AI’s ROLE IN MODERN MEDICINE

Gaurisha Sisodia
Assistant Professor
Galgotias College of Engineering and Technology Greater Noida
gaurisha@[Link]

Dr. Imtiaz Ahmed

Assistant Professor
Galgotias College of Engineering and Technology Greater Noida
imtiazahmed@[Link]

This chapter introduces Artificially intelligent systems are also being applied in the
healthcare sector to enhance patient experience, patient care, and provide support to
physicians through the use of AI assistants. Companies, such as BotMD have built
systems .
Abstract— Medical sciences make substantial use of artificially intelligent computer systems.
Common uses include patient diagnosis, complete drug development and discovery,
enhancing patient-physician contact, transcribing medical records, including prescriptions,
and patient treatment from a distance. Modern computer algorithms have lately attained
accuracy levels comparable to those of human experts in the medical sciences, despite the
fact that computers still often perform tasks more quickly than humans. There are conjectures
that human replacement of humans in specific medical science roles is just a question of time.
This article aims to disentangle hype from reality by examining the ways in which artificial
intelligence is transforming the field of medical science.

Keywords: Assistant, artificial intelligence, training, clinical, surgery, discovery,

nursing

1.1 Introduction
AI i.e Artificial intelligence is the technique that enables robots to carry out tasks assigned
by humans that need intellect comparable to that of humans, such as voice recognition for
virtual assistants like Alexa and Siri and facial recognition for identity recognition. Elderly or
blind passengers are also receiving assistance from autonomous or self-driving cars.
Computers can now read retinal impulses with the same accuracy as a trained professional
according to Google DeepMind. According to He et al. (2019), Babylon, a wellness
application, asserts that its chatbot can easily pass exams for general practitioners. AI in
healthcare offers patients a noble return. In addition to larger elements to track and identify
sick and reasonably healthy individuals, researchers must assess detailed patient data to
determine the optimal strategy for the customisation of medicine. This will help to clinical
decision support (CDS) for physicians and other healthcare professionals can improved with
machine learning. This offers the resources to increase potential earnings. Within the field of
artificial intelligence, machine learning employs algorithms and data to identify patterns and
give medical practitioners automatic perception. AI "learns" features with the help of a vast
amount of healthcare data by utilizing sophisticated algorithms. After then, the new
perspective applied to enhance clinical practice (Jiang et al. 2017). It can also be given the
understanding and self-correction capabilities to improve its accuracy in response to input. In
order to notify medical professionals about the proper course of action, an AI application
provides them with current health information gleaned from professional processes, journals,
and newspapers. AI employs tools to find relationships that are too complicated to be
streamlined patient monitoring and administration. As a result, we must monitor current
developments in AI (Ahmed et al. 2020). This reviewed article covered the usage of artificial
intelligence in healthcare and related goods and applications in this review article.

Fraud
detection
Robot-
Training assisted
surgery
Artificial
inteligence
in healthcare
Cardiac
Medical
tissue
daigonosis
chips
Clinical
trail

Figure 1 shows the twelve main uses of AI in healthcare.

1.1 Artificial intelligence and AI-enabled devices Artificial neural networks (ANN)

These are similar to the human brain in that they function on nonlinear correlations between
input and output variables. (Park and Lek 2016). The introduction of the artificial neural
network in 1958 led to its widespread popularity because of the rise in data complexity and
capacity (Ravi et al. 2017). An input layer, a hidden layer, and an output layer make up an
ANN. Each node has a function inserted by the input layer to directly receive input data..
Next, each element in the input layer provides a balanced linear combination as input to the
nodes in the hidden layer, which use a non-linear transformation activation feature. The
output layer performs comparable duties.
AIMIS Open Lab and the Medical Image Cloud will support the creation of medical AI
applications and aid in the administration of medical data. Patients may see their CT, MRI,
and X-ray pictures as well as safely and easily share their medical information with others via
the Tencent AIMIS Medical Image Cloud. By sharing its medical AI capabilities with other
entities, including as academic institutions, engineering and research innovation centers, and
scientific research organizations, Tencent AIMIS Open Lab may be able to develop medical
AI apps. By use of cloud-based Picture Archiving and Communication Systems (PACS),
Tencent's 'AIMIS Image Cloud' links healthcare facilities at each phase of the Medical
Treatment Combination. In primary care facilities, this enables doctors to digitally receive
professional diagnoses and exams. Tencent provides its real-time interactive to physicians.

Figure 2. An illustration of the system at the system level and pertinent parts for combining
music therapy, ANN, and BAN.
The group uses the AI-driven enhanced workflow technologies of the An electronic health
platform to lessen the strain of routine, fundamental duties and improve diagnostic accuracy
while examining medical photos. Through flexibility and scalability, it ensures future-
readiness while Making it possible for users to quickly obtain answers for therapeutic,
operational, and strategic decision assistance (Siemens Healthcare 2021AI-RAD Companion
Chest CT is an AI-powered computed tomography device assistance. The AI-RAD
companion of picture datasets using algorithms driven by AI automates the post-processing.
To help radiologists focus on more pertinent concerns, large patient volumes and repetitive
activity routines can be automated.
Compared to hand comments, which take 30 minutes to calculate severity scores, our system
can do it in around 10 seconds every instance. These results can be used to monitor the
progression of abnormalities in COVID-19 patients and promptly assess the severity of the
lung infection (Gouda and Yasin 2020). These developments show how AI has progressed
from the day it was created to the present, showing how it has surpassed human intellect. In
the future, AI is expected to outperform humans in some tasks, which might be advantageous
to humans (Hosny et al. 2018). Through the use of AI-powered algorithms to produce
medical pictures, AI Rad companion helps radiologists decrease errors and effort while
improving diagnostic accuracy.

1.2 Robot-Assisted Surgery

v While examining the surgical site in three dimensions, a computer and software program
helps a surgeon efficiently operate surgical equipment connected to the robotic arms through
small incisions (Lanfranco et al. 2004). Under a laparoscopic technique, tiny (less than one
centimeter) abdominal incisions are made in order to insert short, thin tubes (called trocars),
which are then used to insert long, thin pieces of equipment. The surgeon manipulates, cuts,
and sutures tissue with these tools. Patients typically experience less pain during laparoscopic
procedures, and their recovery time is notably shorter. Growing robotic technology have
solved some of the laparoscopic technique's drawbacks, revolutionizing controlled access
surgery (Garry 2006).

Figure 3: shows a surgery assisted by robotic machine with hands.

Instead of taking the place of the surgeon or performing tasks on their own, robotic devices
can increase flexibility and ergonomic performance. The reason for calling them "master-
slave systems" is that the surgeon is the one operating them. The instruments are connected to
the patient's side via the slave unit, which is operated by robotic arms. Its components
include: (a) the master console, which has a user port that lets the operator view a three-
dimensional representation of the operating area; (b) manipulators for monitoring
instruments; and a monitor panel for focusing and adjusting the camera position. (Singh et al.
2018).
1.3 Virtual Nursing Assistant

Patients in hospitals undergoing treatment or surgery are supported by virtual nursing

assistants. The empathetic virtual personal healthcare aides from iCare Navigator, Sally and
Walt, empower patients to actively participate in their recovery and well-being.
Figure 4: illustrates a virtually assisted machine for nursing
A tablet or hospital TV set can be used to access them (Sadhika 2019). Research indicates
that patients do not experience criticism for their questioning when their healthcare assistant
is [Link] and Sally have the patience necessary to make sure that patients
comprehend the instructions and information about their discharge so they can take good
care of them after they are released. Patients engage truthfully and bravely as a result, and
their commitment is acknowledged (Barrett et al. 2019). The company that developed iCare
Navigator, TeleHealth Services, asserts that it uses a patient's electronic health records and
machine learning to build a personalized connection. Built on cutting-edge artificial
intelligence technology and cloud-based technologies, iCare Navigator keeps a close eye on
the patient's reaction, demeanor, openness, sensitivity, and overall commitment to providing
fully customized patient experiences (Raleigh 2017).
1.4 Clinical Trial Assistance

Because of regulatory ambiguity, risk aversion, and concerns about rapidly advancing
technologies (such as machine learning and wireless health monitoring tools and sensors),
clinical pharmaceutical research has basically ceased for the past 30 years. Documenting the
results of the majority of current clinical studies that show typical patient effects but are
difficult to translate into individualized medical recommendations at the standard point of
care is a significant barrier throughout the drug development phase (Shah et al. 2019).
Artificial intelligence seeks to transform clinical decision-making processes. Because it can
improve the protection and caliber of healthcare outcomes by utilizing the vast amounts of
genetic, biomarker, and phenotypic data gathered throughout the healthcare system, including
patient records and guidance systems (Magrabi et al. 2018).
Investigating the use of AI and ML are research institutes, biotechnology companies, and
development enterprises. Their three primary areas of concentration have been: (i) finding
compounds with medicinal properties and potential targets for drug development using
machine learning; (ii) diagnosing images (such as retinal scans, autopsy samples, and body
surfaces, bones, and essential organs) with the use of neural networks and optimization
techniques (iii) applying deep learning techniques on integrative data sources, such as the
fusion of genetic and therapeutic data, to uncover novel prediction techniques; (iv) using
dynamic algorithms for a numerical expansion of available therapeutic and diagnostic data
sets, to enable faster tracking and detection of disease progression (Shah et al. 2019).
Treatment decision-making is intended to be revolutionized by artificial intelligence. This is
because it could improve the safety and caliber of healthcare results by utilizing the enormous
volumes of phenotypic, genetic, and biomarker data that have been collected from patient
records and guidance systems, among other sources, throughout the healthcare system
(Magrabi et al. 2018).
Research institutes, biotechnology corporations, and development firms looked into three
primary areas for the use of AI and ML: using neural networks and optimization techniques
on diagnostic images (such as retinal scans, autopsy samples, and body surfaces, bones, and
essential organs); (ii) using machine learning to determine the therapeutic impacts of
molecular products and drug discovery targets iii) applying deep learning techniques on
integrative information sources, such as the combination of genetic and therapeutic data to
identify new prediction method; to enable faster detection and monitoring of disease
progression and dynamic algorithms for a numerical increase of current therapeutic and
diagnostic data sets (Shah et al. 2019).
1.5 Training
Logically organized training data about a relatively consistent process across time is
necessary for improving machine learning models; deviating from this leads to over-fitting,
when AI places an excessive amount of emphasis on erroneous connections in records (Buch
et al. 2018
Liability concerns and the "black box" dilemma are two of the most anticipated obstacles,
along with changing the way doctors practice medicine traditionally. A deep learning
algorithm created by Black Box engineers at Mount Sinai Hospital was evaluated using data
from 700,000 patients. An early sign of a disorder like schizophrenia could be accurately
predicted by this algorithm. This is even more amazing considering that this illness is
difficult to diagnose, even for experts. It is impossible to determine how the machine makes
this forecast or which variables are taken into account, which is the fundamental problem
with this method. The Black Box phenomenon is the term used to describe this idea
(Paranjape et al. 2019)
The processing of digitally recorded information for learning purposes is a new field that has
emerged as a result of recent developments in computing power and data collection. The
application of data science in the medical field has the potential to significantly accelerate the
rate at which information is generated and the range of issues that can be addressed, even
though costly prospective research and clinical trials have largely determined (Celi et al.
2016). Electronic health records (EHRs), which facilitate the faster digitization of healthcare,
encourage creative approaches to gathering and gaining access to relevant data that may be
utilized to make better decisions. In order to manage evidence, monitor AI software, and use
AI apps to Physicians must use their knowledge and skills to make well-informed decisions
(Paranjape et al. 2019).
1.6 Fraud Detection
Three categories exist for healthcare fraud: Healthcare fraud is committed by (a) individuals
(such as doctors, dentists) or provider organizations (such as hospitals); (b) businesses that
frequently engage in unethical tactics that target other service providers (such as laboratory
services) or suppliers of pharmaceutical and medical supplies in order to collect fees on
payout; (c) Fraud involving caregivers frequently involves specific groups, including as
insurers or patients. The three categories of projects that aim to prevent, detect, and address
Group activities can be utilized to combat fraud and abuse in the healthcare industry (Joudaki
et al. 2015).

In an effort to stop fraud, a small number of auditors handle thousands of healthcare reports
using traditional methods. The issue lies in their limited time for each claim, which causes
them to concentrate on particulars rather than the big picture of the provider's activities. As a
result, this method is ineffective and takes time the rising use of computer-based applications
and electronic medical data have made it possible to spot fraud and misbehavior in new ways.
Automated fraud detection solutions are becoming more popular thanks to advances in
machine learning and artificial intelligence (Bauder and Khosh goftaar 2018). Data mining is
a fundamental component of knowledge discovery from databases (KDD) and involves the
use of methods to evaluate data, create specialized models, and identify trends in the data that
were not before identified. Third-party payers, such as health insurance companies, can
classify a small fraction of instances or applicants for additional scrutiny and inspection of
fraud and abuse, and extract useful information from a large number of cases with the aid of
data mining (Joudaki et al. 2015).

The two main categories of data mining techniques utilized in healthcare fraud analysis are
supervised and unsupervised approaches. When there are no prior collections of reliable and
fraudulent findings, unsupervised procedures are used. In order to determine whether two
claims are similar to one another or different from one another, unsupervised techniques
typically compare the attributes of one claim to [Link] also makes it simpler to locate
anomalous or related entries inside classes by outlining the correlation and order criteria in
detail. A dataset of discovered fraud and legitimate cases is used to train a supervised
methods model that uses rating to classify observations as fraudulent or non-fraudulent.. They
require assurance of the authenticity of the documents. They assist in recognizing patterns of
fraud and wrongdoing that have already been noted. Thus, in order to reflect changes in
regulations and conditions as well as advancements in dishonest actions, the models will be
updated on a regular basis (Bolton and Hand 2002).
1.7 Drug Discovery and Other Research

Artificial neural networks that aid in the development of AI more quickly include deep neural
networks (DNN) and recurrent neural networks (RNN). Since deep learning algorithms
produced better outcomes in property prediction, artificial intelligence technologies have
drawn a lot of attention in the field of pharmaceutical scienceAI's use in early drug
development has significantly increased; examples include the de novo synthesis of peptides
and chemical compounds as well as the production of formulations (Hessler and Baringhaus
2018).

A Deep Neural Network (DNN) is composed of numerous stages of non-linear functions,

such as multiple hidden layers of neural networks. Understanding feature hierarchies, where
features at lower levels of the hierarchy are utilized to construct features at higher levels, is
the primary objective of deep learning approaches. levelsDeeper structures could produce
even better outcomes when each layer is retrained using an unsupervised learning technique
(Gudivada et al. 2016). The Tox21 challenge in 2014 was the scientific community's "largest"
endeavor to test computer algorithms for toxicity [Link] designed especially for
this purpose were used to screen 12,000 drugs and environmental contaminants for 12
different adverse effects. As part of the "Tox21 Data Challenge" (Tox21 challenge), the
efficacy of computer algorithms for preliminary toxicity evaluation has been investigated to
determine if they may reduce the amount of animal testing and in vitro research. Mayr and
colleagues (2016).
DNNs perform better than other machine learning methods in terms of predicting attributes
including biological behavior, ADMET (Absorption, Distribution, Metabolism, Excretion,
and Toxicitycharacteristics, as well as physicochemical factorsPhysicochemical descriptors
and extended connectivity fingerprinting (ECFP) have been combined with the existence or
lack of established toxicophores to improve the descriptor set. Molecular characteristics
allegedly associated with recognized toxicophoric components can be distinguished by the
DNN. These networks usually have a deeper comprehension of complex ideas in the
numerous underlying layers (Hessler and Baringhaus 2018).
Protease inhibitor molecules were then used to retrain the network, and at the end, the
network's degree of association with the target protein, chymotrypsin-like protease, was
assessed (Lee et al. 2014). Desikan 2020 discovered a startling resemblance between two of
the thirty-one possible compounds they discovered and aurantiamide, an antiviral component
that occurs naturally.
After employing protease inhibitor compounds for retraining, the network's connection to the
target protein, chymotrypsin-like protease, was assessed (Lee et al. 2014). Desikan 2020
discovered that aurantiamide, an antiviral component found naturally, was very similar to two
of the thirty-one candidate compounds they identified.

Artificial Intelligence in Medicine (AIM) has seen a substantial growth in specialization in

recent years due to the growing impact of AI on the medical field. This underscores the
necessity of conducting a systematic review of study findings and advancements in AIM.
When it comes to treating cancer patients, IBM Watson-Oncology has become as effective as
or more so than human specialists in medicine. Utilizing scientific data, Microsoft's Hanover
Project in Oregon has tailored cancer patients' therapy options. By analyzing smartphone
application data and diagnostic photographs acquired from NHS patients, the National Health
Service of the United Kingdom (NHS) employed Google's DeepMind tool to detect health
risks (Bali et al. 2019).

Such high-throughput approaches to biology and illness present challenges as well as

opportunities for the pharmaceutical sector to identify potential treatment approaches. The
pharmaceutical business is using machine learning (ML) technologies more frequently as a
result of recent advancementsThe goal of the eminent drug development strategy is to create
drugs (small molecules, antibodies, or peptides, or more recent approaches such short RNAs
or cell treatments) that enhance a molecular target's activity in order to treat a disease (Liu et
al. 2019).
Even though phenotypic testing has recently been more popular, starting a product
development strategy necessitates changing the target, which changes the disease condition.
Target recognition and prioritizing refers to the process of choosing the target based on the
data that is currently available.
The next step is target validation, which involves using in vivo and ex vivo physiologically
relevant methods to validate a discovered target for disease. For instance, machine learning
may assess large datasets knowing the function of a putative target to make inferences about
potential cause and effect based on known real targets. In order to determine the druggable
genes associated with morbidity, Costa et al. (2010) have created a tree-based meta-classifier
that focuses on the topological network of transcriptional, protein-protein, and metabolic
interactions, as well as tissue expression and subcellular localization. Using various sets of
genomic data, Jeon et al. (2014) created a support vector machine (SVM) classifier for
ovarian, pancreatic, and breast malignancies that can distinguish between proteins that are
therapeutic targets and those that are not.

1.8 AI Application for Stroke Management

Over 140,000 individuals die in the US each year from strokes, which are a common and
recurring illness. Therefore, research on the prevention and treatment of strokes is extremely
important. Recent years have seen an increase in the use of AI techniques in stroke-related
trials, particularly in the three critical areas of stroke care: early illness prediction, diagnosis,
and treatment (Astrakas et al. 2012).According to scientists in the UK and the USA, the AI
system will be able to predict heart attacks and strokes with accuracy. Kristopher Knott, a
research associate at the British Heart Foundation, conducted the largest-ever study
combining AI and cardiovascular magnetic resonance imaging (CMR) to date. By evaluating
how much of a single cardiac muscle contrast product is absorbed, cardiac magnetic
resonance imaging (CMR) analyzes the heart's blood circulation. The higher the blood flow,
the less likely it is that there are cardiac artery blockages. The reports require extensive and
time-consuming interpretation, which is frequently more qualitative than quantitative. They
developed an artificial intelligence system that examined photos and trained to recognize
signs of impaired blood flow in order to develop a more intelligent planWhen the device was
tested on scans of over 1,000 patients who desired CMR because they were either already
diagnosed with cardiac failure or at risk of developing it, researchers found that the AI model
performed well in identifying which people were more likely to have a heart attack or stroke.
It may be possible for machine learning algorithms to discern between an ischemic stroke and
a hemorrhagic or any other kind of stroke, which would lessen the possibility of overlooking
conditions including subdural hematoma, meningitis, encephalitis, coma, acute
demyelination, and abscess. The FDA approved the use of the Viz AI algorithm in 2018 to
help clinical decision-making through triage. In order to make a provisional diagnosis, AI
Contact can analyze CT scans and recognize stroke symptoms in visual films. The gadget can
identify stroke instances and send an alarm to a neurovascular specialist via laptop or
smartphone. This way, the specialist can attend to the most urgent cases, while the radiologist
can examine less urgent pictures. Patients who might not follow the routine examination
protocol might receive timely treatment from this AI-enabled system management without
risking their health or possibly death (Liebeskind 2018).
Rapid AI's data-driven, scientifically verified solutions enable doctors to diagnose stroke
patients more quickly and accurately, as well as determine the best course of treatment. Every
day, clinicians throughout the world improve patient safety and results by utilizing a
reputable and tested network that was created by stroke specialists and is currently in use in
over 1300 hospitals.
Another company focused on medical technology, Another company focused on medical
technology, NeuroView, wants to automate the prediction of stroke abnormalities in the
region. A shared objective of numerous well-known professional associations, healthcare
infrastructure companies, and even hospitals, who are increasingly collaborating to address
shared problems, is the development of departmental and user-friendly AI practices.
However, one should not undervalue the accomplishments of researchers over the past ten
years. According to Tran et al. (2019), it provides a generally optimistic outlook for the use
of AI in healthcare going forward for enterprises, patients, and healthcare practitioners.
1.9Cardiac Tissue Chips

In research on artificial organs and organs-on-a-chip, the heart has typically proven to be one
of the hardest organs to replicate. The University of Alabama in Birmingham research teams
have developed Cell culture techniques known as biomimetic cardiac tissue chips (CTCs) are
able to precisely reproduce complex blood flow pressures linked to variations in the pressure
volume of the heart. Through the identification of potentially harmful drugs before they enter
clinical trials, scientists anticipate that tissue chips will expedite the development of new
drugs and medical devices while also improving public health. In order to simulate changes
in the left ventricle's pressure-volume relationship, the cardiac tissue chip (CTC) prototype is
made up of entrapped cardiac cells that may be cultivated in three-dimensional (3-D) fibers
and subjected to hemodynamic stress. According to Rogers et al. (2019), It is possible to
generate heart cell cultures under diseased, normal, and developmental circumstances by
closely monitoring heart rate, peak-systolic pressure, end-diastolic pressure-volume, and the
ratio of systolic and diastolic duration, among other cardiac function-related metrics
 Figure 5: illustrates cardiac tissue cells
According to Kong et al. (2019), the CTC is capable of simulating the pressures associated
with a pressure and volume overload. Researchers Rogers et al. (2019) conducted
experiments using cardiomyogenic cell line-derived H9c2 cells. Their findings demonstrated
that, in the presence of pathological hemodynamic pressure, the culture inside the CTC
produces alterations in morphology and gene expression that are similar to those observed in
cases of hypertrophic and dilated cardiomyopathy.
When subjected to prolonged volume overload, the cells inside the CTC experience rapid
cardiac hypertrophy remodeling and fibrosis, while the cells susceptible to prolonged volume
overload experience significant cellular size alterations as a result of thinning and elongating
of the designed tissue. Based on these results, CTC can be used to create meaningful designs
that allow accurate reproduction of hemodynamic loading and unloading to mimic cardiac
conditions. On the other hand, systems such as Organ-on-chip and CTC generate vast
amounts of data pertaining to shape, size, structure, interaction, and composition, which
require analysisv

1.10 Artificial Neuron:

According to Oboudiyat et al. (2013), Alzheimer's disease is a neurodegenerative disorder

that causes progressive neuronal loss, which can have emotional, behavioral, and motor
consequences. It also results in the loss of an individual's soul. Patients and their family
experience significant distress from the condition. Still, scientists are exploring possible
nanotechnological avenues that could significantly enhance the patients' quality of life.
priority conditions as part of the mental health gap action program (mhGAP). The predicted
number of people with dementia worldwide was 35.6 million in 2010. This number is
expected to almost double every 20 years, reaching 65.7 million in 2030 and 115.4 million in
2050. Globally, there are around 7.7 million new cases of dementia detected annually, or one
case every four seconds (Tanna 2013). Recent research revealed an atypical build-up of
immunoglobulin (Ig) in the brain parenchyma of AD tissues, as well as unique neurons
indicating the degenerative and apoptotic nature of these vascular-derived
antibodiesSubsequent research revealed that the conventional complimentary elements, C1q
and C5b-9, that these Ig-positive neurons possessed were more frequently spatially linked
with Ig-negative neurons that had reactive microglia. Therefore, since there isn't a common
Blood-Brain Barrier (BBB) malfunction that would enable the detrimental effects of such
autoantibodies to reach their targets, the mere presence of anti-neuronal autoantibodies in
previously rejected serum cannot have any pathogenic implications (D'Andrea 2005).

Figure 6: shows analysis of artificial neurons with human neurons.

Artificial neurons that might be implanted into the brain to heal damage caused by
neurodegenerative disorders like Alzheimer's disease have been developed by an international
research team. The scientists created tiny silicon chips that mimic the operation of real
neurons by simulating biological ion channels (Abu-Hassan et al. 2019). The goal is to create
chips that can heal the essential processes of the nervous system and undo the damage caused
by autoimmune reactions. Actually, they resemble connecting bridges that cause a disruption
in a neural canal. The silicon chips are perfect for use as medical implants or in other
bioelectronic devices since they behave similarly to biological neurons and only require 140
nanoWatts of electricity, which is one-billionth of the power required by the microprocessors
used to construct artificial neurons. The next goal for researchers is to find the least invasive
and non-surgical ways to provide deep brain stimulation to individuals with Alzheimer's
disease. By doing so, it will be possible to support the use of artificial intelligence in
healthcare (University of Bath, 2019).

1.11 Plastic Surgery

Dedicated to staying at the forefront of contemporary scientific advancements, cosmetic
surgeons are imaginative individuals. Our capacity to swiftly and efficiently incorporate
advancements has allowed the field of plastic and reconstructive surgery to grow
tremendously, beginning with the invention of skin grafts through transplantation. Dedicated
to staying at the forefront of contemporary scientific advancements, cosmetic surgeons are
imaginative individuals. Our capacity to swiftly and efficiently incorporate advancements has
allowed the field of plastic and reconstructive surgery to grow tremendously, beginning with
the invention of skin grafts through transplantation.
Figure 7: shows plastic surgery assisted by Robotic machines
Applications for data-driven surgical modeling that may identify empirical asymmetries in
pre-operative photos can provide insight into the best course of action for achieving a desired
cosmetic outcome. Software for altering pictures that suggests such possibilities creates
irrational expectations and is unable to make up for the limitations of actual cosmetic
procedures. Artificial intelligence (AI) algorithms have been developed to classify patients
for whom specific therapies are too risky and to keep them out of the pre-operative setting in
terms of surgical screening.
According to Kim et al. (2019), both rely on obscure risk factors that were hidden in prior
surgical error statistics and have a major impact on both health and the economy. The
aesthetic surgery group has recently demonstrated its ability to integrate large amounts of
data through the establishment of databases such as ASAPS CLOUD, Cosmet Assure, the
General Register of Autologous Fat Transfer (GRAFT), the Regional Surgical Quality
Improvement Plan (NSQIP), and Plastic Surgery Monitoring Operations and Results (TOPS).
This indicates both a readiness to embrace the disruptive nature of artificial intelligence and
the need for methodical data processing from plastic surgery procedures. Aesthetic Neural
Network (ANN) is the latest technique that the American Society for Cosmetic Plastic
Surgery (ASAPS) developed. It is a preliminary approach to improving treatments and
modeling economics (Chandawarkar et al. 2020).

Health decisions about wound treatment are based on an assessment of wound features, such
as dimensions and location, as well as patient-related parameters, such as skin texture,
genetic details, and living environment. The seriousness of certain casualties meets the
human eye clearly which signifies the emergence of artificial intelligence to render
assessment easy and more effective (Yeong et al. 2005). By matching wound pictures against
explicit measures of the patient's body, a thinking machine may estimate the magnitude of
infected/damaged tissues (Kim et al. 2019).
In addition to other smart imaging techniques, angiograms from computed tomography (CT)
based on AI-assisted analysis may help surgeons develop surgical flaps. The new equipment
allows doctors to view three-dimensional (3D) photographs as strips, thereby saving lapse of
time and strengthening their capacity to see the internal structure of the human body. A
computer can assess every slice of a 3D imaging specimen simultaneously, but radiologists
need to slice CT angiography images into 64 pieces for accurate diagnosis. Cosmetic
surgeons frequently select the same flap configuration when performing a different flap
operation. Still, surgeons may find it easier to create a personalized plan for each patient with
the help of AI's quick, integrated thinking abilities improving reliability in their ability to see
the human body's inner framework. While radiologists must divide CT angiography images
into 64 slices for precise diagnosis, a computer can evaluate all slices of a 3D imaging
specimen at the same time. While conducting a separate flap procedure, cosmetic surgeons
very often choose the same configuration of flaps. Nevertheless, AI‟s fast, integrative
thought capabilities may assist surgeons in developing a strategy that can be customized to
each particular patient

Osteocytes may become infected with cutaneous wound infections, leading to progressive
swelling and osteomyelitis. Because it takes a while for an osseous lesion to show up on basic
radiographic imaging, osteomyelitis presents a number of challenges during the diagnosis
stage. AI-assisted radiographic image evaluation could shorten the time it takes for
osteomyelitis to show up on a radiograph. Similar to this, radiologists have strengthened the
differences in intervals between two radiological pictures by using temporal subtraction, a
function of computer-assisted diagnosis (Kim et al. 2019).

Craniosynostosis is one of the most severe craniofacial malformations that cosmetic surgeons
have encountered (Johnson and Wilkie 2011). Contributions to the epidemiology of
craniosynostosis come from both hereditary and environmental sources. The main
disadvantage appears to be the paucity of additional data regarding the outcomes of patients
of craniosynostosis. Scientists and medical professionals can now identify the genetic and
environmental origins of this anomaly.. These days, surgical preparation is supported and
improved by combining many images using artificial intelligence (AI) technologies. This
appears to be particularly true in syndromic disorders, where recurrent abnormalities may be
more frequent as a result of the pathophysiology of osteogenesis. Artificial neural networks
(ANNs) have made it possible for cosmetic surgeons to predict postoperative problems after
craniofacial surgery. Repeated cranial deficiencies may arise from syndromic
craniosynostosis, and following corrective surgery, the bone may start to form improperly. AI
and precision medicine can often be utilized to enhance surgical modifications in order to
optimize postoperative results. Moreover, cranial remodeling can be configured with the use
of Big Data picture processing to better suit individual children (Kim et al. 2019).
1.12 Organ Transplantation

Around 100,000 organs are transplanted every day worldwide, and there are still more people
waiting for organs (Healio 2019). Organ Secure is an AI-powered program that aims to match
organ donors with people in need of an organ transplant in real time in order to address this
dilemma.. Organ Secure is currently under alpha testing with the goal of resolving the two
primary issues facing the organ donation ecosystem: encouraging more people to donate their
organs and ensuring that those in need may obtain organs when they need them. People can
sign up to become patients, learn which organs to donate based on their medical data, and
learn about local legislation all through this application, which begins by educating users
about organ donation (Pradhan et al. 2020). An AI-powered real-time rating system on the
donor list, together with the time required to rise to the top of the list, would be beneficial to
organ recipients.

Figure 8: shows artificially assisted organ transplant.

Now that patients and their families have access to information about expected costs, local
organ banks, and other pertinent details, they may appropriately prepare for the procedure.
The program uses Azure Machine Learning to identify an organ match and estimate the rank
and time required for an expecting receiver, taking into account the several criteria governing
organ donation, such as blood type and antigen form (Pahl et al. 2020). There is no practical
way to search the records in the event that a possible donor is hurt or dies, making it difficult
to effectively preserve the organs. Before starting the extraction procedure, hospitals can use
Organ Secure to verify the donor's identity.
Conflict of Interest

The authors have no conflict of interest to declare.

REFERENCES

The Role of AI in Hospitals and Clinics: Transforming Healthcare in the 21st Century
by Shiva Maleki Varnosfaderani and Mohamad Forouzanfar / Published: 29 March 2024

Binhowemel, S.; Alfakhri, M.; AlReshaid, K.; Alyani, A. Role of Artificial Intelligence in
Diabetes Research Diagnosis and Prognosis: A Narrative Review. J. Health Inform. Dev.
Ctries. 2023, 17, 1–12. [Google Scholar]

Guan, Z.; Li, H.; Liu, R.; Cai, C.; Liu, Y.; Li, J.; Wang, X.; Huang, S.; Wu, L.; Liu, D.
Artificial intelligence in diabetes management: Advancements, opportunities, and
challenges. Cell Rep. Med. 2023, 4, 101213. [Google Scholar] [CrossRef] [PubMed]

Hu, Y.; Kothapalli, S.V.; Gan, W.; Sukstanskii, A.L.; Wu, G.F.; Goyal, M.; Yablonskiy,
D.A.; Kamilov, U.S. DiffGEPCI: 3D MRI Synthesis from mGRE Signals using 2.5 D
Diffusion Model. arXiv 2023, arXiv:2311.18073. [Google Scholar]

Artificial Neural network (ANN) Enabled Internet of Things (IoT) Architecture for Music
Therapyby Shama Siddiqui,Rory Nesbitt,Muhammad Zeeshan Shakir, Anwar Ahmed
Khan, Ausaf Ahmed Khan, Karima Karam Khan,Naeem Ramzan.

Cardiac Tissue Chips to Fly to the ISS to Refine Drug Screening

Abu-Hassan, K., Taylor, J. D., Morris, P. G., et al. (2019). Optimal solid-state neurons.
Nature communications, 10(1), 5309.

Ahmed, Z., Mohamed, K., Zeeshan, S., & Dong, X. (2020). Artificial intelligence with multi-
functional machine learning platform development for better healthcare and precision
medicine. Database: the journal of biological databases and curation, 2020:baaa010. doi:
10.1093/database/baaa010.

Almada, M., & Maranhão, J. (2021). Voice-based diagnosis of covid-19: ethical and legal
challenges. International Data Privacy Law, ipab004. DOI:10.1093/IDPL/IPAB004.

Bali, J., Garg, R., & Bali, R. T. (2019). Artificial intelligence (AI) in healthcare and
biomedical research: Why a strong computational/AI bioethics framework is required?.
Indian Journal of Ophthalmology, 67(1), 3–6.
Barrett, M., Boyne, J., Brandts, J., Brunner-La Rocca, H.P., et al. (2019). Artificial
 intelligence supported patient self-care in chronic heart failure: a paradigm shift from
reactive to predictive, preventive and personalised care. The EPMA Journal, 10(4), 445-
464.

Cell Press. (2020) Researchers restore injured man‟s sense of touch using brain-computer
interface technology. ScienceDaily. Retrieved from
[Link] [Link]

Chandawarkar, A., Chartier, C., Kanevsky, J., & Cress, P. E. (2020). A Practical Approach to
Artificial Intelligence in Plastic Surgery. Aesthetic surgery journal. Open forum, 2(1),
ojaa001.

Ciuti, G., Skonieczna-Żydecka, K., Marlicz, W., et al. (2020). Frontiers of Robotic
Colonoscopy: A Comprehensive Review of Robotic Colonoscopes and Technologies.
Journal of clinical medicine, 9(6), 1648.

Corredor, G., Wang, X., Zhou, Y., et al. (2019). Spatial Architecture and Arrangement of
Tumor-Infiltrating Lymphocytes for Predicting Likelihood of Recurrence in Early-Stage
Non-Small Cell Lung Cancer. Clinical cancer research : an official journal of the
American Association for Cancer Research, 25(5), 1526–1534.

cury, R. C., Megyeri, I., Lindsey, T., et al. (2021). Natural Language Processing and Machine
Learning for Detection of Respiratory Illness by Chest CT Imaging and Tracking of
COVID- 19 Pandemic in the US. Radiology. Cardiothoracic imaging, 3(1), e200596.

Davenport, T., & Kalakota, R. (2019). The potential for artificial intelligence in healthcare.
Future Healthcare Journal, 6(2), 94–98.

Desikan, S. (2020). Coronavirus: TCS uses AI for drug discovery. The Hindu. Retrieved
from [Link] content/uploads/2020/04/april-5-articles-
science-and-technology- [Link]

Drori, I., Thaker, D. D., Srivatsa, A., et al. (2019). Accurate Protein Structure Prediction by
Embeddings and Deep Learning Representations. Retrieved from
[Link]

Ganzer, P. D., Colachis, S. C., 4th Schwemmer, M. A. et al. (2020). Restoring the Sense of
Touch Using a Sensorimotor Demultiplexing Neural Interface. Cell, 181(4), 763-773.

Garry R. (2006). Laparoscopic surgery. Best practice & research. Clinical obstetrics &
gynaecology, 20(1), 89–104.

Ghose, A., Roy, S., Vasdev, N., Olsburgh, J., & Dasgupta, P. (2020). The Emerging Role of
Artificial Intelligence in the Fight Against COVID-19. European Urology, 78(6), 775–
 776.

Gouda, W., & Yasin, R. (2020). COVID-19 disease: CT Pneumonia Analysis prototype by
using artificial intelligence, predicting the disease severity. The Egyptian Journal of
Radiology and Nuclear Medicine, 51(1), 196.

He, J., Baxter, S. L., Xu, J., et al. (2019). The practical implementation of artificial
intelligence technologies in medicine. Nature Medicine, 25(1), 30–36.

Healio. (2019). Transplant tourism: A „pervasive‟ and dangerous

„shadow world of medicine‟, Infectious disease news. Retrieved from
[Link] transplant-tourism-a-
pervasive-and-dangerous-shadow-world-of- medicine

Hithaishi, C. B. (2020). Tencent launches AI-powered medical imaging and data

management devices. Biospectrum. Retrieved from
[Link]

Jiang, X., Coffee, M., Bari, A., et al. (2020). Towards an artificial intelligence framework for
data-driven prediction of coronavirus clinical severity. Computers, Materials & Continua
63: 537–551.

Kim, Y. J., Kelley, B. P., Nasser, J. S., & Chung, K. C. (2019). Implementing Precision
Medicine and Artificial Intelligence in Plastic Surgery: Concepts and Future Prospects.
Plastic and Reconstructive Surgery. Global open, 7(3), e2113.

Kong, M., Lee, J., Yazdi, I. K., et al. (2019). Cardiac Fibrotic Remodeling on a Chip with
Dynamic Mechanical Stimulation. Advanced Healthcare Materials, 8(3), e1801146.

Lee, S. I., Celik, S., Logsdon, B. A., et al. (2018). A machine learning approach to integrate
big data for precision medicine in acute myeloid leukemia. Nature Communications, 9(1),
42.

Li, L., Qin, L., Xu, Z., et al. (2020). Using Artificial Intelligence to Detect COVID-19 and
Community-acquired Pneumonia Based on Pulmonary CT: Evaluation of the
Diagnostic Accuracy. Radiology, 296(2), E65–E71.

Liebeskind D. S. (2018). Artificial intelligence in stroke care: Deep learning or superficial

insight?. EBioMedicine, 35, 14–15.

Linder, S. M., Rosenfeldt, A.B., Bay, R.C., et al. (2015). Improving Quality of Life and
Depression After Stroke Through Telerehabilitation. The American journal of
occupational therapy, 69(2), 6902290020p1–6902290020p10.
Liu, B., He, H., Luo, H., Zhang, T., & Jiang, J. (2019). Artificial intelligence and big data
facilitated targeted drug discovery. Stroke and Vascular Neurology, 4(4), 206–213.

Magrabi, F., Ammenwerth, E., McNair, J. B., et al. (2019). Artificial Intelligence in Clinical
Decision Support: Challenges for Evaluating AI and Practical Implications. Yearbook of
Medical Informatics, 28(1), 128–134.

Mashamba-Thompson, T. P., & Crayton, E. D. (2020). Blockchain and Artificial Intelligence

Technology for Novel Coronavirus Disease-19 Self-Testing. Diagnostics (Basel,
Switzerland), 10(4), 198.

Mayr, A., Klambauer, G., Unterthiner, T., & Hochreiter, S. (2016). DeepTox: Toxicity
Prediction using Deep Learning. Frontiers in Environmental Science, 3, 80.

McCall, B. (2020). COVID-19 and artificial intelligence: protecting health-care workers and
curbing the spread. Lancet Digit Health, 2, 166-167.

Mehta, V., Rammohan, R., Sanivarapu, R., Sayedy, N., & Akella,
J. (2020). The role of virtual assistants with coronavirus diagnosis and treatment. Chest,
158(4), A1176.

Mencattini, A., Mattei, F., Schiavoni, G., et al. (2019). From Petri Dishes to Organ on Chip
Platform: The Increasing Importance of Machine Learning and Image Analysis. Frontiers
in pharmacology, 10, 100.

Mohlman, J. S., Leventhal, S. D., Hansen, T., et al. (2020). Improving Augmented Human
Intelligence to Distinguish Burkitt Lymphoma From Diffuse Large B-Cell
Lymphoma Cases. American Journal of Clinical Pathology, 153(6), 743–759.
Mori, Y. (2019). Artificial intelligence and colonoscopy: the time is ripe to begin clinical
trials. Endoscopy, 51, 219-220.

Murff, H. J., FitzHenry, F., Matheny, M. E., et al. (2011). Automated identification of
postoperative complications within an electronic medical record using natural
language processing. JAMA, 306(8), 848–855.

Narula, S., Shameer, K., Salem Omar, A. M., Dudley, J. T., & Sengupta, P. P. (2016).
Machine-Learning Algorithms to Automate Morphological and Functional Assessments in
2D Echocardiography. Journal of the American College of Cardiology, 68(21), 2287–
2295.

Oboudiyat, C., Glazer, H., Seifan, A., Greer, C., & Isaacson, R. S. (2013). Alzheimer's
disease. Seminars in Neurology, 33(4), 313–
329.
Pahl, E., Emery, R. W., Noce, M., et al. (2020). Mobile Application for Communication
Increases the Efficiency of Organ Procurement and Transplantation. Progress in
transplantation (Aliso Viejo, Calif.), 30(2), 172–176.

Pandey, S. K., & Sharma, V. (2019). Robotics and ophthalmology: Are we there yet?. Indian
Journal of Ophthalmology, 67(7), 988– 994.

Paranjape, K., Schinkel, M., Nannan Panday, R., Car, J., & Nanayakkara, P. (2019).
Introducing Artificial Intelligence Training in Medical Education. JMIR Medical
Education, 5(2), e16048.

Park, Y. S., & Lek, S. (2016). Artificial Neural Networks: Multilayer Perceptron for
Ecological Modeling. In Developments in Environmental Modelling, 28, 123-140.

Pavlou, M., Ambler, G., Seaman, S. R., et al. (2015). How to develop a more accurate risk
prediction model when there are few events. BMJ (Clinical research ed.), 351, h3868.

Pradhan, P., Mishra, A. S., & Habeebullah, H. (2020). Organ donation-an android
application. SSRN Electronic Journal, 7(12), 85-89.

Raleigh. (2017).TeleHealth Services Redefines Patient Engagement,

Launches iCare Navigator with AI-Driven Virtual Health Coach. TeleHealth Press
Release. Retrieved from
[Link]
Press_Release_iCN_10-[Link]

Ravi, D., Wong, C., Deligianni, F., et al. (2017) Deep Learning for Health Informatics. IEEE
Journal of Biomedical and Health Informatics, 21(1), 4-21.
ennie, S., Buchbinder, M., Juengst, E., et al. (2020). Scraping the Web for Public Health
Gains: Ethical Considerations from a 'Big Data' Research Project on HIV and
Incarceration. Public Health Ethics, 13(1), 111–121.

Rogers, A. J., Miller, J. M., Kannappan, R., & Sethu, P. (2019). Cardiac Tissue Chips (CTCs)
for Modeling Cardiovascular Disease. IEEE Transactions on Bio-medical Engineering,
66(12), 3436–3443.

Rouillard, A. D., Hurle, M. R., & Agarwal, P. (2018). Systematic interrogation of diverse
Omic data reveals interpretable, robust, and generalizable transcriptomic features of
clinically successful therapeutic targets. PLoS Computational Biology, 14(5), e1006142.

Sadhika. (2019). Fintalk. Bank of Baroda. Retrieved from

[Link] [Link]

Saltz, J., Gupta, R., Hou, L., et al. (2018). Spatial Organization and Molecular Correlation of
 Tumor-Infiltrating Lymphocytes Using Deep Learning on Pathology Images. Cell Reports,
23(1), 181– 193.

Santosh, K. C. (2020). AI-Driven Tools for Coronavirus Outbreak: Need of Active Learning
and Cross-Population Train/Test Models on Multitudinal/Multimodal Data. Journal of
Medical Systems, 44(5), 93.

Shah, P., Kendall, F., Khozin, S., et al. (2019). Artificial intelligence and machine learning in
clinical development: a translational perspective. NPJ Digital Medicine, 2, 69.

Shen, J., Zhang, C., Jiang, B., et al. (2019) Artificial Intelligence Versus Clinicians in Disease
Diagnosis: Systematic Review. JMIR Medical Informatics, 7(3), e10010.

Siemens Healthcare GmbH. (2021). Siemens Healthineers Digital Marketplace Online Help –
Online Guide. Retrieved from
[Link]
gital_marketplace_user_guide.pdf

Singh, S., Dalla, V. K., & Shrivastava, A. (2021). Combating COVID-19: Study of robotic
solutions for COVID-19, AIP Conference Proceedgs. Retreived from
[Link] doi/pdf/10.1063/5.0050148
Spasic, I., & Nenadic, G. (2020). Clinical Text Data in Machine Learning: Systematic
Review. JMIR Medical Informatics, 8(3), e17984

The Hindu. (2020). Mumbai students develop AI-based voice tool to detect COVID-
19. Sci-tech. Retrieved from
[Link] develop-ai-based-voice-
tool-to-detect-covid-19/[Link]

Tran, B. X., Latkin, C. A., Vu, G. T., et al. (2019). The Current Research Landscape of the
Application of Artificial Intelligence in Managing Cerebrovascular and Heart Diseases: A
Bibliometric and Content Analysis. International Journal of Environmental Research and
Public Health, 16 (15), 2699.

University of Bath. (2019). World first as artificial neurons developed to cure chronic
diseases. Press release. Retrieved from [Link]
first-as-artificial- neurons-developed-to-cure-chronic-diseases/

Urias, M. G., Patel, N., He, C., et al. (2019). Artificial intelligence, robotics and eye surgery:
are we overfitted?. International Journal of Retina and Vitreous, 5, 52.

Vamathevan, J., Clark, D., Czodrowski, P., et al. (2019). Applications of machine learning in
drug discovery and development. Nature reviews. Drug discovery, 18(6), 463-477.
NYU Langone Health / NYU School of Medicine. (2020). Experimental AI tool predicts
which COVID-19 patients develop respiratory disease. ScienceDaily. Retrieved from
[Link]/releases/2020/03/[Link]