BIO ASSIGNMENT

Topic : AI in Healthcare
Roll number : CB.EN.U4AIE19024
Name : Dharshan Kumar
Subject code : B.Tech..2019.R.AIE.1.19BIO103
Introduction:

AI allows cheaper, better, earlier detection of diseases, which lowers the

cost of treatment and improves odds of a full recovery.

 AI is fundamentally good at pattern matching and data correlation.

 Healthcare starts with diagnosis.
 An AI can study millions of radiology images, MRI scans, x-rays,
multi-spectral photos of skin lesions, and “learn” which images are
likely to be indications of a problem. AI Is Now Helping Doctors
Diagnose Skin Cancer Faster
 At some point in the not-too-distant-future, as sensors get better
and cheaper, AI will have more data to analyze, learn from, and
make faster and cheaper diagnoses. This should be a virtuous cycle
for a period of time.

 AI can help make sure patients take all their medications in the right
amount at the right time, to improve outcomes.
 AI and wearable sensors can observe at-risk patients 24x7 and alert
them or their support people in advance or simply immediately if they’re
having a medical crisis.
 AI could listen to changes in your voice and diagnose cardiac
problems. Talking into an app could help your doctors diagnose you .
 AI can help people manage their health, by watching what they eat, how
active they are, whether their body is exhibiting chronic tension, and
other factors that can be managed with lifestyle interventions to guide
them towards better health.
 AI can operate an artificial pancreas for diabetics, monitoring blood
glucose continuously.
 AI can help doctors treat more patients, improving access to health care.

I) Inventions using AI:

• 1950s
– Computer assisted tomography
– Artificial Kidney
– Cardiac Pacemaker
– Antibiotic Production technology
• 1960s
– Ultrasound
• 1970s
– Computer assisted tomography
– Artificial hip and knee replacements
• 1980s
– Magnetic resonance imaging
– Laser surgery

Artificial kidney
II) Three types of machine learning algorithms:

1. Unsupervised (ability to find patterns)

2. Supervised (classification and prediction algorithms based on previous
examples)
3. Reinforcement learning (use of sequences of rewards and
punishments to form a strategy for operation in a specific problem
space)

III) Benefits of Artificial intelligence:

AI can definitely assist physicians

1. Clinical decision making - better clinical decisions
2. Replace human judgement in certain functional areas of
healthcare (eg, radiology).
3. up-to-date medical information from journals, textbooks and
clinical practices
4. Experienced vs fresh Clinician
5. 24x7 availability of expert
6. Early diagnosis
7. Prediction of outcome of the disease as well as treatment
8. Feedback on treatment
9. Reinforce non pharmacological management
10. Reduce diagnostic and therapeutic errors
 11. Increased patient safety and Huge cost savings associated
with use of AI
12. AI system extracts useful information from a large patient
population
13. Assist making real-time inferences for health risk alert and
health outcome prediction
14. Learning and self-correcting abilities to improve its
accuracy based on feedback.

IV)Growth drivers of AI in healthcare:

• Increasing individual healthcare expenses

• Larger Geriatric population
• Imbalance between health workforce and patients
• Increasing Global Health care expenditure
• Continuous shortage of nursing and technician staff. The number of
vacancies for nurses will be 1.2 million by 2020
• AI is and will help medical practitioners efficiently achieve their tasks
with minimal human intervention, a critical factor in meeting
increasing patient demand.

IV) Potential challenges:

• Development costs
 • Integration issues
• Reluctance among medical practitioners to adopt AI
• Fear of replacing humans
• Data Privacy and security
• Mobile health applications and devices that use AI
• Data exchange
• Need for continuous training by data from clinical studies
• Incentives for sharing data on the system for further
development and improvement of the system.
• All the parties in the healthcare system, the physicians, the
pharmaceutical companies and the patients, have greater
incentives to compile and exchange information
• State and federal regulations
• Rapid and iterative process of software updates commonly used to
improve existing products and services

V) How can AI be used:

Artificial intelligence can prove extremely beneficial in healthcare. It can

bring patients answers and advice whenever they need it, help diagnose
disease, analyze patient history, and assist medical professionals in
administrative tasks. Here are some examples of AI in healthcare:

 Administrative workflow automation

Doctors, nurses, and other care providers have to dedicate too much
time to paperwork and administrative tasks. Of course, these tasks
could be very important (keeping patient information up-to-date is a
must!) but they could use this time to actually help people. AI can
 assist medical practitioners in such work, e.g. with voice-to-text
transcriptions and advanced data analysis.
 Virtual nurses
Being stuck in a huge line in the waiting room sucks. And sometimes
you have to sit in a room full of coughing people because your knee
hurts. Let's get infected trying to fix another problem, yay! You
probably know this issue. Additionally, huge lines in clinics put
pressure on doctors to solve problems faster. Virtual nurses can
reduce the number of unnecessary hospital visits, which is beneficial
to both patients and doctors. Check out this example from the
Babylon app:

A chatbot is shown in this picture

The app uses AI in interactive symptoms checker to provide medical care.

The app asks questions and analyzes the answers, it assesses known
symptoms and risk factors and provides informed medical information.

 Diagnosis aid
AI can detect disease from medical scans. It’s successful with
identifying cases of skin cancer and in 2018, Google reported that
their AI had 99% of accuracy in metastatic breast cancer detection.
 AI diagnosis is still a work in progress but it’s exciting to read stories
like that and see the development of life-saving technology.
 Health monitoring
Wearable health and fitness monitors are gaining popularity. These
accessories provide users with information about how much they
move, how many steps they make, how fast they run, how many
calories they burn, how much (and how well) they sleep, and what
their heart rate is. All this data can be very valuable to medical care
providers and could be shared with e.g. the doctor to provide real-
time alerts about possible health issues.
 Robot-assisted surgery
Robotic surgeries are considered minimally invasive and extremely
precise. In a robotic procedure, the pre-op medical records are
integrated with real-time operating metrics to improve the
outcomes. This technique allows for better physician precision and
can lead to a 21% reduction in length of patient’s post-operation
hospital stay.
AI in healthcare is more than just analyzing medical records, it also presents
opportunities like AI diagnostics, virtual assistants, wellness management. AI
can sense, understand, and perform actions to support people in both
administrative and clinical functions.

1. One of AI's greatest potential advantages is to enable individuals to

remain healthy like they don’t need a specialist. The utilization of AI and the
Internet of Medical Things (IoMT) in healthcare applications is now helping
individuals. For e.g. Healthify, Google Health etc.

2. Innovative applications related to medical or health energize more

beneficial conduct in people and help with the proactive administration of a
strong way of life.

3. AI expands the capacity for health experts to understand & analyze more
the everyday routine and necessities of the general group of people they
care for, and with that understanding, they can give better input, direction,
and support for remaining healthy.

4. AI is already in complete utilization for detecting diseases, like cancer and

more predictive in the early stage. This saves time and even the treatment
gets accurate and in time. A survey says it works 30 times faster and predicts
99% accurate.

5. The various customer wearables and other health-related gadgets joined

with AI are additionally being connected to regulate heartbeats to check
illness, empowering doctors and different specialists to all the more likely
screen and distinguish conceivably hazardous situation at prior, more
treatable stages.

6. Google's DeepMind Health works in associate with clinicians, scientists,

and patients to take care of genuine medicinal services issues. The
innovation joins machine learning and frameworks neuroscience to
manufacture incredible universally useful learning calculations into neural
systems that work with human brains.

7. Utilizing design acknowledgment to distinguish patients in danger or

building up a condition – or analyzing, break down because of the way of life
or different factors – is another zone where AI is starting to grab hold in
medicinal services.

Recent work has shown promising results in image classification:

 Skin lesions
 Retinal hemorrhage
 Pathology images
 Signal processing – EEG, ECG data
 Taking Pathology image Pathology image
Conclusion:

“AI will have an incredible impact on healthcare… and already has. I would
redefine the acronym ‘AI’ from ‘Artificial Intelligence’ to ‘Augmented
Intelligence’ since the benefits of cognitive computing is focused on the
ability to augment and enhance clinical decision making. Augmented
intelligence lets both humans and machines do what they do best. AI holds
the promise of improving quality while reducing cost. Just a quick look
at these results reveals multiple discussion illustrating this.

What makes AI compelling is that data, even about a single patient can be
staggering. It is impossible for a provider to review a comprehensive medical
record or the contents of an Electronic Medical Record (EMR) during a brief
patient encounter. Cognitive systems change the landscape by consuming
specific patient data and comparing it massive amounts of medical research
to provide the clinician with relevant patient data and supporting treatment
research, driving a more accurate differential diagnosis.

Clinicians are not replaced by cognitive intelligence but instead enhance by a

tool that can consume and analyzes data beyond human comprehension.
These systems synthesize extensive amounts of data into actionable events.
Cognitive systems excel by providing confidence levels for
various healthcare recommendations.

An excellent example is where Watson provides evidence-based treatment

options for oncology patients based on outcomes and allows
for personalizing therapy.

In medical research, this cognitive capability will allow analysis of research in

various languages, something that until recently would have been
impossible. Watson is fluent in nine languages (and counting). ‘Rather than
pattern matching, Watson is taught to understand the structure of
languages, using a combination of natural language processing and machine
learning. Language experts use representative sentences to teach Watson
how to parse - to identify the verb, nouns, adjectives and other parts of
speech’.

In summary, cognitive systems like Watson can quickly analyze and index:
research, web text, video, and medical data at unprecedented speed and
scale. This capability does not replace the clinician but augments the
physician’s capacity by providing analytics from data sets too large for
human’s consumption.”

Figures showing the AI improvements in healthcare:

Year wise increase of AI in different healthcare streams:

Year wise increase of AI in finding solution against top 10 diseases:

Percentage of AI algorithms used in health care: