INTRODUCTION TO

AI & ML
Overview of Artificial Intelligence
• AI is composed of two words, “artificial” and “intelligence,” where “artificial” stands for “human
created’ and “intelligence” stands for “thinking power.”
• AI is “a man-made object with thinking power’.
• Intelligence is the ability of a system to
• calculate,
• reason,
• perceive relationships and analogies,
• learn from experience,
• store and retrieve information from memory,
• solve problems,
• comprehend complex ideas,
• use natural language fluently,
• classify,
• generalize, and
• adapt to new situations.
• For AI to work, the availability of “data” is the main key
• Humans can learn faster and process certain things faster even with a limited amount of data, but
AI-based systems need massive amounts of data to generate any useful inferences.
• The answers are present in the data, which can be obtained by applying AI to get them out.
• The AI techniques speed up the implementation of complex programs.
• AI is currently being applied in a variety of fields, ranging from playing chess and music to making
complex decisions, creating models, predicting patterns, even self-driving cars and the Medical field.
The Importance of AI
• AI can automate repetitive learning through the datasets.
• AI adds intelligence to existing processes, it cannot be viewed as an independent application.
• AI uses progressive learning algorithms that allow the data to carry out the programming.
• It can find structure and irregularities in the data to be used in classification and/or prediction.
• The back-propagation technique allows the algorithm to refine itself, with the help of training data and
new data, if the predicted results are not accurate.
• AI can analyze large data with hidden layers of neural networks.
• It can obtain higher accuracy through deep neural networks
• In the medical field, AI-based DL, image classification, and object recognition techniques can be
employed to possibly detect the disease on MRIs with almost as much reliability as when it’s done by
trained radiologists.
• AI algorithms learn entirely differently from humans, they ought to perceive things differently, and can
easily visualize the relationships and patterns that cannot be seen by humans.
• Thus the human–AI partnership can offer many opportunities:
(i) It can provide further support to our existing abilities, and allow for better perception and
understanding.
(ii) It can introduce analytics to industries in which AI is currently being used.
(iii) It can be used to improve the analytic technologies such as computer vision, time-series analysis, etc.
(iv) It can bridge the economic, language, and translation barriers.
(v) It provides know- how of ML to be used to build predictive models for AI.
 (vi) It can learn how software is to be utilized to process, analyze, and derive meanings from natural
language.
(vii) It can process images and videos for several real-time applications.
(viii) It can build intelligent systems to provide interactive communications between humans and AI
systems.
• Process involved with AI
• Reasoning Process
• Learning process
• Problem solving process
• Prediction process
• Self-correction process
ML overview
• ML is a programming computer to optimize a performance criterion using example data or past
experience.
• We have a model defined up to some parameters and learning is the execution of a computer program to
optimize the parameter of the model using the training data or past experience.
• ML uses the theory of stastics in building mathematical models, because the core task is making
inferences from the same.
• In ML the role of learning is twofold: First in training we need efficient algorithms to solve the optimized
problem as well as to store and process the massive amount of data we generally have.
• Second, once model is learned its representation and algorithmic solution for inference needs to be
effective as well.
Importance of ML
• The importance of machine learning stems from its wide-ranging applications and its ability to transform
industries by automating processes, improving decision-making, and uncovering insights from large
datasets.
• Here are some key points highlighting the importance of ML:
• Automation and Efficiency
• Enhanced Decision-Making
• Personalization
• Financial Services
• Natural Language Processing (NLP)
• Research and Development
• Healthcare and Medicine
• Machine learning (ML) is playing an increasingly important role in medicine, transforming how
healthcare is delivered, improving patient outcomes, and enhancing the efficiency of medical processes.
Here are some key areas highlighting the importance of ML in medicine:

Medical Diagnostics
1.Early Detection of Diseases:
1. ML algorithms can analyze medical images (like X-rays, MRIs, and CT scans) to detect
diseases such as cancer, heart disease, and neurological disorders at an early stage, often
with high accuracy.
2. For instance, ML models can identify early signs of diabetic retinopathy in eye scans
before noticeable symptoms occur.
2.Pathology:
1. ML can assist pathologists by automating the analysis of tissue samples and identifying
abnormal cells, leading to faster and more accurate diagnoses of conditions such as cancer.
 Personalized Medicine
1.Tailored Treatment Plans:
1. ML can analyze patient data, including genetic information, medical history, and lifestyle factors, to
develop personalized treatment plans.
2. This approach ensures that patients receive treatments that are most effective for their specific condition
and genetic makeup.
2.Predictive Analytics:
1. ML models can predict patient outcomes based on historical data, helping doctors make informed decisions
about treatment options.
2. For example, ML can predict which patients are at higher risk of complications from surgeries or certain
treatments.

Drug Discovery and Development

1.Accelerating Drug Discovery:
2. ML algorithms can analyze vast amounts of biological and chemical data to identify potential drug
candidates more quickly and at a lower cost than traditional methods.
3. This speeds up the process of finding new medications and bringing them to market.
2.Predicting Drug Efficacy and Safety:
2. ML models can predict the efficacy and potential side effects of new drugs by analyzing preclinical and
clinical trial data.
3. This helps in optimizing drug development and reducing the likelihood of adverse effects.
 Medical Imaging
1.Image Analysis and Interpretation:
1. ML is widely used in radiology to enhance the analysis and interpretation of medical images.
2. It helps in identifying abnormalities such as tumors, fractures, and infections with high precision.

2.Automated Image Segmentation:

1. ML can automate the segmentation of medical images, dividing them into regions of interest for
further analysis, which is crucial for planning surgeries and treatments.

Predictive Maintenance of Medical Equipment

1.Equipment Monitoring:
2. ML models can predict when medical equipment is likely to fail or require maintenance by analyzing
usage patterns and performance data.
3. This ensures that equipment is always in optimal working condition, reducing downtime and
improving patient care.
 Patient Monitoring and Management
1.Remote Monitoring:
1. ML-powered devices and wearables can continuously monitor patients' vital signs and health metrics, alerting
healthcare providers to any concerning changes.
2. This is especially valuable for managing chronic conditions and post-surgery recovery.
2.Disease Management:
1. ML can help manage chronic diseases such as diabetes, hypertension, and asthma by analyzing patient data and
providing personalized recommendations for lifestyle changes and medication adjustments.

Clinical Decision Support

1. Decision-Making Assistance:
2. ML systems can assist healthcare providers in making clinical decisions by providing evidence-based
recommendations and identifying potential diagnosis options.
3. This support helps reduce diagnostic errors and improve patient outcomes.
2. Risk Assessment:
4. ML models can assess patients' risk for developing certain conditions or experiencing adverse events, enabling
proactive interventions.
5. For example, predicting the likelihood of hospital readmissions or the risk of sepsis in ICU patients.
 Health Records and Data Management
1.Efficient Data Analysis:
1. ML can analyze electronic health records (EHRs) to extract valuable insights, identify trends, and improve patient
care management.
2. This enables more efficient use of healthcare data for research and clinical practice.
2.Natural Language Processing (NLP):
1. NLP, a subset of ML, can be used to analyze and interpret unstructured data in medical records, such as doctors'
notes, to extract meaningful information and support clinical decision-making.

Research and Public Health

1.Epidemiology:
2. ML models can analyze data from various sources to track the spread of infectious diseases, predict outbreaks, and
support public health interventions.
3. This is particularly useful for managing pandemics and other public health crises.
2.Genomics and Precision Medicine:
2. ML helps in analyzing genomic data to understand the genetic basis of diseases and develop targeted therapies.
3. It plays a crucial role in advancing the field of precision medicine, where treatments are tailored to individual
genetic profiles.
Multifaceted Discipline
• AL and ML are the multifaceted discipline as it has many different aspects or features.
• AI a subset of computer science, has origins in mathematics, logic, philosophy, psychology, cognitive
science, and biology.
• AI, machine learning, and their place in computer science
• Alan Turing published “Computing Machinery and Intelligence” in which he asked, can machines think?
• Turing proposed that a computer could be considered to be able to think if a human evaluator could have
a natural language conversation with both a computer and a human and not distinguish between either
(i.e., an agent or system that is successfully mimicking human behavior)
Four approach's
• Acting humanly: The Turing Test approach
• The Turing Test, proposed by Alan Turing (1950), was designed to provide a satisfactory
operational definition of intelligence.
• A computer passes the test if a human interrogator, after posing some written questions, cannot
tell whether the written responses come from a person or from a computer.
• For that computer would need to possess the following capabilities:
• natural language processing: to enable it to communicate successfully in English;
• knowledge representation: to store what it knows or hears;
• automated reasoning: to use the stored information to answer questions and to draw new
conclusions;
• machine learning: to adapt to new circumstances and to detect and extrapolate(make
judgements) patterns.
• Thinking humanly: The cognitive modeling approach
• If it is said that a given program thinks like a human, we must have some way of determining how humans
think.
• There are three ways to do this:
• through introspection—trying to catch our own thoughts as they go by;
• through psychological experiments—observing a person in action; and
• Through brain imaging(Functional magnetic resonance imaging)—observing the brain in action.
• Thinking rationally: The “laws of thought” approach
• The Greek philosopher Aristotle was one of the first to attempt to codify “right thinking,” that is,
irrefutable(cannot be wrong and must be accepted) reasoning processes.
• —for example, “Socrates is a man; all men are mortal; therefore, Socrates is mortal.”
• These laws of thought are supposed to govern the operation of the mind; their study initiated the field called
logic.
• Acting rationally: The rational agent approach
• Operate autonomously,
• perceive their environment,
• persist over a prolonged time period,
• adapt to change, and
• create and pursue goals.
• A rational agent is one that acts so as to achieve the best outcome even if there is uncertainty.
 Disciplines that contribute to AI
• Philosophy
• Can formal rules be used to draw valid conclusions?
• How does the mind arise from a physical brain?
• Where does knowledge come from?
• How does knowledge lead to action?
• Mathematics
• What are the formal rules to draw valid conclusions?
• What can be computed?
• How do we reason with uncertain information?
• Economics
• How should we make decisions so as to minimize payoff?
• How should we do this when others may not go along?
• How should we do this when the payoff may be far in the future?
• Neuroscience • Linguistics
• How do brains process information? • How does language relate to thought?

• Psychology
• How do humans and animals think and act?
• Computer engineering
• How can we build an efficient computer?
• Control theory and cybernetics
• How can artifacts operate under their own control?
Omniscience, learning, and autonomy
• Difference between rationality and omniscience.
• An omniscient agent knows the actual outcome of its actions and can act accordingly; but
omniscience is impossible in reality.
• Rationality maximizes expected performance, while perfection maximizes actual
performance.
• Expecting just perfection will not be fair to agents.
• Doing actions in order to modify future percepts—sometimes called information gathering—
is an important part of rationality.
• When agent relies on the prior knowledge of its designer rather than on its own percepts, we
say that the agent lacks autonomy.
• A rational agent should be autonomous—it should learn what it can do to compensate for
partial or incorrect prior knowledge.
AGENTS AND ENVIRONMENTS
• An agent is anything that can perceive its environment through sensors and acts upon that
environment through actuators.
• This simple idea is illustrated in Figure bellow
THE NATURE OF ENVIRONMENTS

• We are ready to think about building rational agents.

• First, however, we must think about task environments, which are essentially the
“problems” to which rational agents are the “solutions.”
• Specifying the task environment
• The PEAS (Performance, Environment, Actuators, Sensors) description.
• In designing an agent, the first step must always be to specify the task environment as fully
as possible.
• Let us consider a more complex problem: an automated taxi driver.
• The full driving task is extremely open-ended. There is no limit to the novel combinations of
circumstances that can arise.
• Figure below summarizes the PEAS description for the taxi’s task environment.
Properties of task environments
• Fully observable vs. partially observable
• Single agent vs. multiagent
• Deterministic vs. stochastic
• Episodic vs. sequential
• Static vs. dynamic
• Discrete vs. continuous
• Known vs. unknown
THE STRUCTURE OF AGENTS
Simple reflex agents
Model based reflex agent

Goal-based agents
• Utility-based agents

Utility-based agents

Learning agent
Applications of AI in Healthcare
• It is not likely to happen that artificially intelligent agents will ever completely replace doctors and
nurses.
• But Machine learning and AI are transforming the healthcare industry.
• Improving diagnostics, predicting outcomes.
• It has began the personalized care systems.
Prediction
• Technologies monitoring the data to predict disease outbreaks already exists .
• This is often done using real-time data sources such as social media as well as historical information from
the Web and other sources.
• Malaria outbreaks have been predicted with artificial neural networks, analyzing data including rainfall,
temperature, number of cases, and various other data points.
Diagnosis
• Many digital technologies offer an alternative to non-emergency health systems.
• Digital tools gives a more holistic(person's physical, emotional, mental, social, spiritual, and financial
health) view of patient health, through access to data and giving patients more control over their health.
• These technologies empowers consumers to make better decisions about their own health
• Facilitates new option for prevention, early diagnosis of life-threatening diseases.
• Management of chronic conditions.
• People are using digital health technologies to:
• Reduce inefficiencies
• Improve access
• Reduce costs
• Increase quality and
• Make medicine more personalized for patients based in his/her genetic profile.
• In future, combining the genome(set of nucelic acid coded as DNA) with machine learning algorithms
could provide the opportunity to learn about the risk of disease.
• Improvement in pharmacogenetics (analyzes how the genetic makeup of a patient affects their response
to drugs), can provide better treatment pathways for patients.
Personalized Treatment and Behavior
Modification
• Digital therapeutics, a subset of digital health, is a evidence-based therapy
• Digital therapeutics (DTx) are high-quality software systems that help patients cope with various
medical conditions and illnesses through health-related tips, behavior recommendations, exercise plans,
meds intake alerts, etc.
• DTx proven products provide clinical results that impact a condition.
• This is what separates DTx from a bunch of other wellness apps and medication reminders.
• There’s a wide array of disorders and diseases DTx products cover, including obesity, ADHD (attention
deficit hyperactivity disorder), type 2 diabetes, anxiety, depression, congestive heart failure, and many
more.
Digital therapeutics key principles
• Prevent, manage, or treat medical disorders or diseases.
• Produce a software-driven medical intervention.
• Use best practices for product design, creation, deployment, management, and maintenance.
• Involve end-users in the processes of product development and testing.
• Incorporate patient privacy and security protections.
• Validate the results by appropriate regulatory bodies.
• Publish the idea in peer-reviewed journals with trial results and clinically proven outcomes.
• Collect, analyze, and apply real-world evidence.
• Examples are
• Digital therapy from Diabetes Digital Media, the Low Carb Program, assists people with type 2 diabetes
and prediabetes to reverse their condition.
Drug Discovery
• The use of machine learning in preliminary
drug discovery is useful in many ways.
• From Initial screening of drug compounds
to predicting their success rate based on
biological factors.
• This includes R&D discovery technologies
like next-generation sequencing.
Follow-Up Care
• Hospital readmittance is a huge concern in healthcare.
• Doctors, as well as governments, are struggling to keep patients healthy, particularly when returning
home following hospital treatment.
• Few Organizations have developed digital health coaches, similar to a virtual customer service
representative on an e-commerce site.
• Digital Health Coaching is equipped with current innovations and technology and combines them with
advanced behavioral science (behavioural interactions between organisms in the natural world) to aid
individuals in making better health choices.
• It also prompts questions about the patient’s medications and reminds them to take medicine, queries
them about their condition symptoms, and conveys relevant information to the doctor.
Realizing the Potential of AI in Healthcare
• For AI and machine learning to be fully embraced and integrated within healthcare systems, several key
challenges must be addressed.
Understanding Gap
• There is a huge disparity between stakeholder understanding and applications of AI and machine
learning.
• Communication of ideas, methodologies, and evaluations are pivotal to the innovation required to
progress AI and machine learning in healthcare.
• Encourage the adoption of data-driven strategies,
• Include the sharing and integration of data
• Develope data science teams, focused on learning from data, is key to a successful healthcare strategy.
• Improving value for both the patient and the provider requires data and hence data science professionals.
Fragmented Data
• Data is currently fragmented and difficult to combine.
• Patients collect data on their phones, Fitbits, and watches, while physicians collect regular biomarker and
demographic data.
• At no point is this data combined.
• Nor do infrastructures exist to analyze this larger set of data in a meaningful and robust matter.
• In addition, electronic health records (EHRs) (computer software used to capture, store, and share
patient data in a structured way), which at present are still messy and fragmented across databases,
require digitization in a mechanism that is available to patients and providers at their convenience.
Appropriate Security
• Organizations face challenges of security and meeting government regulation specifically with regards to
the management of patient data and ensuring its accessibility at all times.
• Many healthcare institutions are using legacy versions of software that can be more prone to attack.
Data Governance
• Medical data is personal and not easy to access.
• It is widely assumed that the general public would be reluctant to share their data because of privacy
concerns.
• However, a survey conducted found that 17% of people would never consent to their anonymized data
being shared with third parties.
• In this case convincing the patients and assuring them, that their data will not be miss used is a task.
Bias
• A significant problem with learning is bias.
• Within machine learning, learning to learn creates its own inductive bias based on previous experience.
• Essentially, systems can become biased based on the data environments they are exposed to.
• That is when certain elements of a dataset are overweighted or overrepresented.
• Biased datasets leads to skewed outcomes, systematic prejudice, and low accuracy
• This demands more transparent algorithms to meet the stringent regulations on drug development and
expectation.
• Transparency is not the only criteria; but decision-making must be unbiased to fully trust its abilities.
Software
• Traditional AI systems had been developed in Prolog, Lisp, and ML. Most machine learning systems
today are written in Python due to many of the mathematical underpinnings of machine learning that are
available as libraries.
• However, algorithms that “learn” can be developed in most languages, including Perl, C++, Java, and C.
Healthcare Data—Little and Big Use Cases
• Healthcare stakeholders are surrounded by masses of data from patients, professionals, and transactions.
• It is key to know how to drive value and meet KPIs (key performance indicators).
• The following are a selection of exciting healthcare data use cases.
• Predicting Waiting Times:
• Time series analysis techniques were used in certain healthcare center to predict admission rates at
different times.
• Time series analysis was used for analyzing a sequence of data points collected over an interval of time.
• Models of time series analysis include:
• Classification: assigns categories
• Curve fitting: Plots the data along a curve to study the relationships of variables within the data
• Descriptive analysis: Identifies patterns in time series data
• Explanative analysis: understand the data and the relationships within it
• Exploratory analysis: Highlights the main characteristics
• Forecasting: Predicts future data
• Intervention analysis: Studies how an event can change the data.
• Segmentation: Splits the data into segments to show the underlying properties of the source information.
• Reducing Readmissions:
• Same above approach can be used to help manage hospital costs.
• Using data analytics at risk patient groups, based on their medical history, demographics, and behavioral
data can provide the necessary care to reduce readmission rates.
• Electronic Health Records:
• EHRs haven’t yet come to fulfilment.
• The idea is theoretically simple: that every patient has a digital health record consisting of their details,
demographics, medical history, allergies, clinical results, and so forth.
• Records can be shared, with patient consent, via secure computer systems and are available for healthcare
providers from both public and private sectors.
• Each record comprises one modifiable file, which means that doctors can implement changes over time
with no danger of data replication or inconsistencies.
• Healthcare IoT—Real-Time Notifications, Alerts, Automation:
• Devices connected to the Internet currently include weighing scales; activity monitors (such as Fitbit,
Apple Watch, Microsoft Band) that measure heart rate, movement, and sleep; and blood glucose meters