Brain as CPU, Eye as Camera PDF

Functions and Classification of Nervous system Dr.D.Sivaraman Department of Biotechnology SoET, PDEU  Functions of the nervous system Basic Function Higher-order...

Functions and Classification of Nervous system Dr.D.Sivaraman Department of Biotechnology SoET, PDEU  Functions of the nervous system Basic Function Higher-order functions Receive Inputs Memory Storage and Retrieval Analyze and interpretation Regulation and Maintenance Response- Output- Homeostasis Action  Functions of the nervous system Processing & Input Interpretation Output Sensory function Detection & Action and Response interpretation Motor Function Regulation Higher-order functions Control and Storage and Retrieval Maintenance  Functions of the nervous system Sensory function Detection & Action and Response interpretation Motor Function Information from the environment Integrates sensory information from The brain sends signals to through our senses (sight, hearing, different parts of the body. It the muscles to produce touch, taste, and smell). movement. This allows us interprets the information and determines the action to walk, talk, eat, and perform other actions. Regulation Higher-order functions Maintain homeostasis, which is a constant Responsible for higher-order functions internal state in the body. This includes such as thinking, learning, memory, and regulating body temperature, heart rate, blood emotion. pressure, and breathing. Fundamental Classification Location and Axis Movement of Signals Actions in response of the Stimuli External and Internal Stimuli Voluntary / Involuntary actions Fundamental Classification  Afferent and Efferent Sensory nerve tracts Motor nerve tracts Afferent Sensory Division Efferent Motor Division Receives and transmit the signals Receives and transmit the signals from effector organ towards CNS from CNS to the effector organs External Environment : Hearing/ Smell/ Somatic NS Autonomic NS Touch Voluntary Involuntary Sympathetic Parasympathetic Brain Vs CPU  BRAIN CPU The brain is composed of neurons, which are CPU has cores that act as the brain's neurons. the basic working units. Neurons A CPU core is a processing unit that executes communicate with each other using electrical instructions. and chemical signals, Cores Neurons It's always busy, fetching instructions, decoding them, executing them, and storing the results. The more cores a CPU has, the more workers it has, and the faster it can get things done.  BRAIN CPU Control Unit (CU): Acts like the brain’s Different regions of the brain (e.g., cerebrum, executive function (frontal lobe). It directs the cerebellum, brainstem) specialize in various operation of the processor functions—such as decision-making, motor control, and sensory processing. Arithmetic Logic Unit (ALU): This is like the cerebellum, performing mathematical and logical operations. Regions of Brain CU /ALU  BRAIN CPU The brain receives information through CPU: Input devices like keyboards, mouse, and sensory inputs (eyes, ears, skin, etc.), which sensors are analogous to the brain's sensory send signals to the brain for processing. For organs. example, visual information is processed in the occipital lobe. Input Devices Sensory inputs  BRAIN CPU Short-term memory: Temporary and limited (like Random Access Memory (RAM): This is the remembering a phone number for a few seconds). short-term memory of a computer. Long-term memory: Stores information permanently or Hard Drive or Solid State Drive (HDD/SSD): for long periods (like learned skills or knowledge). This is where long-term data is stored, similar to how the brain stores long-term memory. STM /LTM RAM/ SSD  BRAIN CPU The brain requires a constant supply of oxygen A CPU requires a power supply to function. and nutrients from blood vessels to function. Without electrical power, it cannot perform computations. Blood Supply Power Supply Artificial Neural Network (ANN)  Artificial Neural Network (ANN) Complex problems ANN- Mimicking Recognition, decision-making, human brain function and problem-solving Solution with accurate prediction An Artificial Neural Network (ANN) is a computational model inspired by the way biological neural networks in the human brain function. ANNs are used in machine learning and artificial intelligence to solve complex problems by mimicking human brain activities such as pattern recognition, decision-making, and problem-solving.  Neurons Nodes The basic units of an Brain are neurons Nodes are typically arranged in layers: input layer, hidden layers, and output layer. Interface between the neural network and the external world.  Process neurons Input neurons Output neurons Input neurons: Receive information from the outside world. Hidden neurons: Process the information received from input neurons and transmit it to output neurons. Output neurons: Produce the final output of the network. 1 2 Cancer Detection Process the Image Images Detect the presence or Absence of Cancer  Weights Determines the Importance of Each Inputs when there are multiple inputs Connections between neurons are associated with weights Weight determine the strength of the influence of one neuron on another. Weights are adjustable parameters learned during the training process. These weights control how much impact an input will have on the output.  Assign Weights to Input Inputs Cholesterol level: Let’s say the weight for 220 mg/dL cholesterol is 0.7. Blood pressure: Systolic 140 mmHg The weight for blood pressure is 0.3  Cardiovascular Risk Assign Weights to Multiply Inputs by Input Inputs Weights Multiply cholesterol (220 mg/dL) by its weight (0.7): Cholesterol level: Let’s say the weight for 220 * 0.7 = 154 220 mg/dL cholesterol is 0.7. Multiply blood pressure (let’s use 140 mmHg for simplicity) by its Blood pressure: weight (0.3): Systolic 140 mmHg 140 * 0.3 = 42 The weight for blood pressure is 0.3 Add the weighted inputs together: 154 + 42 = 196 Neural network has Activation Function learned that if the summed value is Since 196 is greater than 150, the above 150, there is model predicts that this person is a high risk of heart at risk for heart disease disease.  Activation Function Activation function allows the neural network to learn complex patterns  After inputs are multiplied by weights and summed, the result is passed through an activation function.  The activation function introduces non-linearity, which allows the neural network to learn complex patterns. Forward Propagation Initial Prediction Data Input Layer Hidden Layer Output layer Assign Weights to Data Processing/ Input Inputs Initial Prediction 28×0.5=14 BMI: 28 Weight for BMI: 0.5 Prediction: 70% Blood glucose level: 150 chance of Weight for Blood glucose 150×0.6=90 diabetes. mg/dL level: 0.6 Total 14+90=104 In forward propagation, data moves from the input layer, through the hidden layers, to the output layer. The neuron’s output is calculated using its inputs, weights, and the activation function.  Comparing Prediction to Actual Training Phase & Result Backpropagation Training Backward Propagation Repeating the Process Adjusting weights after each Prediction: 70% chance of Weight for BMI might prediction. diabetes. change from 0.5 to 0.4. Over time, the network’s Actual Result: 0% (no Weight for blood glucose predictions become more diabetes). might change from 0.6 to accurate. 0.7  ANNs learn by adjusting the weights using a method called backpropagation.  Initially, the network makes predictions using random weights. The difference between predicted and actual outputs (the error) is calculated.  In backpropagation, the network adjusts its weights in reverse (from output to input) to reduce the error  Learning Process  Supervised Learning: The network is trained on labeled data, where the correct output is known.  Unsupervised Learning: The network learns patterns from data without labeled outcomes.  Reinforcement Learning: The network learns based on feedback and rewards from its actions.  Supervised Learning Input Learning Prediction Output To learn which patterns Once trained, the Gene expression levels of gene expression are network can predict the Correct cancer type for each cancer patient associated with different type of cancer for new (label) for each patient cancer types. patients Increased Gene Expression Type of Cancer AR (Androgen Receptor) gene Prostate Cancer VEGF (Vascular Endothelial Growth Ovarian Cancer Factor) EGFR (Epidermal Growth Factor Lung Cancer Receptor) Human Epidermal Growth Factor Breast Cancer Receptor- HER2 gene  Unsupervised Learning Input Learning Prediction Output The network uses The network identifies Predict the organism DNA sequences from techniques like clusters of species with based on the DNA various organisms clustering to group similar genetic sequence similar DNA sequences sequences DNA sequences from various Species  Reinforcement Learning Network interacts with the patient's data Rewards & Input Feedback Penalties Output If the blood sugar After administering the levels improve, the Different dosage of dosage, the network gets network gets a reward. Optimize drug dosage Insulin to diabetic feedback in the form of If they worsen, the for different patients patients patient response network receives a penalty. Insulin Dose Patient ID Control on Blood Sugar Dose 1 1121 Good Dose 2 1176 Very Bad Dose 3 2086 Very Good Dose 4 2141 Bad Dose 3 is optimised for Patient ID 2086  Machine Learning Trial Data Set Training models Model familiarise with the Data pattern Predictions Model Can able to With the Memory of New Data set Predict Training Machine learning is a branch of artificial intelligence (AI) that focuses on developing algorithms and statistical models that enable computers to perform tasks based on past experience.  Steps Involved in Machine Learning Gathering Data Data Preparation Data Wrangling Data Analysis Creating the patterns Patient Data Cleaning the Data Priorities the most X and Y axis charts Medicine Stock Handling the missing important data set Identifying the patterns Salary of the Employers data chosen for the task Deployment Test the Model Train the Model Deploy it into the real- Evaluate how well the Learn the relationships world application. In model performs on between the input biology, unseen data features and the output  Cyclin D1 Gene MYC gene Angiogenesis – New Blood vessel formation DNA replication Abnormal - Cell Enhance the Tumor Multiplication Growth by supplying Nutrients Breast Cancer Lung Cancer Head and Neck Cancer  Machine Learning Techniques Doctor is Analysing the Gathering Data Data Preparation Data's from the Cancer Patient Data's from the 10 patients Cleaning the Data Cyclin D1 Handling the missing Generating the MRN- MYC Gene data Medical Record Number Datas on Gene expression report Analysis the MRN for the two Genes Tumour tissue biopsies If the data is missing in any of the MRN then quantitative PCR error should be rectified RNA Sequencing  Data Wrangling Data Analysis Creating the patterns Priorities the most X and Y axis charts important data set Identifying the patterns chosen for the task Gene expression report having expression of many genes Analyzing the Gene expression Report Giving High focus to the Cyclin D1 MYC Gene  Train the Model Test the Model Learn the relationships between the input Evaluate how well the features and the output model performs on unseen data Higher the Expression Model is been tested of the Cyclin D1 with unknown set of MYC Gene data's which is not used during the trial session Greater the Chances of Cancer Analysis the predictions with the test Choose the Drug that datas used to manage the cancer in these patients  Deployment Deploy it into the real- world application. In biology, Model is now ready for application in cancer hospitals, research centers etc The eye functions as a camera Anatomy and Physiology of Eye Sclera White outer coat of the eye, called the sclera. It protects the inner workings of the eye and provides a place for the muscles that move the eye to attach. Cornea Transparent, dome-shaped structure that covers the front of the eye. It's essential for vision as it plays a major role in focusing light rays entering the eye.  IRIS Pupil The iris is the colored part The pupil is the dark, circular opening of the eye that surrounds in the center of the iris. the pupil. It is a muscular The size of the pupil is controlled by the diaphragm that controls the muscles of the iris. amount of light entering the eye by adjusting the size of the pupil. Controls - Pupil In bright light, the iris constricts the pupil to reduce the amount of light entering the eye. In dim light, the iris dilates the pupil to allow more light to enter the eye.  LENS The lens of the eye is a transparent, biconvex structure that focuses light onto the retina. Ciliary Muscle Ciliary muscle, which is responsible for A thin lens has less curvature and typically a longer focal changing the shape of the lens. When the length. ciliary muscle contracts, allowing the lens to become thicker A thick lens has more curvature and a shorter focal length.  Vitreous humor The vitreous humor is a clear, jelly-like substance that fills the space between the lens and the retina in the back of the eye. Structure of Retina  Structure of Retina and Vision formation The retina is a light-sensitive layer of tissue that lines the back of the inner eye. It is responsible for converting light into electrical signals that are then transmitted Cornea Pupil Lens Vitreous Humor Retina Photoreceptor cells Rods (opsin +retinal) Cones (photopsins) Rods function in low light and are Cones function in bright light and are responsible responsible for night vision, for color vision and high-acuity vision. The opsin in red cones absorbs light in the long-wavelength range. green cones Rhodopsin, also sometimes referred to as visual purple, is a light- absorbs light in the medium-wavelength range,.blue cones absorbs light in the short- sensitive pigment found in the rod cells of the retina in the eye wavelength range, Electrical Signal Optic Nerve ->Brain The brain interprets the electrical signals from the retina as vision. The brain's visual center is located in the occipital lobe  Mechanism/ Physiology of Vision Light falls on the object and get reflected back – Rays enters the cornea were it transmit and refracts the light Then it enters the pupil (Based on the brightness Iris adjust the size of pupil using sphincter/ dilator muscle) From the pupil light now enters the Lens (Ciliary muscles controls the shape of the lens) From lens it passes across the Vitreous fluid (Gels like watery substance secreted by ciliary epithelium) Finally the light falls on the retina (Light sensitive layer) which comprises of photoreceptors cells In retina rod cells (dim vision) responsible for formation of shape which converts light rays in to nerve signal using rhodopsin and cone cells (bright vision) responsible for color vision which converts the light rays in to nerve signals. Rods and cones together send the signals to the optic nerve via bipolar neurons Optic nerve carried the information to the visual processing center of the brain (visual cortex). It is located in the occipital lobe  How a camera works Entry of Light Focuses the Light Shutter Shutter Opens and Closes: The Light Enters the Camera: When you The camera's lens focuses the light, camera's shutter opens for a brief take a photo, light from the scene making the image clear. By adjusting moment, allowing light to hit the enters the camera through the lens. the lens (zooming or focusing), you sensor or film. The length of time the The lens focuses this light onto a control how sharp or blurry the subject shutter stays open is called the sensor or film inside the camera appears. shutter speed, which affects how much light is captured. Image Process & Captures the Image Storage The sensor (in digital cameras) or film In digital cameras, the camera (in old cameras) captures the light. The processes the information from the sensor converts the light into electrical sensor to form the image. The signals, forming a digital image, while photo is then saved to a memory film chemically reacts to light to create card, ready for viewing, editing, or a picture. printing.  How a camera works Entry of Light Focuses the Light Shutter Shutter Opens and Closes: The Light Enters the Camera: When you The camera's lens focuses the light, camera's shutter opens for a brief take a photo, light from the scene making the image clear. By adjusting moment, allowing light to hit the enters the camera through the lens. the lens (zooming or focusing), you sensor or film. The length of time the The lens focuses this light onto a control how sharp or blurry the subject shutter stays open is called the sensor or film inside the camera appears. shutter speed, which affects how much light is captured.  How a camera works Image Process & Captures the Image Storage In digital cameras, the camera The sensor (in digital cameras) or film processes the information from the (in old cameras) captures the light. The sensor to form the image. The photo is sensor converts the light into electrical then saved to a memory card, ready for signals, forming a digital image, while viewing, editing, or printing. film chemically reacts to light to create a picture.