptic fit, also wanted to avoid damaging critical regions * No pain receptors (nociceptors) on cerebral cortex, applied local anaesthetic to margin of wound , it's possible to electrically stimulate brain and them feeling no pain * Stimulation of precentral gyrus **Results** * *Electrical stimulus resulted in movement rather than sensation - confirmed that human precentral gyrus is the motor cortex* * Sensory & motor homunculus maps developed - consistent with Leyton & Sherrington * But it's a rather coarse map - not that precise

Question 50

What two predicitions about a Motor muscle map can be made using the Penfield Homunculus ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 50

Each muscle should be represented in one location of M1 Each M1 neuron should activate one muscle

Question 51

# For the Penfield Homunculus' hypothesis of a precise muscle map How was the prediction of 'Each muscle should be represented in one location of M1' tested and what was the result ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 51

Microstimulation of M1 can test the first prediction One muscle can be stimulated over a large region of M1 - goes against first prediction

Question 52

# For the Penfield Homunculus' hypothesis of a precise muscle map How was the prediction of 'Each M1 neuron should activate one muscle' tested and what was the result | (experiment, Aim, method & result) ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 52

**Spike triggered averaging (STA) - test of second prediction** * **Aim** - identify which muscles are activated by a given neuron in M1 * Each time neuron fires AP, extract corresponding EMG trace * Do this lot of times as EMG is noisy so the average will be much more clear **Method** * Monkey performs task with hand * Measure activity in M1 * Measure muscle activation (EMG) in response to each action potential * Average these responses * ‘bump’ at short latency indicates monosynaptic connection from neuron to muscle **Results** * Each M1 neuron should activate one muscle to be consistent with prediction * Each neuron activates several muscles – M1 is NOT a precise muscle map

Question 53

# For the Penfield Homunculus' hypothesis of a precise muscle map Why did Graziano et al (2002) conduct an experiment with long udration microstimulation ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 53

Reasoned natural behaviours are longer than 10ms so conducted long duration microstimulation in M1 (500ms)

Question 54

What did Graziano et al (2002) find when they conducted long duration microstimulation on Monkeys and what hypothesis did this lead to ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 54

They would elicit coordinated Behaviours such as hand to mouth action or defensive actions **Hypothesis**: M1 contains an ethological action map

Question 55

Can one M1 neuron alone cause movement? If so/not why? ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 55

M1-LMN connections are 100-200 uV (in terms of strength) Depolarisation needed to evoke an AP is roughly 15mV Conclusion - one M1 neuron cannot cause movement alone

Question 56

Geogopoulos et al (1986) conducted a reaching task experiment involve monkeys. Outline this experiment and its conclusions ## Footnote Week 5 - Voluntary motor control and descending pathways

Answer 56

* Reaching task, monkey reaches to green light * Recorded activity of neurons from M1 * Neurons ‘tuned' to reach direction * Tuning curves board * Bump has maximum at 140 degrees - neurons preferred direction of reach (differs per neuron) * Each M1 neuron has a preferred direction (PD) * Each neuron ‘votes’ for its preferred direction with a strength (‘weight’) given by its firing rate * **Population vector** * Average PD weighted by firing rate * Direction encoded by the population of neurons * This is population coding

Question 57

What is a Brain-Machine Interface (BMI)?

Answer 57

A medical device that measures or alters electrophysiological activity at the level of populations of neurons.

Question 58

What do the types of BMI do? (4)

Answer 58

Restore lost sensory abilities. Restore lost motor abilities. Regulate pathological neural activity. Restore lost brain processing capabilities.

Question 59

Why don't hearing aids work for everyone?

Answer 59

They work as an amplifier and are effective for certain types of deafness. However, there are certain forms of deafness caused by the destruction of hair cells. Therefore, there is an issue relaying the sound to the auditory afferents and higher. A different solution is needed...

Question 60

How do cochlear implants work and what damage in the ear do they bypass?

Answer 60

The cochlear implant bypasses damaged cochlear hair cells by directly stimulating surviving neurons of the auditory nerve. It does this by receiving input from the external microphone, which is transduced to tonotopic organisation of pure tones, which then stimulates the associated electrode that is implanted in the cochlear.

Question 61

What is frequency decomposition? What is its significance in BMI?

Answer 61

The dividing of a sound into it's pure tones. This is what the cochlear does, and therefore any BMI needs to be able to do this if it's going to help circumnavigate the damage to cochlear hair cells.

Question 62

How does frequency decompisition in the ear occur?

Answer 62

Sound makes the basilar membrane vibrate Location of maximum vibration depends on the tone frequency. Cochlea performs a frequency decomposition of sound: - The lower the frequencie, the further along the basilar membrane it is coded.

Question 63

What are the three main components of cochlear implants? What do they do?

Answer 63

Microphone: receive sound information and input it into the sound analyser. Sound analyser: splits sounds into its pure tones. Cochlear stimulator electrode array: stimulates each location in the cochlea with the matching frequency component (in same tonotopy as cochlea ofc).

Question 64

Outline the study into how cochlear implants perform under different conditions.

Answer 64

METHODS: - Tested under quiet condtions the difference between word and sentence recognition. - Tested over 1-24 months to see change over time. RESULTS: - Gradual increase in performance over time in both conditions. - Single word recognition reached a 50-60% accuracy after 24 months. - Sentence recognition reached ~90% accuracy after 24 months. Shows people are way better at recognising sentences - the contextual information allows for better clues and predicition of what is coming next.

Question 65

What is the most widespread and successful BMI?

Answer 65

Cochlear implants, as of 2008 there were 120,000 of them worldwide.

Question 66

What is the result of a spinal injury in the region C1-C4?

Answer 66

Lose function from the neck down. Requires respirator.

Question 67

What is the result of spinal injury in the region C5-C8?

Answer 67

Lose function in the waist down, and varying loss in upper body depending on severity. E.g., C7 = loss of function from chest down but retain use of arms and hands

Question 68

What is tetraplegia?

Answer 68

Paralysis of the arms and legs.

Question 69

What is dysarthria?

Answer 69

Difficulty speaking

Question 70

What is Dysphagia?

Answer 70

Difficulty eating.

Question 71

What is Locked in State?

Answer 71

Near-complete paralysis of all the muscles in your body (can usually move eyes at this stage).

Question 72

What are the systoms and then the late symptoms of ALS?

Answer 72

Symptoms: - Paralysis - Dysarthria - Dysphagia Late Symptoms: - Respiratory failure - Locked in State

Question 73

What is the basic BMI schematic? What do they each do?

Answer 73

Electrode array (recording from the brain). Decoding algorithm (computing neuronal signals into a usable signal). Actuator takes information from decoding algorithm and transduces it to complete the task. - E.g., moving a robot arm, moving a cursor on a screen.

Question 74

Outline what is meant by a population vector in relation to M1.

Answer 74

Each M1 neuron has a preferred direction (PD) in which each neuron 'votes' its PD with a 'weight' given by its firing rate. Therefore, the Population Vector is the average of PD's of all neurons in relation to their firing rate. Direction is encoded by the summation of neural signals from the population of neurons.

Question 75

Outline briefly the process and significance of population vector decoding in relation to reaching task.

Answer 75

Decoding: - Record many M1 neurons during reaching task. - Measure the PD of each M1 neuron. - On each trial, use neuron responses to construct population vector. Result: - Population vector closely matches actual reach direction. Significance: - Gives insight into how networks of MN's in motor cortex work to bring about motion. - In BMI's this is critical because we can decode the intention of the monkey without observing it's movement. - Insane potential clincial applications.

Question 76

Outline the study by Moran & Schwartz (1999) into complex movements. | (hint:monkeys make spiral movements)

Answer 76

METHODS: - Monkeys trained to make spiral movements with their arms. - Recorded the neurons in M1. - Used the population vector decoding to predict the monkeys movement 100ms ahead of movement. RESULTS: - Predicted ENTIRE trajectory. - Complex movements can be decoded from motor cortex activity

Question 77

Outline the Serruya et al (2002) study into Monkeys using Utah arrays.

Answer 77

METHODS: - Recorded from monkey M1 with Utah array to track hand movement. - Monkeys trained to use joystick to move cursor on the screen and track target. - Then they disconnected the joystick from the computer and continued to record the PD from M1. - Used a 'linear filter' algorithm to decode hand movement from M1 activity. RESULTS: - Monkey used neural control to track target successfully (fairly good accuracy). - Proof-of-principle of neural control.

Question 78

Outline the Hochberg et al (2006) study 'Braingate'.

Answer 78

METHODS: - Participant MN had a spinal cord transection C3-C4 = complete tetraplegia. - Implanted Utah array in M1 arm area. - Then asked him to make imagined movements with his arms. - Trained algorithm to decode imagined action from M1 activity. RESULTS: - Was able to complete a centre-out task in 4 directions with 80% accuracy. - Shows neural control.

Question 79

What, in brief, did Willett et al (2021) discover when they implanted 2 utah arrays into a participant with a C4 spinal cord injury. ## Footnote Week 6 - Brain Machine interfaces

Answer 79

* The participant, T5, was able to produce 90 characters a minute with only a 5% error rate * this is comparable to the smart-phone typing speed of T5's age group

Question 80

Describe the Willet et al (2021) experiment in full | (Hint, Participant T5, imagined handwriting) ## Footnote Week 6 - Brain Machine interfaces

Answer 80

**Participant** * Had a C4 spinal cord injury - paralysed from the neck down * 2 Utah arrays implanted in M1 (primary motor cortex) , particularly hand area **Procedure** * T5 asked to imagine handwriting complete sentences * Neural activity recorded from motor cortex * this produces data to train an algorithim * Algoeithim ('recurrent neural network') trained to decode characters from neural activty **Results** * 90 characters produced per mijn * 5% error rate (no backspace allowed either) * Comparable to smart-phone typing speed of T5’s age group * Performance fluctuated a bit day by day but performance was overall high as trial progressed * Reason why performance so high- algorithm was fed through a language model ( a model that knew of the grammar and spelling of the english language)

Question 81

What is a critical evaluation of the Willett et al (2021) study | (hint- P T5, imagined handwriting - 2 strengths, 2 limitations) ## Footnote Week 6 - Brain Machine interfaces

Answer 81

**Strengths** * Addresses degrees of freedom limitation * Achieves peer typing speed **Limitations** * One participant * Needs a dedicated lab

Question 82

What is locked in syndrome in ALS ## Footnote Week 6 - Brain Machine interfaces

Answer 82

only limited eye movement remains, otherwise complete paralysis

Question 83

What is completely locked in syndrome in ALS ## Footnote Week 6 - Brain Machine interfaces

Answer 83

No movement at all (including eye movement)

Question 84

Moses et al (2021) & Chaudhary et al (2022) both conducted studies looking at the potential application of BMIs for communication. What did they do in order to investigate this ## Footnote Week 6 - Brain Machine interfaces

Answer 84

the interface to the microelectrode is implanted in the brain Microelectrode is a sheet of electrodes that sits on the surface of the brain

Question 85

What are the pros and cons of using a micro electrode sheet to record brain activity ## Footnote Week 6 - Brain Machine interfaces

Answer 85

**pros** * less invasive - less likely to evoke an immune response **cons** * signals that electrodes get are weaker and nosier (surface electrodes)

Question 86

What are the four key components of physically producing speech? ## Footnote Week 6 - Brain Machine interfaces

Answer 86

Lips, tongue, jaw , larynx

Question 87

What are the major peripheral nerves involved in speech control ## Footnote Week 6 - Brain Machine interfaces

Answer 87

* Facial nerve (VII) * hypoglossal (XII) * controls tongue * trigeminal motor (V) * controls jaw * laryngeal * controls larynx

Question 88

What does the brainstem control that is involved in speech ## Footnote Week 6 - Brain Machine interfaces

Answer 88

Muscles (lips, tongue, jaw , larynx) Major peripheral nerves Lower motor neurons in the brainstem Promotor nuclei in the brainstem

Question 89

Which areas are involved in the cortical control of speech ## Footnote Week 6 - Brain Machine interfaces

Answer 89

M1 face area Broca’s area

Question 90

What can a lesion to the M1 face area result in? ## Footnote Week 6 - Brain Machine interfaces

Answer 90

The complete loss of voluntary control over speech - might scream if spooked but won’t be able to give a coherent lecture

Question 91

What did Penfield find regarding cortical areas and speech? ## Footnote Week 6 - Brain Machine interfaces

Answer 91

A widespread network of areas that interfere with speech

Question 92

What is Broca’s area? ## Footnote Week 6 - Brain Machine interfaces

Answer 92

A region of the brain in the left frontal lobe (areas 44-45) responsible for the production of speech and language processing

Question 93

What does lesioning to Broca’s area cause ## Footnote Week 6 - Brain Machine interfaces

Answer 93

**Broca's Aphasia** * Impaired ability to produce language even though basic ability to control speech related muscles is unimpaired * Fine with comprehension but cannot produce

Question 94

What do lesions to Broca’s area show about the cortical processing of speech ## Footnote Week 6 - Brain Machine interfaces

Answer 94

* Broca’s is not a unique site of speech control * Lesion restricted to Broca’s alone causes only transient Broca’s aphasia - a network of areas are involved

Question 95

Outline the Moses et al (2021) experiment regarding the use of BMI's in communication | (participant details, implant, tasks, decoding & performance) ## Footnote Week 6 - Brain Machine interfaces

Answer 95

**Participants**: * One participant - 'Bravo-1', 36 y/o. * History of stroke in the pons region at age 20, * anarthria (inability to speak) & quadriparesis (paralysis of all four limbs). * Prior communication abilities limited to computer-assisted typing, averaging 5 correct words per minute. **Implant**: * Implanted with electrocorticography (ECoG) electrodes over the left sensorimotor cortex, positioned above pia and below dura. * implant was placed to cover the primary sensory and motor cortices. **Tasks**: * Isolated word task: Bravo-1 imagined speaking 50 isolated words. * Sentence task: Presented with sentences to imagine speaking, aimed at training an algorithm. **Decoding:** * Neural data recorded during tasks. * A deep learning algorithm trained to classify words from neural signals. * Classifier output filtered using a language model. **Performance:** * Isolated words: Achieved 47% accuracy. * Sentence decoding: * Initial word error rate: 61%. * Improved to 26% with language modeling. * Language model integration led to a decrease in error rate. * Production rate increased to 15 words per minute, three times faster than the previous eye tracker system | (big card, so best to rememeber in chunks)

Question 96

Outline the findings from the Chaudhary et al (2022) experiment regarding the use of BMI's for communication ## Footnote Week 6 - Brain Machine interfaces

Answer 96

Though BMIs can be used to increase communication capacity for CLIS (completely locked in state) ALS patients, this communication is not always quick and may not be entirely reliable | (check notes for further details on the study )

Question 97

What conclusions can be reached about BMI's uses in communication from both the Moses et al (2021) study and the Chaudhary et al (2022) study? (4 +, 3 -) ## Footnote Week 6 - Brain Machine interfaces

Answer 97

**Successes** * Application of basic science insights into neural coding * Novel treatments for severe paralysis * proof -of principle established * Progress in refining the methods **Limitations:** * Studies are of single participants * Systems cannot be used independently by caregivers * Commercialisation required