Paper
Paper
Paper
Abstract:
Artificial Intelligence (AI) and Machine Learning (ML) has completely transformed
the Life sciences. AI has reshaped Research and Development by Data science and
Machine Learning to a enormous amount of data which has resulted in discovery of
New molecule. AI can access cross-references published scientific literature to
discover new promising drugs that has the capabilities to cure the patients. Applying
AI to big data in life sciences can help companies to reshape their business models,
streamline bio-pharma manufacturing, and enhance everything from new molecule
research and clinical trial data flow, to self-healing supply chain applications and
product intelligence. This paper tries to study the various application and challenges
of AI/ML in Healthcare industry and the role played by AI/ML to transform the whole
Healthcare industry across the globe by reviewing the existing work. Further, we
discussed the limitations of current works and identify the directions for future works.
Keywords: AI, ML, clinical Trials, Drug Development, Healthcare
1. Introduction
Artificial intelligence (AI) is presently one of the scorching lingoes in technology. The last
few decade have seen many innovations and advancements that have previously perceived
as a fiction of science.
Artificial intelligence is the simulation of human intelligence processes by machines,
especially computer systems. Specific applications of AI include expert systems, natural
language processing, speech recognition, and machine vision. AI programming focuses on
cognitive skills that include the following:
Learning. This aspect of AI programming focuses on acquiring data and creating rules
for how to turn it into actionable information. The rules, which are called algorithms,
provide computing devices with step-by-step instructions for how to complete a specific
task.
Reasoning. This aspect of AI programming focuses on choosing the right algorithm to
reach a desired outcome.
Self-correction. This aspect of AI programming is designed to continually fine-tune
algorithms and ensure they provide the most accurate results possible.
Creativity. This aspect of AI uses neural networks, rules-based systems, statistical
methods and other AI techniques to generate new images, new text, new music and new
ideas.
2. Modus operandi of ML
UC Berkeley breaks out the learning system of a machine learning algorithm into three main
parts.
i. A Decision Process: In general, machine learning algorithms are used to make a
prediction or classification. Based on some input data, which can be labeled or unlabeled,
your algorithm will produce an estimate about a pattern in the data.
ii. An Error Function: An error function evaluates the prediction of the model. If there are
known examples, an error function can make a comparison to assess the accuracy of the
model.
iii. A Model Optimization Process: If the model can fit better to the data points in the
training set, then weights are adjusted to reduce the discrepancy between the known
example and the model estimate. The algorithm will repeat this iterative “evaluate and
optimize” process, updating weights autonomously until a threshold of accuracy has been
met.
3. Common ML algorithms
A number of machine learning algorithms are commonly used. These include:
Neural networks: Neural networks simulate the way the human brain works, with a
huge number of linked processing nodes. Neural networks are good at recognizing
patterns and play an important role in applications including natural language translation,
image recognition, speech recognition, and image creation.
Linear regression: This algorithm is used to predict numerical values, based on a linear
relationship between different values. For example, the technique could be used to predict
house prices based on historical data for the area.
Logistic regression: This supervised learning algorithm makes predictions for
categorical response variables, such as “yes/no” answers to questions. It can be used for
applications such as classifying spam and quality control on a production line.
Clustering: Using unsupervised learning, clustering algorithms can identify patterns in
data so that it can be grouped. Computers can help data scientists by identifying
differences between data items that humans have overlooked.
Decision trees: Decision trees can be used for both predicting numerical values
(regression) and classifying data into categories. Decision trees use a branching sequence
of linked decisions that can be represented with a tree diagram. One of the advantages of
decision trees is that they are easy to validate and audit, unlike the black box of the neural
network.
Random forests: In a random forest, the machine learning algorithm predicts a value or
category by combining the results from a number of decision trees.
4. Objectives of the Study
i. To study the impact of AI and ML in revolutionizing Healthcare industry.
ii. To study the challenges in implementing the AI and ML in Healthcare sector.
5. Literature Review
iii. Abhishek Parashar et al, 2022 1: the authors concluded from their study that with the
help of AI/ML application they were able to use different machine learning algorithms to
predict different diseases with satisfying accuracy. Also the application involved huge
use of personal medical data and in order to protect that huge data from unauthorized
access they included a login facility to maintain privacy.
iv. Dora Gondocs and Viktor Dorfler, 2019 2: They concluded that the current processes
are not designed to incorporate AI; hence, need new processes. The future of AI design is
collaborative; it will involve medics collaborating with AI developers, an essential part of
which will be looping back the experience of using AI into the process design.
v. Mahmoud nasr et al, 20213: This paper studied the wearable devices and smart phones
for basic signs monitoring, machine learning for three significant diseases such as
(COVID-19, heart disease, and diabetes) diagnosis, and the frameworks developed to aid
the adults in ambient assisted living. They concluded that almost all the wearable and
smart phones were able to monitor the vital essentials. They also discussed the major
challenges of recently developed smart healthcare frameworks that are the main obstacles
to develop assistive prototypes discussed the major challenges of recently developed
smart healthcare frameworks that are the main obstacles to develop assistive prototypes.
vi. Binggui Zhou et al, 20224: this paper studied the impact of NLP in healthcare industry by
reviewing the existing literature. The authors had elaborated on various NLP approaches
for smart healthcare. This study concluded that the NLP driven smart healthcare played a
crucial role in understanding two important medical issues namely Corona virus and
mental health.
vii. Rashika Raina1 and Rakesh Kumar Jha, 20225: The authors of this research have
proposed a novel architecture for an Intelligent and Interactive Healthcare System that
incorporates edge/fog/cloud computing techniques and focuses on Speech Recognition
and its extensive application in an interactive system for remote areas. This paper
explains why speech interaction method is taken as the core of proposed framework.
Various algorithms available and applied in speech recognition are also discussed. Also,
several energy and power efficient techniques along with few resource allocation
techniques are discussed. The fundamental reason for using speech in the healthcare
sector is that it is easily available and can easily predict any physical or psychological
discomfort. Simply put, human speech is the most natural form of communication.
viii. Suryakanta Sahu, 20226: this study, a contextual Chabot empowered by deep
learning algorithm and natural language processing techniques was developed and the
model’s performance was evaluated against a Naive Bayes model and numerous state-of-
the-art solutions. For training the conversational system, data was gathered from a
number of open-source websites like Wikipedia, WebMD etc. The suggested model is
anticipated to assist users with healthcare related enquiries instantly and without the need
for an active internet connection. Upon evaluating the model with a number of accuracy
metrics, a train accuracy of 98% and validation accuracy of 81% was achieved.
ix. Nayif Saleh, 20177: This study came up with a solution faced by the healthcare industry in
deploying wireless sensor nodes for real world application. The use of this technology
requires huge energy and to resolve this energy issue the authors have proposed a new
technique to resolve the problems due to limited energy sources. They proposed use of a
quaternary transceiver, instead of a binary one, which will use the amplitude/phase,
modulator/demodulator units to increase the number of bits transmitted per symbol. The
system will reduce the consumption of energy in the transmission phase due to the
increased bits transmitted per symbol.
6. Application of AI/ML in Healthcare industry:
i. Identification of diseases: AI-powered diagnostics refer to the use of advanced
computer algorithms and machine learning techniques to analyze medical data and assist
in disease diagnosis. This technology leverages large datasets and complex algorithms to
identify patterns and correlations that may go undetected by traditional diagnostic
methods. The process starts with inputting medical data into an AI system, including
electronic health records, lab results, imaging scans, genetic information, and even
lifestyle factors. The more data available, the more accurate the AI algorithm can
become. The system then uses this data to identify patterns or trends that may indicate a
particular disease or condition.
ii. Predicting future health outcomes: AI also has the potential to predict the likelihood of
developing certain conditions based on factors such as genetics, lifestyle choices, and
environmental factors. By analyzing large datasets and identifying risk factors associated
with different diseases, AI systems can provide personalized risk assessments for
individuals
iii. Improved Accuracy and Speed: One of the major advantages of using AI-powered
diagnostics for early disease detection and prevention is the significant improvement in accuracy
and speed compared to traditional methods. By combining advanced algorithms, machine
learning, and vast amounts of data, AI systems are able to analyze medical images and patient
data with unprecedented precision and speed.
iv. Clinical Trials: the application of AI/ML in clinical trials involves various processes
a. Screening of Patients: with the help of AI/ML the selection of right candidates
can be done efficiently and effectively for clinical trials. It speeds up the
screening and selection process of right patients for the clinical trials and also
makes sure that diversified pool of patients is selected.
b. Repurposing of Drugs: AI/ML can analyze the existing chemical composition
of drugs to identify potential new uses of existing drugs and accelerates the drug
discovery process.
c. Real time Monitoring: Ai/ML can analyze real time data of patients who have
participated in clinical trials. It helps to identify adverse events faster and with
accuracy than done through manual methods.
d. Predictive Analysis: AI/ML will predict the responses of patients with respect to
the different treatments and thus helps researches regarding dosing and treatment
strategies.
e. Enhances efficacy: AI/ML models used in clinical trial databases will help in
predicting trial success in early stages, hence minimizing unrecoverable costs.
v. Automation: automation in pharmaceutical manufacturing industry has multiple
benefits. The benefits that automation provides are Safety, efficiency, reliability and cost.
Most importantly enhanced safety in formulation, labeling, and packaging addresses the
crucial concerns of human error.
Automation handling under controlled parameters can lead to improved reliability and
consistency, resulting in enhanced quality assurance.
vi. Blocking Counterfeit drugs: AI/ML algorithms can analyze huge data to identify
patterns associated with counterfeit drugs, helping not only authorities but also
companies to detect and prevent their distribution.
vii. Quality Control: AI tools like Trials.AI employ Natural Language Processing (NLP)
and knowledge graphs to scrutinize clinical trial protocols for potential risks or
compliance issues. This oversight enhances the quality of new drug rollouts and helps
mitigate risks.
viii. Pharmaceutical Marketing: One of the most important applications of AI in
healthcare marketing is through the use of personalized marketing. By analyzing
humongous amounts of data on consumer behavior, AI can provide precise insights into
individual preferences and develop targeted marketing campaigns to increase engagement
and loyalty.
ix. Customer engagement with Chatbots: Chatbots are intelligent virtual assistant and are
capable of understanding local languages, providing instant responses and
recommendations according the needs of the patients. AI and ML will become the go-to
solution for healthcare providers looking to enhance support across multiple channels.
x. AI/ ML in Surgical Training: The use of AI in surgical learning has the potential to
transform the way in which surgeons are trained. Surgical training has undergone
significant changes in recent years through the introduction of simulation and task-based
training. AI complements this path with great promise
7. Challenges of AI/ML in Healthcare Industry.
x. Quality of Data: the data related to health care industry is often incomplete,
isolated and stored in different formats. Thus it makes it harder for the analyst
to access and analyze the data effectively. Poor quality of data can lead to
incorrect diagnosis and treatment.
xi. Regulatory compliance: the regulatory compliance ensures that the AI/ML
system adheres with existing regulatory requirements so that a transparent and
trustworthy system is built at a relatively lower cost. The objective of
regulatory compliance is to protect the identity of patients, including personal
information of patients and personal health information
xii. Privacy and security concerns: Doctors are not allowed to share any medical
information about their patients with any individual or entity without
obtaining the exclusive consent of the patients. The healthcare data is sensitive
and subject to strict privacy regulations. The Government introduced The
Digital Information Security in Healthcare Act (DISHA), India’s HIPAA
equivalent, aimed at providing healthcare data privacy, security,
confidentiality and standardization.
xiii. Recruiting data science experts: recruiting the experts for supervising the
AI/ML in healthcare is a big challenge. There is a large gap between demand
and supply for data science and AI experts and the healthcare industry.
xiv. Ethical considerations: this is one of the biggest challenges that AI/ML
experts, healthcare experts and pharmaceutical firms have to face related to
patient privacy, healthcare information, efficacy of drugs and many more. To
address these ethical challenges a deep understanding of data science
principles, healthcare ethics and regulatory compliance is required.
xv. Interoperability: the existing healthcare systems all over the world generally
use different standards and formats for storing the data, making
interoperability a serious challenge. The ideal integration of AI applications
with existing healthcare IT infrastructure is crucial for effective and efficient
communication and data exchange within the system.
xvi. Need of standardization: The current healthcare system lacks the presence of
well established guidelines for development and instills of AI which has lead
to inefficiency and inconsistencies in practices.
8. Limitations of the study
Following are the limitations of the study:
i. The data used by AI and ML are not cent percent accurate. So to outcome of
the AI and ML technology will suffer from certain biases.
ii. The AI and ML technology have the potential to change the whole healthcare
industry landscape but in order to do that it requires people with ample
knowledge of technology which is the biggest challenges nowadays as this
technology is in its early stages in the healthcare industry.
9. Conclusion
The use of AI/ML is increasing by every passing day. Using AI and ML in healthcare
industry has immense potential to revolutionize various aspects of healthcare delivery,
diagnosis, treatment, and patient outcomes. These technologies can analyze huge amount
a data which will lead to more accurate diagnosis, customized treatment and effective and
improved patient care. The AI and ML have the potential to revolutionize the healthcare
industry. However, having said that it is crucial to address challenges such a privacy of
patients, regulatory compliance, and ethical considerations to ensure that the adaptation
of AI and ML is done effectively and transparently.
References
1. Abhishek Parashar, Yukti Mohan and Sayoni Ghosh, “AI-DOC: a Mini Healthcare Assistant
for the DigitalWorld”, January 2022. https://easychair.org/publications/preprint_open/5WwW
2. Dora Gondocs and Viktor Dorfler, “AI in medical diagnosis: AI prediction & human
judgment”, January 2022. https://r2hcai.github.io/AAAI-23/files/CameraReadys/34.pdf
3. Binggui Zhou, Guanghua Yang, Zheng Shi, and Shaodan Ma, “Natural Language Processing for
Smart Healthcare”, September 2022. https://arxiv.org/pdf/2110.15803
4. Mahmoud nasr , Md. Milon Islam , Shady Shehata , Fakhri Karray , and Yuri Quintana,
“Smart Healthcare in the Age of AI: Recent Advances, Challenges, and Future Prospects”,
November 2021. https://arxiv.org/pdf/2110.15803
5. Rashika Raina1 and Rakesh Kumar Jha, “Intelligent and Interactive Healthcare System
(I2HS) Using Machine Learning”, November 2022.
https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9853209
6. Suryakanta Sahu, “Contextual Healthcare Chatbot using Deep Neural Network”, A Master’s
thesis submitted to national college of Ireland, 2022.
https://norma.ncirl.ie/6288/1/suryakantasahu.pdf
7. Nayif Saleh, Abdallah Kassem, And Ali M. Haidar, “Energy-Efficient Architecture for Wireless Sensor
Networks in Healthcare Applications”, 2017.
https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8247183