AGRA PUBLIC PHARMACY COLLEGE,

ARTONI

Event: Artificial Intelligence In Heath Care Workshop

Date: 26th-28th Oct 2023
Day: Thursday Friday Saturday
Number of participants: 52
Venue: Seminar hall and Computer Lab

Event Coordinator

ISLAMUN JAHAN
 EVENT REPORT

TRAINER DETAIL:
Mr Deepesh Data science trainer and consultant at Ernst & Young
Gurgaon, India.

Consultant at Adobe Hyderabad

Core Competencies of Mr Deepesh are Entrepreneurship, Statistical Modeling, Statistical

Data Analysis, Statistical Consulting, Statistical Programming, Business Analytics, Risk
Management, SAS Programming, SAS E-Miner, Python, Micro Strategy, JMP

Mr. Deepesh conducted trainings at various renowned institutions like IIT, Jodhpur, IIT,
Delhi,Wingfotech, Delhi,Kyiron, Delhi, Ion Media, Noida, Aptron, Noida, Marathwada Mitra
Mandal, Pune, BMS College Of Engineering, Gurgaon, IIT, Kanpur, RCPL, Unique Infosystem,
Zamia Hamdard, Delhi, Zamia Islamia, Delhi, VGI, Dadri, AKGEC, Ghaziabad, IIT –
Bhuvneswer, NIT –Warangal, IIT Indore,Techtrunk Ventures –Hyedrabad, Cetpa Infotech –
Noida.

SUMMARY:
A workshop on Artificial Intelligence was conducted in Agra Public Pharmacy College from
26th to 28th October 2023.This session started at 10:00am. The workshop was conducted for
Undergraduate Students, Post Graduate Students. The workshop was conducted in association
with makeintern and E-Cell IIT Kharagpur.The speaker and trainer: Mr. Deepesh Kumar Mishra.

All students actively participate in the workshop. Students found workshop very useful. It helps
them to gain the knowledge in use of advance technology i.e., artificial intelligence, data science,
use of these technology for industrial growth, modern advancement era of industrialization that
how we move toward industry 3.o to industry 4.o model in which we deal with data science and
development of artificial intelligence in health care science for example of neuralink corp. is an
American neurotechnology company that is developing implantable brain–computer interfaces,
based in Fremont, California, as of 2022 which is Founded by Elon Musk.

Now from workshop students really got to understand the real meaning of industry 4.o model,
data science, artificial intelligence & the developing science behind artificial intelligence and
more much.
The purpose of workshop was too introduced to the exciting world of Artificial Intelligence in
Healthcare. Students explore the fundamental concepts of AI and Machine Learning and
discover its wide-ranging applications within the pharmaceutical industry.

The workshop is divided for 3 day program, the Time Table of 3 days
program on Artificial Intelligence in Healthcare
Day One Day Two Day Three
Introduction to AI and Data Cleaning and Predictive Making AI Bots with Medical
Visualization with POWER Regression with Python Content
BI
Workshop Report

Day- 1

Date: 26th October 2023

Session 1: Introduction to AI and Visualization with POWER BI

In this session students cover topics that are

 Understanding the basics of AI and its applications in healthcare

 Introduction to data visualization with POWER BI
 Exploring and visualizing healthcare data with POWER BI
 Case studies and examples of AI in healthcare

Introduction to Artificial Intelligence (AI):

Artificial Intelligence (AI) is a rapidly advancing field in computer science that focuses on
creating systems and machines capable of performing tasks that typically require human
intelligence. These tasks include understanding natural language, recognizing patterns, solving
problems, and making decisions. These technologies aim to simulate human cognitive functions,
such as reasoning, learning, problem-solving, and adaptation, and apply them to a wide range of
applications.

Artificial Intelligence (AI) is making significant strides in the healthcare industry, enhancing
various aspects of patient care, diagnosis, treatment, and administrative processes as we can see
that how in healthcare AI is use:

1. Medical Imaging: AI is widely used in medical imaging to assist with the interpretation
of complex images, such as X-rays, MRIs, and CT scans. It can help identify
abnormalities, tumors, and other conditions. Examples include:

 Deep Learning for Diabetic Retinopathy: AI can analyze retinal images to

detect diabetic retinopathy and other eye diseases.

 CAD (Computer-Aided Diagnosis): AI aids radiologists by highlighting

potential anomalies in mammograms or lung CT scans.

2. Disease Identification and Diagnosis:

 Pathology Analysis: AI can analyze tissue samples for cancer diagnosis,

classifying different types and stages.

 Skin Cancer Detection: AI-powered apps can identify and classify skin lesions
for potential malignancies.
 3. Drug Discovery and Development:

 Drug Repurposing: AI analyzes existing drugs to identify new uses, potentially

accelerating the discovery of treatments for various diseases.

 Molecule Analysis: AI helps researchers predict the biological activity of

molecules, speeding up the drug development process.

4. Personalized Treatment Plans: AI can analyze patient data, including genetic

information, to create personalized treatment plans and predict how individuals will
respond to specific medications or therapies.

5. Predictive Analytics: AI models can forecast disease outbreaks, patient readmissions,

and identify individuals at higher risk of specific health conditions. This is valuable for
preventive care and resource allocation.

6. Natural Language Processing (NLP): NLP techniques are used for mining electronic
health records (EHRs) to extract valuable information for research, billing, and clinical
decision support.

7. Telemedicine: AI-driven chat bots and virtual health assistants help answer patient
queries, schedule appointments, and provide general healthcare information.

8. Robotic Surgery: AI-assisted robotic surgical systems can enhance the precision and
dexterity of surgeons in procedures like minimally invasive surgery.

9. Medication Management: AI-powered systems help patients manage their medications,

sending reminders and monitoring adherence.

Data visualization with Power BI

Introduction

Data visualization with Power BI involves using Microsoft Power BI, a business intelligence
tool, to create visually compelling and interactive reports and dashboards from your data. These
visualizations make it easier to understand, analyze, and communicate insights from your data,
helping organizations make informed decisions and gain valuable insights. Power BI provides a
user-friendly interface for connecting to data sources, transforming and modeling data, and
creating various types of charts, graphs, and tables. These visualizations can be customized,
shared, and accessed on various devices, making it a powerful tool for data-driven decision-
making in businesses and organizations.

Exploring and visualizing healthcare data with Power BI enables healthcare professionals and
organizations to make data-driven decisions, enhance patient care, optimize operations, and
improve healthcare outcomes. It also facilitates better communication and collaboration among
healthcare teams and supports research and development in the field.

Using this tool we gain insights and create visual representations from healthcare-related data.
This process can help healthcare organizations, professionals, and researchers better understand
patient outcomes, treatment effectiveness, and operational efficiency. Here's how it works:

 Data Integration: Import healthcare data from various sources, such as electronic health
records (EHRs), clinical databases, or health monitoring devices, into Power BI. You can
connect to different data repositories and perform data integration to create a unified
dataset.

 Data Cleaning and Transformation: Cleanse and transform the data to ensure its
accuracy and consistency. This step involves handling missing values, standardizing
formats, and converting data into a suitable structure for analysis.

 Data Modeling: Define relationships between different data tables and create calculated
fields or measures. This step helps in preparing the data for visualization and analysis.

 Visualization: Use Power BI's wide range of visualization options, such as charts,
graphs, maps, and tables, to create interactive reports and dashboards. You can represent
patient demographics, disease trends, treatment outcomes, and more.

 Interactive Dashboards: Develop interactive dashboards that allow users to filter, drill
down, and explore the data. This interactivity helps in uncovering meaningful insights
and patterns.

 Real-time Monitoring: Incorporate real-time data feeds from health monitoring devices
or sensor networks to track patient vital signs, disease trends, or other critical metrics.

 Predictive Analytics: Leverage AI and machine learning models within Power BI to

predict disease outbreaks, patient readmissions, or individual patient outcomes.

 Patient Outcomes: Visualize and compare patient outcomes and treatment effectiveness,
which can be valuable for healthcare providers and researchers.

 Cost Analysis: Analyze healthcare costs, resource allocation, and cost-effectiveness

using Power BI's data visualization capabilities.

 Secure Sharing: Share the reports and dashboards securely with relevant stakeholders,
ensuring data privacy and compliance with healthcare regulations.

Case studies and examples of AI in healthcare

 AI is increasingly being applied in healthcare to improve patient care, diagnosis, and
operational efficiency. Some case studies and examples of AI in healthcare:
1. IBM Watson for Oncology:
 Case Study: IBM Watson for Oncology analyzes vast amounts of medical
literature and patient records to provide personalized treatment recommendations
for cancer patients.
 Outcome: It assists oncologists in making treatment decisions based on the latest
medical knowledge, potentially leading to more effective and tailored cancer
therapies.
2. Path AI:
 Case Study: Path AI uses deep learning to assist pathologists in diagnosing
diseases from pathology slides.
 Outcome: By automating the analysis of pathology slides, PathAI can enhance
diagnostic accuracy and reduce the workload on pathologists.
3. Zebra Medical Vision:
 Case Study: Zebra Medical Vision employs AI to analyze medical imaging data,
including X-rays, CT scans, and MRIs, to detect a wide range of diseases.
 Outcome: This technology can identify conditions such as lung cancer,
osteoporosis, and liver disease at an early stage, potentially improving patient
outcomes.
4. Google Health:
 Case Study: Google Health developed an AI model to detect diabetic retinopathy
from retinal images.
 Outcome: This AI system helps identify eye disease in diabetic patients, allowing
for early intervention and preventing vision loss.
5. Tempus:
 Case Study: Tempus applies AI to analyze clinical and molecular data to assist
oncologists in making data-driven treatment decisions.
 Outcome: It supports personalized cancer treatment by identifying the most
effective therapies and clinical trial options for each patient.
6. Pathfinder Software's Cardio Care:
 Case Study: Cardio Care is a cloud-based AI platform that analyzes ECG data to
detect cardiac arrhythmias.
 Outcome: The system can provide early detection of heart irregularities, enabling
timely medical intervention and reducing the risk of heart-related complications.
7. Butterfly Network:
 Case Study: Butterfly Network developed an AI-powered handheld ultrasound
device called the Butterfly iQ.
 Outcome: This portable ultrasound system, combined with AI, provides cost-
effective and accessible imaging for healthcare professionals, particularly in
resource-constrained environments.
8. Early Sense:
 Case Study: Early Sense uses AI to monitor patients in hospitals and nursing
homes by analyzing data from under-the-mattress sensors.
 Outcome: It helps in early detection of patient deterioration, reducing the risk of
adverse events and improving patient safety.
Day- 2

Date: 27th October 2023

Session 2: Data Cleaning and Predictive Regression with Python

In this session students cover topics that are

 Introduction to Python and its applications in healthcare

 Understanding data cleaning and its importance in AI

 Data cleaning techniques and strategies using Python

 Predictive regression models in healthcare using Python

 Case studies and examples of predictive regression in healthcare

Introduction to Python and its applications in healthcare

Python is a versatile, high-level programming language known for its simplicity and readability.
It has gained popularity in healthcare and medical research due to its extensive libraries, data
analysis capabilities, and ease of use. Python's versatility and its extensive ecosystem of libraries
make it a valuable tool for healthcare professionals, researchers, and organizations.

Here's an introduction to Python and its applications in healthcare:

Python Application in Healthcare:

1. Data Analysis and Visualization:

 Python, along with libraries like Pandas and Matplotlib, is used to process and
analyze healthcare data, such as electronic health records (EHRs), clinical trial
data, and medical images.

 Researchers and healthcare professionals can create interactive visualizations to

gain insights from large datasets, helping in disease monitoring, patient outcomes,
and resource allocation.

2. Machine Learning and AI:

  Python is a primary language for implementing machine learning and artificial
intelligence (AI) algorithms. Libraries like Scikit-Learn and TensorFlow are
extensively used in healthcare applications.

 Machine learning models can predict patient outcomes, detect diseases from
medical images, and aid in personalized treatment plans.

3. Medical Imaging:

 Python, along with libraries like OpenCV and PyTorch, is used to process and
analyze medical images from various modalities, including X-rays, CT scans, and
MRIs.

 Machine learning models can identify anomalies and assist radiologists in

diagnostics.

4. Healthcare Analytics:

 Python is employed to develop healthcare analytics solutions, such as patient risk

prediction models, cost analysis, and resource optimization.

 It helps healthcare organizations make data-driven decisions to improve patient

care and operational efficiency.

5. Telemedicine and Mobile Apps:

 Python is used to develop telemedicine platforms and mobile healthcare

applications, enabling remote consultations, health monitoring, and medication
management.

6. Genomic Analysis:

 Python, along with libraries like Biopython, is applied in genomic data analysis to
study genetic variations and their association with diseases.

 It helps in personalized medicine and genetic research.

7. Epidemiology and Public Health:

 Python is used in epidemiological modeling to study disease spread and plan

interventions.

 Public health agencies rely on Python for data analysis and visualization to make
informed decisions in disease control and prevention.

8. Clinical Trials:
  Python is used for data management and statistical analysis in clinical trials,
helping researchers evaluate the safety and efficacy of new treatments.

Understanding data cleaning and its importance in AI

Data cleaning, also known as data-preprocessing or data cleansing, is a crucial step in the data
preparation process for artificial intelligence (AI) and machine learning (ML) applications. It
involves identifying and rectifying errors, inconsistencies, and inaccuracies in the dataset to
ensure that the data used for AI models is accurate, reliable, and free from noise.

Why to Understanding Data Cleaning:

1. Handling Missing Values: Data cleaning involves addressing missing data points.
Missing data can lead to inaccurate conclusions in AI models. Techniques like
imputation, which fill in missing values, are applied to ensure that the dataset is
complete.

2. Dealing with Outliers: Outliers are data points that significantly deviate from the norm.
Data cleaning identifies and handles outliers, which can affect the performance of AI
algorithms by skewing results.

3. Data Type Conversion: Sometimes, data may be in the wrong format, requiring
conversion to the appropriate data type (e.g., converting text to numeric values).

4. Handling Duplicates: Identifying and removing duplicate entries or records is essential

to maintain data integrity. Duplicate data can distort model training and lead to biased
results.

5. Encoding Categorical Variables: AI models often require numerical input data. Data
cleaning includes encoding categorical variables into numerical values, such as one-hot
encoding.

6. Data Scaling and Normalization: Data cleaning also encompasses scaling and
normalizing features to ensure that different features are on similar scales. This helps
models perform optimally.

Importance of Data Cleaning in AI:

1. Improves Model Performance: Clean data provides a more accurate representation of

the real world, leading to AI models that perform better, make more accurate predictions,
and yield more reliable results.

2. Reduces Bias: Biased or incomplete data can introduce bias into AI models, leading to
discriminatory or inaccurate predictions. Data cleaning helps mitigate this issue by
ensuring a more representative dataset.
 3. Enhances Generalization: Cleaned data aids AI models in generalizing patterns and
relationships from the training data to new, unseen data, increasing their effectiveness in
real-world applications.

4. Saves Time and Resources: Training AI models on dirty data can result in inefficient
model convergence and the need for extensive retraining. Data cleaning reduces the time
and computational resources required for model development.

5. Improves Data-Driven Decision-Making: Reliable and accurate data, free from errors
and inconsistencies, supports data-driven decision-making in various industries, including
healthcare, finance, and marketing.

6. Mitigates Legal and Ethical Risks: Data cleaning helps prevent the use of inaccurate or
outdated information, reducing the risk of legal and ethical issues associated with making
decisions based on flawed data.

Data cleaning techniques and strategies using Python

Data cleaning in Python involves a combination of techniques and strategies to identify and
address errors, inconsistencies, and inaccuracies in your dataset.

Predictive regression models in healthcare using Python

Predictive regression models in healthcare using Python involve using regression analysis to
predict healthcare outcomes, such as patient health metrics, disease progression, or treatment
effectiveness.

Case studies and examples of predictive regression in healthcare

Predictive regression models are widely used in healthcare for a variety of applications. Here are
some case studies and examples of predictive regression models in healthcare:

1. Diabetes Progression Prediction:

 Case Study: Researchers developed a predictive regression model to estimate the

progression of diabetes in patients based on their historical health data.

 Outcome: The model allowed healthcare providers to identify individuals at

higher risk of diabetes complications and tailor treatment plans accordingly.
2. Blood Pressure Control:

 Case Study: A predictive regression model was used to predict blood pressure
levels in hypertensive patients based on lifestyle factors, medication adherence,
and genetic markers.

 Outcome: Physicians were able to adjust treatment plans and interventions to

better manage blood pressure in these patients.

3. Mortality Risk Assessment:

 Case Study: A hospital implemented a regression model to predict the risk of

mortality for critically ill patients in the intensive care unit (ICU).

 Outcome: The model assisted healthcare professionals in prioritizing care and

resource allocation for patients at higher risk of adverse outcomes.

4. Cardiovascular Disease Risk Prediction:

 Case Study: Researchers developed a predictive regression model to estimate an

individual's risk of developing cardiovascular disease based on factors like age,
gender, cholesterol levels, and lifestyle.

 Outcome: This tool allows healthcare providers to identify at-risk patients for
early intervention and lifestyle modification.

5. Cancer Survival Prediction:

 Case Study: Researchers developed regression models to predict cancer survival

rates based on patient characteristics, disease stage, and treatment data.

 Outcome: These models provide valuable insights for oncologists in assessing the
prognosis of cancer patients and determining the most appropriate treatment
options.
Day- 3

Date: 28th October 2023

Session 3: Making AI Bots with Medical Content

In this session students cover topics that are

 Introduction to AI bots and their applications in healthcare

 Understanding medical content and its relevance in AI bots
 Building AI bots for medical content using popular frameworks
 Integrating AI bots with healthcare systems and workflows
 Case studies and examples of AI bots in healthcare

Introduction to AI bots and their applications in healthcare

AI bots, also known as chat bots or virtual assistants, are computer programs powered by
artificial intelligence (AI) that are designed to interact with users through natural language
processing and respond to their queries or requests. In healthcare, AI bots are being increasingly
adopted for various applications to improve patient care, streamline administrative tasks, and
enhance the overall healthcare experience.

Introduction to AI Bots in Healthcare:

AI bots in healthcare are intelligent software applications that leverage AI technologies,

including natural language processing, machine learning, and expert systems, to perform a range
of functions in the healthcare domain. These bots can interact with patients, healthcare
professionals, and administrative staff, providing timely information, support, and assistance.

Applications of AI Bots in Healthcare:

o Patient Engagement and Education:

 AI bots can engage with patients to provide information about medical conditions,
treatment options, and lifestyle recommendations.

 These bots can also assist patients in adhering to medication regimens and
tracking their health progress.

o Appointment Scheduling and Reminders:

 AI bots can schedule medical appointments, send appointment reminders, and

help patients reschedule or cancel appointments as needed.

 This streamlines the administrative process and reduces no-shows.

o Telemedicine Support:

 In telemedicine, AI bots can facilitate virtual consultations between patients and

healthcare providers.

 They can assist with gathering patient medical history and guiding patients
through the virtual visit.

o Health Monitoring and Alerts:

 AI bots can monitor patient data, such as vital signs and symptoms, and provide
real-time alerts to healthcare providers or patients when there are concerning
changes.

 This is especially valuable for remote patient monitoring and early intervention.

o Chat-Based Triage:

 AI bots can help in the initial assessment of patients' symptoms and recommend
appropriate actions, such as seeking immediate medical attention or providing
home care advice.

o Administrative Tasks Automation:

 In healthcare administration, AI bots can handle routine tasks like verifying

insurance information, processing billing inquiries, and managing medical
records.

o Access to Medical Information:

 AI bots can provide healthcare professionals with quick access to medical

databases, research articles, and treatment guidelines, aiding in clinical decision-
making.

o Mental Health Support:

 AI chat bots can offer mental health support, assisting individuals in managing
stress, anxiety, and depression.

 They can also provide crisis intervention and guide users to appropriate resources.

o Clinical Documentation:

 AI bots can assist clinicians in documenting patient encounters, converting

spoken words into text, which can save time and reduce administrative burdens.
 o Patient Feedback and Surveys:

 AI bots can collect patient feedback and conduct surveys to gauge patient
satisfaction and identify areas for improvement in healthcare services.

Understanding medical content and its relevance in AI bots

Understanding medical content is crucial for the development and effectiveness of AI bots in the
healthcare domain. AI bots need to have a solid grasp of medical knowledge and content to
provide accurate, reliable, and contextually relevant information and support to healthcare
professionals, patients, and administrators. The accuracy and relevance of the information
provided by these bots are critical for their acceptance and impact on patient care and healthcare
decision-making. Combining advanced natural language processing, machine learning, and
comprehensive medical knowledge allows AI bots to play a valuable role in the healthcare
ecosystem by providing accurate, up-to-date, and contextually relevant information and support.

The importance of medical content and its relevance in AI bots:

1. Accurate Information and Decision Support:

 AI bots must possess a deep understanding of medical terminology, diseases, treatments,

and procedures. This knowledge enables them to provide healthcare professionals with
accurate and relevant information.

 AI bots can assist in clinical decision support by offering evidence-based

recommendations and aiding healthcare providers in diagnosis and treatment planning.

2. Patient Education and Engagement:

 Healthcare AI bots can interact with patients, explaining medical conditions, treatment
options, and lifestyle modifications. To be effective, they need to convey complex
medical concepts in a patient-friendly manner.

 These bots can provide up-to-date information on health topics, medications, and
preventive measures, helping patients make informed decisions.

3. Clinical Documentation and Data Entry:

 AI bots can assist healthcare providers with clinical documentation, making it important
that they understand medical content and can accurately transcribe spoken words into
medical records.
  They must also be aware of the latest coding and billing guidelines for insurance and
reimbursement purposes.

4. Telemedicine and Virtual Consultations:

 AI bots are increasingly used in telemedicine. They need to understand medical content
to facilitate virtual consultations by collecting patient medical history, symptoms, and
other relevant information.

 The bot's ability to ask relevant follow-up questions and provide accurate information is
vital for quality virtual care.

5. Interpretation of Medical Images:

 In applications like radiology, AI bots must comprehend medical imaging terminology

and be able to assist in interpreting medical images such as X-rays, CT scans, and MRIs.

 Their ability to provide insights into image findings is essential for aiding radiologists in
the diagnostic process.

6. Compliance and Regulations:

 AI bots need to be aware of healthcare regulations, privacy laws, and data security
standards, ensuring that they provide responses and services that align with legal and
ethical guidelines.

7. Continuous Learning and Updates:

 Medical knowledge is constantly evolving, with new research, treatments, and guidelines
emerging regularly. AI bots should be capable of continuous learning and updating their
knowledge base to stay current.

8. Multi-Lingual Support:

 In healthcare, access to medical information and services is often required in various

languages. AI bots must be proficient in multiple languages to cater to diverse patient
populations.

9. Mental Health and Emotional Support:

 AI bots designed for mental health support need to understand psychological concepts,
treatment approaches, and empathy to effectively engage with individuals experiencing
emotional distress.
Building AI bots for medical content using popular frameworks
Building AI bots for medical content involves creating intelligent software applications capable
of understanding, processing, and providing relevant medical information to users. Popular
frameworks and libraries for developing such AI bots include Python-based tools and natural
language processing libraries.

The steps to build AI bots for medical content using these frameworks:

Step 1: Define the Bot's Purpose and Use Cases

 Determine the specific use cases for your medical AI bot, such as patient education,
clinical decision support, or telemedicine assistance. Define the bot's objectives and the
type of medical content it will handle.

Step 2: Choose a Framework and Libraries

 Select the frameworks and libraries that align with your project requirements. For
medical AI bots, popular choices include:

 Dialogflow: A Google Cloud service that simplifies the creation of natural

language processing chatbots.

 Rasa: An open-source conversational AI platform that enables the development

of AI-driven chatbots.

 SpaCy: A natural language processing library in Python for text processing and
understanding.

 Scikit-learn: A Python library for machine learning, which can be used for
classification and regression tasks.

 TensorFlow or PyTorch: Deep learning frameworks for training and deploying

machine learning models.

Step 3: Data Collection and Preparation

 Gather medical data and content that your bot will use for training and reference. This
can include medical literature, textbooks, research articles, and relevant datasets.

 Preprocess the data by cleaning, structuring, and formatting it for use in training and
understanding by the bot.

Step 4: Model Training

  Train machine learning models or natural language processing models using the selected
framework. This step is critical for teaching the bot to understand and respond to medical
content accurately.

 Use labeled medical data for supervised learning, and fine-tune pre-trained language
models for more advanced tasks.

Step 5: Natural Language Processing

 Implement natural language processing techniques to understand and interpret user input.
This involves tasks like named entity recognition, sentiment analysis, and text
classification.

 Utilize the selected libraries and frameworks to build language understanding

capabilities.

Step 6: Intent Recognition

 Define the medical intents your bot needs to recognize and handle. Intents can include
queries about symptoms, treatment options, medication information, or general medical
advice.

 Train the bot to identify these intents in user queries.

Step 7: Context Management

 Implement context management to maintain the conversation flow and remember

previous user interactions. This helps in handling multi-turn conversations effectively.

Step 8: Response Generation

 Develop a mechanism for generating informative and contextually relevant responses to

user queries. This can involve pulling information from medical databases, research
articles, or pre-existing knowledge bases.

Step 9: Integration

 Integrate the AI bot into the desired platforms, such as web applications, mobile apps, or
telemedicine systems. Ensure that it can interact seamlessly with users.

Step 10: Testing and Validation

 Thoroughly test the bot using real-world medical queries and user scenarios. Validate its
responses for accuracy and relevance.

 Fine-tune the bot based on user feedback and performance.

Step 11: Continuous Learning and Updates

 Keep the bot's medical content up to date by regularly updating its knowledge base with
the latest research and medical guidelines.

 Monitor user interactions and adapt the bot to provide better responses over time.

Integrating AI bots with healthcare systems and workflows

Integrating AI bots with healthcare systems involves seamlessly embedding intelligent software
into medical processes. AI bots streamline administrative tasks, facilitate telemedicine, and assist
healthcare professionals and patients. They must ensure data security, compliance with privacy
regulations, and compatibility with electronic health records (EHR) systems. By connecting with
healthcare databases and applications, AI bots enhance decision-making, provide real-time
information, and improve patient care, ultimately contributing to more efficient healthcare
workflows and a better patient experience.

Case studies and examples of AI bots in healthcare

AI bots in healthcare have been applied in various scenarios to enhance patient care, streamline
administrative tasks, and provide information and support.

Case studies and examples of AI bots in healthcare:

1. Woebot - Mental Health Support:

 Case Study: Woebot is an AI chatbot that offers mental health support and
cognitive-behavioral therapy to users through conversational interactions.

 Outcome: Woebot has demonstrated effectiveness in reducing symptoms of

depression and anxiety, providing accessible mental health support to a wide
range of users.

2. IBM Watson Health - Clinical Decision Support:

 Case Study: IBM Watson for Oncology offers clinical decision support by
analyzing vast amounts of medical literature and patient records to recommend
personalized cancer treatment options.

 Outcome: It aids oncologists in making more informed treatment decisions,

potentially improving cancer care outcomes.

3. Ada - Symptom Checker:

 Case Study: Ada is an AI bot that helps users identify medical conditions based
on their reported symptoms and health history.
  Outcome: Ada provides users with initial guidance on potential diagnoses and
encourages seeking medical advice when needed.

4. Suki.AI - Medical Documentation Assistant:

 Case Study: Suki.AI is an AI-powered medical documentation assistant that

helps healthcare professionals with clinical documentation, reducing the time
spent on administrative tasks.

 Outcome: It improves physician productivity, allowing them to focus more on

patient care.

5. Buoy Health - Diagnostic Chatbot:

 Case Study: Buoy Health's AI chatbot assists users in understanding their

symptoms and offers guidance on whether to seek medical attention.

 Outcome: It helps users make informed decisions about their health and can
reduce unnecessary doctor visits.

6. Cleo - Maternity Support:

 Case Study: Cleo is an AI bot that provides maternity support, guidance, and
information to expectant mothers.

 Outcome: It helps women navigate the challenges of pregnancy and postpartum

care, offering personalized advice and support.

7. ChatGPT - Healthcare FAQs:

 Case Study: ChatGPT and similar AI models have been used on healthcare
websites and chat platforms to answer common medical questions and provide
information to users.

 Outcome: They assist in delivering accurate and readily available information to

users, improving the healthcare information-seeking process.

8. Mediktor - Triage and Symptom Analysis:

 Case Study: Mediktor is an AI bot used for medical triage and symptom analysis,
helping users determine the urgency of their health concerns.

 Outcome: It contributes to more efficient healthcare resource allocation by

guiding patients to the appropriate level of care.
Practically students perform all the workshop day wise

On Day 1:

Introduction to Python and some basic coading of Python

After introduction, here at day 1 we learn about tools like Power BI for visualization, Google
colab and jupyter from anaconda which are use for code python language and run it.

 By this code students learn about class of our value which students add in program

Cod
e
a=5

b=2.3

c='victor'

print(type(a))

print(type(b))

print(type(c))
Output

<class 'int'>

<class 'float'>

<class 'str'>
  By this code students learn about create list of data in program

Code
list1=['physics','chem','1','2.3']

print (list1)

print(type(list1))

len(list1)

print(list1[-1])
Output list1[0]='pharmacy'
['physics', 'chem', '1', '2.3']

<class 'list'>

2.3

['pharmacy', 'chem', '1', '2.3']

  By this code students learn about tuple in program in which students can’t change input

Code
#tuple

xyz=('physics','chem','1','2.3')

type(xyz)
Output
tuple

 By this code students learn about import pandas and use for data frame as df in program

Code
import pandas as pd

df=pd.DataFrame(pharmacy)

df
Output

Class Interaction on day 1

Introduction to AI

Seeding the knowledge of Python

On Day 2:

Introduce to the tools of visualization in coading of Python for representing

data and data cleaning or analyzing
 Importing data from command
data=pd.read_csv('/content/data.csv')

data
Outcome

 data from command for describe for calculate possibilities

data.describe()

Outcome

 Code for import and use of Matplotlib which make your data into visualization output
 import matplotlib.pyplot as plt

x=[4,4,8,8]

y=[4,8,4,8]

x2=[5,6,9,8]

y2=[4,8,4,8]

plt.plot(x,y,marker='o',label='x',color='r')

plt.plot(x2,y2,marker='*',label='y',color='g')

plt.legend()

plt.xlim(1,10)

plt.ylim(1,10)

plt.show()
Outcome

 Some other visualization command

Pie Plot in python

Plot a Heat map in python

Plot a pair plot in python
Class interaction on day 2
On Day 3:

AI Bot coading in Python and Presentation

 Code of bot

Class interaction on day 3

At the end of workshop students explore about the world of artificial intelligence and learn its
use in health science for predicting the disease chance using database, drug discovery , in clinical
trials and also about the AI bots as a helping hand in modern era in many expect such as
suggestion , make our research easier and so on. After workshop using python learn about data
science ,developing AI and more much.