#computational neuroscience | Explore Tumblr Posts and Blogs

pishstudy · 2 years

Text

Neural Code and Neural Encoding

Week 2 - Computational Neuroscience Coursera

2.1 What is Neural Code?

To put it simply, neural code is how information about a stimulus is expressed via neurons by their response to these stimuli.

Neural Coding Problem:

Encoding Problem: How does a stimulus cause a pattern of responses

Decoding Problem: What does these responses tell us about the stimulus

Because neural systems are very noisy, the models we use are probabilistic...

So when we are looking at neural encoding, we are looking at the probability of a certain response given a certain stimulus . . . . p(response|stimulus)

And vice versa for neural decoding; the probability of a certain stimulus being present given a certain response. . . . p(stimulus|response)

How do we quantify the stimulus and the response?

In computational neuroscience it is important to quantify the stimulus and the response so that we can measure them and we can find their relationship.

We often hope that a given stimulus causes a given response in a neuron (though this is not always the case - sometimes a stimulus and a response have no relationship).

Therefore, the stimulus is the independent variable and the response is the dependent variable. So we often visualize it as such, where the stimulus is on the x-axis and the neural response is on the y-axis.

Often the neural response is quantified as the average firing rate of a neuron.

The stimulus parameter varies. Some examples include:

The average amplitude of an audio recording of someone speaking a certain syllable

The angle orientation of a bar of light

The luminance of an array of pixels

Visualizing Stimulus-Response Relationships:

Tuning Curves:

In the below tuning curve a V1 cortical neuron. These V1 cortical neurons often respond to orientated bars of light. In this case the neuron fires more frequently when the bar of light is almost vertical. When it is horizontal it almost doesn’t respond at all.

This tuning curve is more cosine because the motor neuron still fires despite the angle of the arm. It just fires more and more frequently as the arm approaches 150 degrees.

Stimulus Representations:

A stimulus does not have to be a specific sound or picture - it can also be a concept in general.

In an experiment published by Quiroga et al. in Nature(2005), 80 images were shown to epilepsy patients undergoing surgery. A neuron’s firing rate was recorded for each picture and the following tuning curve was made:

The images with the highest responses were actually pictures of Brad Pitt and Jennifer Aniston. This neuron would not fire at such a high frequency at images of Brad Pitt or Jennifer Aniston alone.

In that same experiment, the patients were shown images and another neuron was recorded. A tuning curve (not shown here) indicated that the stimuli which produced the highest firing rate was the concept of Pamela Anderson. The neuron would fire frequently if an image, drawing or voice recording of Pamela Anderson was shown. It even fired frequently if her name written out was shown!

This indicates that neurons can have varying complexity in stimulus representations. Some stimuli are more geometric (such as in the Lateral Geniculate Nucleus) and some are more semantic (such as in the cortex).

2.2 Simple Models

Thus far we have talked about the 2 variables we mostly deal with: the stimulus and its neural response. We understand how we could quantify these variables individually and we know how to visualize them.

But we do not understand their relationship. As mentioned earlier there is encoding and decoding. Let’s first look at neural encoding models - which focus on how a neuron responds to a stimulus: P(response|stimulus).

Basic Coding Models:

1. Linear Response:

This is the simplest response coding model (i.e. relationship between the response and the stimulus). It assumes that the response is linearly dependent on the stimulus at that time (or at some time recorded in the past).

This means that our response function, r(t), will look like our stimulus function, s(t), but scaled by some factor number. It may also be delayed a little bit because the response is not immediate after a stimulus is present.

Pro: It is very simple.

Con: It does not account that there may be more than one input other than the stimulus that the response could be dependent on.

2. Temporal Filtering:

It is also known as delayed amplification

Temporal filtering involves looking back over k time points on s(t). At each of these time points we weight the stimulus at that time by some factor function f(k).

The function f(k) is used to change the weight in noisy signals, such that stimuli signals which are not resulting in the response are “filtered out”. This makes the response function look much less noisy than the stimulus one.

Mathematically, this means that the response function r(t) is a convolution of the stimulus function s(t) and the filtering factor function f(t):

3. Running Average :

The running average is an example of temporal filtering, where the filtering factor function is equal to 1/N :

f(t) = 1/N, where N is the number of time steps back.

So basically we find the value of s(t) at each of the N time steps and then we weight it with 1/N.

The larger N becomes, the more the noise in the stimulus becomes cancelled out, therefore smoothing out s(t). If N=0, it will just be r(t) but slightly delayed.

4. Leaky Average:

The leaky average is another example of a temporal filtering system. This system assumes that r(t) still depends on s(t); however, the memory fades over time. This means that time points which are more recent will be weighted larger than time points which are older. In fact, as we go back in time t by N time steps, the weight for older time steps will exponentially decrease.

This filtering factor function is called a leaky integrator - a filtering function that’s strength exponentially decreases into the past.

This filtering system is important in computational neuroscience because the neuronal cell membrane also behaves like this - i.e. older signals decay.

5. Spatial Filtering:

In spatial filtering the goal is very similar to temporal filtering except now it is in 2 dimensions. Instead of taking time points and weighting them by some factor, now we are taking spatial points (x,y) and then weighting them by some factor.

Spatial filtering is important for understanding responses in neurons which respond to light patterns in space - i.e the receptive filed in the visual cortex (very important).

Let’s start by defining a visual space s(x,y) as shown below:

We have a point (x,y) which is varied by x’ and y’. The x’ and y’ are similar to the k variable in temporal filtering. Whereas, k defines the time interval we look at, x’ and y’ define the area in the space that we are going to look. And at each point defined between x’ and y’ we filter it by f(x’,y’) which is a filtering function of x’ and y’, so that we can get our response function r(x,y):

Just like in temporal filtering, the response at a specific point is dependent on how similar the original point is to the point on the filtering function. The filter will often be a shape - if this is ringing any bells it should because this is a receptive field. Therefore, this is the simplest way receptive fields are modelled!

Spatial Filtering and Retinal Receptive Fields:

Let’s say we have a neuron with the following on-centre and off-surround receptive field:

This is the cartoon representation of a receptive field that we have used thus far. f(x’,y’) is roughly represented as being either negative or positive depending on its shading. Realistically, it actually looks more like this:

As you can see f(x’,y’) is very positive at (0,0,0) and then it has a ring around where f(x’,y’) is negative. Now let’s see how we can apply such a receptive field to an image.

Let’s say we have the following image of the Taj Mahal:

We define our point (x,y) on the image and then create our coordinate frame with x’ and y’ as we did before. Then, we multiply them by f(x’,y’) at those points and sum them together.

If there is a bright point in the centre on the image, then there is going to be a positive response. However, if there is a bright point in the surround on the image, then there is going to be a decreased response (NOT NEGATIVE THOUGH - a neuron cannot fire negatively) because the positive image point will be multiplied by a negative filter point. However, a dark spot on the surround will produce an increase in r(x,y) because a negative brightness will be multiplied by a negative f(x’,y’).

Approximating the Receptive Field:

The receptive field is approximated with 2 gaussian distributions. One narrow positive gaussian with a smaller standard deviation and one broad negative gaussian with a larger standard deviation.

This difference of gaussians filter works very well at detecting differences in brightness - i.e. a dark patch next to a bright patch. It cancels out regions in an image where there is constant intensity, thus it focuses mostly on edges.

This is mathematically indicated below:

w(x,y) is the difference of gaussians filter and you can see the broad gaussian being subtracted from the narrow gaussian.

Just as a refresher here is the gaussian formula so it is more clear:

6. Spatiotemporal Filtering:

Now that we understand spatial filtering and temporal filtering - we can combing them into spatial temporal filtering. Visual sensory neuronal responses depend on both spatial and temporal inputs. Thus, we need a filter that incorporates space and time. Such a filter would look like this:

As you can see that the frames of the receptive field defined in space by x and y are implemented over different points in time. It is also important to note how the spatial receptive field at varying time points is different.

The response function would be a function obviously also be a function of space (defined by x’ and y’) and time (defined by tau).

Shortcoming of the Linear Filters Talked about thus Far:

Linear filtering can result in response functions with negative firing rates or have firing rates that increase indefinitely as the input increases. These 2 situations are obviously impossible as a neuron cannot fire negatively and neurons also have a refractory period.

To fix this by adding another step: a non linear function applied to the filtered stimulus.

This is represented mathematically as such:

Now that we have our stimulus, our filtering feature/function, our nonlinear function, we have all the components of a basic encoding model!!

#coursera #computational neuroscience #neuroscience #math #statistics #studyblr #study blog #science #stay productive #jennifer aniston #brad pitt

9 notes · View notes

pressnewsagencyllc · 16 days

Text

Early Reviews of Humane AI Pin Aren’t Impressed

With the long-hyped Humane AI Pin finally hitting the streets Thursday, those gripping their hands until their knuckles turn white in anticipation since its debut last November might want to hold off for a bit before dropping $699 (plus a $24-a-month subscription) on the small, wearable chatbot. Reviews have made their way out the door, and so far, none of those who got their hands on one have…

View On WordPress

0 notes

sandeep-health-care · 8 months

Text

Computational Technologies for Healthcare

In the modern-era technologies are evolving at a faster rate, resulting in a transformation of our lives, particularly in healthcare, with the objective of saving humanity while also providing advanced solutions for practitioners to enhance their choices. Artificial intelligence, Big Data Analytics, Internet of Things, and machine learning all have significant implications in accomplishing this. This article begins by outlining contemporary computational technologies in healthcare before discussing their significance and use. The technologies that are assisting healthcare to improve will indeed be discussed in this article, which will then explore the several research possibilities and obstacles associated with implementing computational technology in healthcare.

Introduction:

Health care has greatly improved aid to automated systems assisting clinicians in making precise diagnoses and a better organizing of information system. Machine learning, artificial intelligence (AI), the Internet of Things, and data analytics are all important in the healthcare business. Machine learning algorithms, for an instance, detect heart disease and cancer at their earliest stages in addition to assist practitioners in making appropriate decisions and diagnoses. With the advancement of AI algorithms, the medical sector is gaining assistance for more successful surgery. Data analytics and IoT plan to handle data effectively and remotely, allowing doctors to access patient reports from anywhere and at any time. More significantly, to forecast health outcomes and make informed treatment plan decisions. This had also improved as a result of counselling patients to visit physicians depending on their health status. Fig. 1 depicts the several aspects of emerging computing technologies.

Healthcare is an ever-changing and dynamically evolving industry. It is critical to stay up to date on the latest trends and advances, particularly for professionals or for the general public. There are numerous elements at work, including the obstacles, changing lifestyle, extended lifespan, evolving health care systems, and digitization. Everything has led to a rapidly changing healthcare scene. Thus, learning about upcoming trends can assist organizations in analyzing, identifying discrepancies and gaps, and resolving those difficulties. We can also estimate what the main players in the healthcare industry will do this year. Healthcare technology trends possess the potential to significantly transform the industry.

Related Work:

Various advancements in how we identify, prevent, and diagnose diseases have driven the healthcare business during the previous decade. Without the phenomenal growth of technology fueled by AI and the digital transformation of healthcare services in response to tougher global conditions and increasing needs for accessible and excellent medical care, this should not have occurred. From anaesthetics and antibiotics to MRI scanners and radioactive therapy, technical advancements in healthcare have resulted in profound transformations. While technologies, the latest pharmaceuticals and treatments, novel devices, updated social media helping with healthcare, etc., will fuel innovation, human aspects continue to be one of the consistent limitations of breakthroughs. No prediction will delight every person; rather, this article intends to investigate snippets of the future to see ways in which we can speculate a bit more clearly about the best way to go where we would like to go.

The intent of this article is to explain the tactical approach to transformation. The use of digital health technologies in healthcare organizations is more significant because improving healthcare, exchange of information, and the efficiency of systems, particularly in developing countries, is crucial to achieving the Millennium Development Goals. The purpose of the research is to highlight the variables of growing technology usage that influence the transformation of healthcare. This technological innovation effectively manages the records needed to record transactions and procedures. Healthcare providers have little choice but to adapt to the fast-moving technological change in the health industry if they want to stay competitive. Artificial Intelligence, Internet of Things, Big Data, Machine Learning and Cloud Computing. It is crucial that those who must use the technology accept it and gain a deeper grasp of the elements that contribute to the systems' positive effects in order to enhance the deployment process. Essential to society's health and the standard of care. Therefore, these justifications suggest that healthcare providers are currently adopting digital transformation in healthcare units. The advantages of digital transformation in healthcare organizations are very strong. They consist of decreased travel expenses for employees, less time spent away from patients, and simpler information system wait times in the healthcare industry.

In recent days, the entire nation has been pushed down by Corona virus. Techniques using both traditional and cutting-edge technologies are required to combat COVID-19 and bring the situation under control. The article focuses to methodically examine current developments in smart healthcare technologies, such as big data analytics and artificial intelligence, which will ultimately rescue the planet. By creating connected frameworks, these intelligence-based solutions support creative managements, versatility, productivity, and efficacy. The research being investigated specifically addresses Big Data and AI contributions that ought to be incorporated into intelligent healthcare systems. Additionally, it investigates how big data analytics and artificial intelligence are used to provide users with information and assist them in making plans. Finally, it proposes models for intelligent healthcare systems that utilize big data analytics and AI

According to the Heart Attack & Stroke Stats 2021 from the American Heart Association (AHA), coronary artery disease and stroke remain the leading cause of death globally and are spreading at an alarming rate. The current coronavirus (COVID-19) epidemic has made this increase much worse and put more strain on the already overburdened healthcare system. The common healthcare issues can be resolved using Smart and Connected Health. Figure 2 outlines the most significant technology. Through incorporating services that are worth more, medical care can become more proactive, preventative in nature, and individualized. Effective use of intelligent health care facilities was made at several pandemic response stages, encompassing identifying diseases, viral being detected, individual observing, tracking, and regulating, as well as the allocation of resources [2]. It's critical to monitor the developments affecting technology in health care as we advance. Sophisticated hospitals and healthcare facilities rely significantly on legacy hardware and software, but it's important to consider how this equipment might be merged with contemporary hardware or ultimately replaced by more reliable ones. Without sacrificing predictability or connectivity, enhancements in efficiency, earnings, reliability, and privacy should take precedence.

Computing Technologies:

The trending technologies in the healthcare domain are discussed below and the digital transformation is depicted.

Artificial Intelligence:

Healthcare uses artificial intelligence to assess and reduce several diseases' treatments. In a number of medical environments, including diagnosis processes, companies that manufacture medicine, portable medical facilities, etc., Intelligence is used. Intelligence assists in the gathering of historical information for illness identification and avoidance in the healthcare industry through digital health records. Besides pandemic therapy and consequence, artificial intelligence has a variety of other applications. The speed at which data is analyzed and judgements are made has been greatly accelerated by AI. Machine learning has a significant impact on the medical sector's capacity to develop new drugs and enhance the efficacy of testing methods. The examination of CT scans is currently aided by artificial intelligence in order to detect pneumonia in COVID-19 treatment patients. Microsoft developed Project Inner Eye, an artificial intelligence radiological technology. It now takes only a matter of seconds instead of hours to complete the patient's 3D sculpting. To gather biomedical dissertations from PubMed, Microsoft has developed another artificial intelligence dubbed Project Hanover. It also helps in selecting the best treatments for every individual, as well as speeding up the cancer diagnosis process.

Read More: https://www.europeanhhm.com/articles/computational-technologies-for-healthcare

#healthcare #health #medical care #health and wellness #technologies #artifical intelligence #iotsolutions #computational neuroscience #hospitals

1 note · View note

richdadpoor · 8 months

Text

Watch Rupert, an AI, Learn to Play Super Mario Live on TikTok

On TikTok, between the “get ready with me” videos, life hacks, and memes, a few robots are working on a challenge that many of us have faced at some point in our lives: beating Super Mario World. Over the past week, users have been live streaming an AI’s attempts to learn to play Mario, and for one robot in particular, it’s going great. Its name is Rupert, and it just beat level 2. Generating…

View On WordPress

#Artificial general intelligence #artificial intelligence #chatgpt #Computational neuroscience #George #Gizmodo #Mario #Nintendo #OPENAI #Rupert #Seth Hendrickson #SethBling

1 note · View note

philosophiesde · 2 years

Text

Neurozentrismus

Neurozentrismus Der hier veröffentlichte Artikel wurde von dem Autor als Wikipedia-Eintrag veröffentlicht unter: https://de.wikipedia.org/wiki/Neurozentrismus und bezieht sich auf den bereits veröffentlichten Essay “Das neurozentristische Weltbild – bitte wenden !“. Neurozentrismus ist eine Wortneuschöpfung in Anlehnung an den Begriff des Eurozentrismus und geht auf das Buch „Ich ist nicht…

View On WordPress

1 note · View note

newsupdatesbykiara · 2 years

Text

Only 12% companies are utilising AI to outpace their rivals: Report

Only 12 per cent of organisations that use artificial intelligence (AI) are doing it at a maturity level to achieve a strong competitive advantage while the rest are still experimenting with the technology, a new report said on Tuesday.

The research from Accenture puts the median AI maturity of organisations at a moderate score of 36, revealing most companies have significant opportunities to generate greater value with AI.

Tech firms globally already have a high AI maturity score of 54, which will rise moderately to 60 in 2024.

In contrast, carmakers and suppliers will leap from a moderate 39 today to 57 in two years -- betting on a significant surge in sales of AI-powered self-driving vehicles.

#Academic disciplines #Articles #Artificial intelligence #Computational neuroscience #Cybernetics #Ethics of artificial intelligence

0 notes

mindblowingscience · 4 months

Text

Our brains are remarkably energy efficient. Using just 20 watts of power, the human brain is capable of processing the equivalent of an exaflop — or a billion-billion mathematical operations per second. Now, researchers in Australia are building what will be the world's first supercomputer that can simulate networks at this scale. The supercomputer, known as DeepSouth, is being developed by Western Sydney University. When it goes online next year, it will be capable of 228 trillion synaptic operations per second, which rivals the estimated rate of operations in the human brain.

#Science #Neuroscience #Brains #Computers #Super Computer #Australia

79 notes · View notes

why-the-heck-not · 3 months

Text

26.01.24, friday

duolingo french streak day 21

4h of writing an essay abt my thesis topic aka I’m elbow deep in articles about brains even tho I study computer science (I love it tho: finally get read abt neuroscience and actually have it not be ”just for fun” (my 16-year-old self who was 100% set on becoming a neurosurgeon/scientist is thrilled currently))

grocery store

#i’m afraid I’m gonna lean too much into the neuroscience of things instead of the cybersecurity & computer aspect #tho I’m sure someone’s gonna go ’’hey u know u’re studying computer science right?’’ if I do so might be worrying for nothing #and like obv I’m gonna have to explain how BCI functions if I’m gonna talk abt its cybersecurity issues #and for that I’ll have to touch on neurons & brain waves & signals etc.#feeling very good abt this topic tho #like I’m having difficulty going to sleep now bc would want to keep reading abt it #(or bc I’ve been drinking coffee too late in the evening)#I just started doing things way too late today hence not much time spent on this today #but hoping tmrw’ll be more productive #bc today was supposed to be like hella productive but alas #today’s productive(ish) things #january 2024 #2024

35 notes · View notes

compneuropapers · 9 months

Text

Interesting Papers for Week 30, 2023

Adult-born neurons inhibit developmentally-born neurons during spatial learning. Ash, A. M., Regele-Blasco, E., Seib, D. R., Chahley, E., Skelton, P. D., Luikart, B. W., & Snyder, J. S. (2023). Neurobiology of Learning and Memory, 198, 107710.

Behavioral origin of sound-evoked activity in mouse visual cortex. Bimbard, C., Sit, T. P. H., Lebedeva, A., Reddy, C. B., Harris, K. D., & Carandini, M. (2023). Nature Neuroscience, 26(2), 251–258.

Exploration patterns shape cognitive map learning. Brunec, I. K., Nantais, M. M., Sutton, J. E., Epstein, R. A., & Newcombe, N. S. (2023). Cognition, 233, 105360.

Distinct contributions of ventral CA1/amygdala co-activation to the induction and maintenance of synaptic plasticity. Chong, Y. S., Wong, L.-W., Gaunt, J., Lee, Y. J., Goh, C. S., Morris, R. G. M., … Sajikumar, S. (2023). Cerebral Cortex, 33(3), 676–690.

An intrinsic oscillator underlies visual navigation in ants. Clement, L., Schwarz, S., & Wystrach, A. (2023). Current Biology, 33(3), 411-422.e5.

Not so optimal: The evolution of mutual information in potassium voltage-gated channels. Duran-Urriago, A., & Marzen, S. (2023). PLOS ONE, 18(2), e0264424.

Successor-like representation guides the prediction of future events in human visual cortex and hippocampus. Ekman, M., Kusch, S., & de Lange, F. P. (2023). eLife, 12, e78904.

Residual dynamics resolves recurrent contributions to neural computation. Galgali, A. R., Sahani, M., & Mante, V. (2023). Nature Neuroscience, 26(2), 326–338.

Dorsal attention network activity during perceptual organization is distinct in schizophrenia and predictive of cognitive disorganization. Keane, B. P., Krekelberg, B., Mill, R. D., Silverstein, S. M., Thompson, J. L., Serody, M. R., … Cole, M. W. (2023). European Journal of Neuroscience, 57(3), 458–478.

A striatal circuit balances learned fear in the presence and absence of sensory cues. Kintscher, M., Kochubey, O., & Schneggenburger, R. (2023). eLife, 12, e75703.

Hippocampal engram networks for fear memory recruit new synapses and modify pre-existing synapses in vivo. Lee, C., Lee, B. H., Jung, H., Lee, C., Sung, Y., Kim, H., … Kaang, B.-K. (2023). Current Biology, 33(3), 507-516.e3.

Neocortical synaptic engrams for remote contextual memories. Lee, J.-H., Kim, W. Bin, Park, E. H., & Cho, J.-H. (2023). Nature Neuroscience, 26(2), 259–273.

The effect of temporal expectation on the correlations of frontal neural activity with alpha oscillation and sensory-motor latency. Lee, J. (2023). Scientific Reports, 13, 2012.

Describing movement learning using metric learning. Loriette, A., Liu, W., Bevilacqua, F., & Caramiaux, B. (2023). PLOS ONE, 18(2), e0272509.

The geometry of cortical representations of touch in rodents. Nogueira, R., Rodgers, C. C., Bruno, R. M., & Fusi, S. (2023). Nature Neuroscience, 26(2), 239–250.

Contextual and pure time coding for self and other in the hippocampus. Omer, D. B., Las, L., & Ulanovsky, N. (2023). Nature Neuroscience, 26(2), 285–294.

Reshaping the full body illusion through visuo-electro-tactile sensations. Preatoni, G., Dell’Eva, F., Valle, G., Pedrocchi, A., & Raspopovic, S. (2023). PLOS ONE, 18(2), e0280628.

Experiencing sweet taste is associated with an increase in prosocial behavior. Schaefer, M., Kühnel, A., Schweitzer, F., Rumpel, F., & Gärtner, M. (2023). Scientific Reports, 13, 1954.

Cortical encoding of rhythmic kinematic structures in biological motion. Shen, L., Lu, X., Yuan, X., Hu, R., Wang, Y., & Jiang, Y. (2023). NeuroImage, 268, 119893.

Mindful self-focus–an interaction affecting Theory of Mind? Wundrack, R., & Specht, J. (2023). PLOS ONE, 18(2), e0279544.

#science #Neuroscience #computational neuroscience #Brain science #research #cognition #neurons #cognitive science #neural networks #neural computation #neurobiology #psychophysics #scientific publications

75 notes · View notes