Artificial Intelligence

(AI) for Healthcare

BM096IU
LUA NGO, PH.D.
FEB 2024
UNDERGRADUATE PROGRAM
Biomedical Engineering Curriculum
Semester 1 Semester 2 Semester 3 Semester 4 Semester 5 Semester 6 Semester 7 Sem. 8
(19 credits) (21 credits) (20 credits) (18 credits) (21 credits) (21 credits) (15 credits) (10 cr.)
Calculus I Calculus II Calculus III Differential Statistics for Free Elective I Free Elective II
4(4,0) 4(4,0) 4(4,0) Equation Health Sc. 3(3,0) 3(3,0)
4(3,1) 3(2,1)

Physics I Chemistry Chemistry Applied Mechanical Machine Design BME Capstone

2(2,0) for Eng.+Lab for BME Infomatics Manufacturing 3(3,0) Design Course
4(2,2) 4(3,1) 4(3,1) 4(2,2) 4(3,1)

Physics II Biology for Human Anat. Biomaterials Eng. Chal. AI for Healthcare Entrepreneur-
2(2,0) BME & Physiology 4(3,1) in Medicine I 3(3,0) ship in BME
4(3,1) 3(3,0) 3(3,0) 3(3,0)

Critical Principles of Introduction Technical Electives

Thinking EE1+Lab to BME 20(15,5)
3(3,0) 4(3,1) 4(3,1)

Academic Academic HCM thought Technical Electives Bioethics

English I English II 2(2,0) 6(3,0) 3(3,0)
4(4,0) 4(4,0)

Lab 1A Lab 1B Design 2A Design 2B Project 1 Project 2 Pre-Thesis Thesis

1(0,1) 1(0,1) 1(0,1) 1(0,1) 1(0,1) 1(0,1) 1(0,1) 10(0,10)

Triết học Chủ nghĩa Xã 2nd year summer semester 3rd year summer semester
Mác-Lênin hội khoa học Military Internship
(3(3,0) 2(2,0) training 3(0,3)
Mathematics & Basic Design Skills and
Physical Physical A B Take B after passing A
Science Projects Cluster
Training I Training II
Humanity, Social Science, Foundation Cluster
1st year summer semester and Political Cluster Take B after taking A
A B
Academic English Concentration but passing A is not
Kinh tế chính trị Mác-Lênin
Cluster Cluster required
2(2,0)
Technical Elective
Lịch sử Đảng Cộng sản Việt Nam Free Elective Cluster
2(2,0)
Cluster
Information

• Attendance: 10%
• Quiz: 10%
• Group project: 20%
• Midterm exam: 30%
• Final exam: 30%
• TA: Bùi Trung Tín (buitrungtin0105@gmail.com)
• Instructor: Ngô Thị Lụa, Ph.D., Room A1.405,
Email: ntlua@hcmiu.edu.vn
 Week Contents
1 Introduction: Historical Review, Current Status and Future Potential
2 Intro (cont.) + Coding session (Data loading, data visualization)
3 AI in Healthcare problems: Data-driven approach, Supervised vs Unsupervised Learning,
Classification, K-Nearest Neighbor
4 Coding session (Image data, Tabular Data, Signal Data, KNN)
Contents
5 AI in Healthcare problems: Regression
Linear Regression + Coding
6 Logistic Regression + Coding
7 Group Project: Data Analysis Life Cycle (Slide)
8 Unsupervised Learning: K-Means Clustering
MIDTERM
9 Neural Network: Backpropagation, Multilayer Perceptrons
10 Neural Network: Loss Functions, Gradient descent, Dropout
11 AI applications in Cancer
12 Convolutional neural networks: Convolutional, pooling, fully connected layer, data augmentation
13 Convolutional neural networks: cont.
14 Coding session (NN, CNN)
15 Group Project Presentation (Poster)
AI is the new Electricity

• AI is the new Electricity

• Electricity had once transformed
countless industries: transportation,
manufacturing, healthcare,
communications, and more

• AI will now bring about an equally

big transformation.
Artificial Intelligence Applications
What is Artificial Intelligence (AI)?

A machine is said to have artificial

intelligence if it can interpret data,
potentially learn from the data, and use
that knowledge to adapt and achieve
specific goals
(Andreas Kaplan, 2018)
What is AI?
Artificial Intelligence
Engineering of making intelligent
machines and program
(Alan Turing, 1940)

Machine Learning
Ability to learn without being explicitly
programmed
(Arthur Samuel, 1959)
 https://www.iwm.org.uk/history/how-alan-turing-cracked-the-enigma-code

Alan Turing – Father of AI

What is Machine Learning?
• Tom Mitchell (1998) Well-posed Learning Problem: A computer program is
said to learn from experience E with respect to some task T and some
performance measure P, if its performance on T, as measured by P, improves
with experience E.
Suppose your email program watches which emails you do or dot not mark as
spam, and based on that learns how to better filter spam. What is the task T in
this setting?
1. Classifying emails as spam or not spam. T
2. Watching you label emails as spam or not spam. E
3. The number (or fraction) of emails correctly classified as spam/not spam. P
4. None of the above – this is not a machine learning problem
What is Machine Learning?
AI specializations
• - Computer vision
• - Signal Processing
• - Game playing
• - Expert systems
• - Natural language processing (NLP)
• - Robotics
• - Brain Computer Interface (BCI)
Computer Vision

Processing on Image/Video data

Signal Processing
 Natural Language Processing

Auto captioning -> Auto recommendation in healthcare AI music composer

Robotics
• More general concept. We can apply computer vision (face
recognition, emotion recognition) in robotics.
• - Applying BCI in controlling robotics
• - Applying NLP to talk to people
Brain Computer Interface
Expert systems
• Expert systems are the computer based decision-making system

Interview

Clinical
Imaging
infomation

Disease
Diagnosis
 Artificial Intelligence (AI) in Healthcare
Brain
• Brain Imaging: tumor
segmentation, Alzheimer’s
disease diagnosis, etc.
• Brain Computer Interface:
controlling wheelchairs, robots,
user identifications…
• Brain Waves: decoding brain
waves, lie detection…
AI Psychological Consultants
Virtual therapists with
affordable price, chatbot to
determine and treat
depression.
Retina, Chest, Lung,
Intestine, etc.
Assisting in diagnosing
many diseases with high
accuracies such as
ophthalmological
pathologies, breast cancer,
lung cancer, prostate cancer,
hookworm detection, etc.
Smart Watch, Phone
Detecting heartbeat, blood
pressure, diabetes, glucose
level, neurological disorders
(i.e., seizure, epilepsy) etc.
20
Therapist Chatbot
Breast Cancer Mammography
Retinal Diseases Diagnosis
Supervised vs. Unsupervised Learning
• Train the machine using data • Working with unlabeled data
which is well “labeled”. • Clustering data patterns
• The machine then predict • Unsupervised data finds all kind
unforeseen data in the same of unknown patterns in data
domain
• It is easier to get unlabeled data
• Labeling is not easy than labeled data
• Challenge with huge data • Better in case of huge data
Supervised vs. Unsupervised Learning
Why is Deep Learning taking off