NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.

NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

An Amalgamation of Ayurveda and

Evidence-based Medicines with Artificial
Intelligence and Machine Learning: A Synergistic
Approach for Less Expensive and Effective
Diagnosis Approaches
H.M. Manjulaa1*, Dr.S.P. Ananda Raj2

Abstract
Technology plays a very important key role in various diagnosis processes. For example, capturing the patient
information, maintaining the electronic health records for further diagnostic references, improving the clinical
diagnosis process, and reducing diagnostic errors when technology is applied in the right ways with proper knowledge
of technology. In this proposed work, the design of the integrated analysis process of Ayur’veda and Western method
of diagnosis by using the current technologies such as Artificial Intelligence and Machine Learning techniques are 684
discussed. This integration process ensures that technologies simplify the timely and accurate diagnosis process.
Integration of the Ayurveda and Evidence-based western medicines has become very much necessary and important
in today’s situation. But, during this process, one should not disturb the basic principles of both Indian Traditional
Medicine and Evidence-based medicine. During integration, the best of best diagnosis approaches in both the fields
are chosen with disease management, disease prevention, and protection which create a healthy society by developing
healthcare with less expensive and less toxic. The amalgamation approach will be more effective than concentrating
on just one diagnostic method.

Key Words: Amalgamation, Ayurveda, Evidence-based Medicine, Artificial Intelligence, Machine Learning.
DOI Number: 10.14704/nq.2022.20.5.NQ22224 NeuroQuantology 2022; 20(5):684-695

Introduction
1. Data Science in the Field of the Health Sector data systems to improve statistics, healthcare and
Data science is a multidisciplinary area in which we medication distribution, and wellbeing evidence
can extract useful knowledge from structured and reportage on medical conclusions. Using Boolean
unstructured data sets by using the data mining operators, the data-mining tool enables accurate
technique. The invention of AI Artificial Intelligence knowledge searches. [1] Complex searches can be
Techniques, ML and Deep Learning has made the used to search for information about diseases,
Data science more powerful and dynamic. In the causative factors, symptoms, treatment guidelines,
dynamic digital age, a combination of science, medicines, nutritional recipes, lifestyle
technology, and medicine has discovered innovative modifications, and treatment processes.

Corresponding author: H.M. Manjulaa

Address: 1*Ph.D. Scholar, Presidency University, Itgalpura, Rajanakunte, Yelahanka, Bangalore, India; 2Associate Professor
Selection Grade, Presidency University, Itgalpura, Rajanakunte, Yelahanka, Bangalore, India.
E-mail: 1*manjulahm@presidencyuniversity.in; 2anandaraj@presidencyuniversity.in
Relevant conflicts of interest/financial disclosures: The authors declare that the research was conducted in the absence of
any commercial or financial relationships that could be construed as a potential conflict of interest.
Received: 25 March 2022 Accepted: 29 April 2022

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

The huge dataset is of from various domains based Volume, Variety, Velocity, and Veracity. Also Covid-
on its domain, scope, and the type that is been used 19 pandemic increased the medical data set in the
in different fields [2]. As per the information huge amount, especially in the medical field. The
generated from the world economic forum [3] 500 survey [4] shows that by 2027, the overall Big Data
million tweets each day, 294 billion E-mails, in the cyber security market will grow to $64.4
Facebook data of size 4PB, data of about 4TB from billion.
automated cars, 65 billion messages from Data-Science in the Medical field uses data analysis
WhatsApp, and huge data from YouTube and 5 from different resources like physicians, Electronic
billion searches are made each day. It’s been Health records, Laboratories, Search Engines,
estimated [3] that by the end of 2025, 463EB per day Research Studies, wearable Diseases, etc… as shown
will be generated from different resources. This in Figure [1].
analysis made data huge in the form of 4 V’s such as

Physician
s

Electronic
Wearable
Health
Devices
Record

Sources
of Health
Data
Research
Laborataries 685
Studies

Search
Engines

Figure 1. Sources of Health Data

Knowledge
Discovery in
•Systematic process to Database
•Methods appled to
convert information to extract relevent
Knowledge •Rules to apply to
convert data into Knowledge from Data
Knowledge
Data Mining
Data Science

Figure 2. Data Science –Knowledge Discovery- Data Mining

The interrelated process of Data Science, useful are incorporated to extract the relevant data from
Information Discovery in Databases, and analysis of the existing structure and unstructured data set.
data using data mining is shown in Figure 2. Data A detailed explanation of Knowledge Discovery
Science is a huge data field where knowledge is in Database processes like data selection,
discovered from the existing database and methods

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

pre-processing, transformation, data mining, and all resources. We can overcome this disadvantage by
evaluation was given by Fayyad et. al. [5]. applying Big Data Analysis techniques and
Knowledge Discovery in Database becomes slow incorporating AI, ML, and Deep Learning algorithms
and error-prone due to the fast increase of data from as shown in Figure 3.

Figure 3. Data Science Using Artificial Intelligence Technique

Data Science along with Machine Learning, Deep Divodas Dhanvantari learned Ayurveda from Lord
Learning which is subset of AI can provide a cost- Indra and passed it on to Sushruta, who wrote the
effective diagnosis and treatment in the medical Sushruta Samhita4 (sixth – fifth century BC) [7].
field. Thus the invention of AI gave a new way of Ayurveda an Indian Traditional Medicine (ITM) is
analysis of huge data in Data Science. played an important role since the ancient period in
providing health care in developing countries.
2. Ayurveda Before the invention of Modern Medicines, Herbal
plants have a great history from the ancient period
The science which tells about life or “Ayu” is
that was used in the treatment of many diseases. In
Ayurveda. Ayurveda is considered a science of life
developing countries like India, this information of 686
that is eternal. The knowledge or “Veda” is nothing
the Ayurvedic treatment is been lost due to
but the knowledge of happiness (good health) and
improper documentation of the Ayurvedic
unhappiness (unhealthy) along with this Ayurveda
medicines that were used in curing many diseases.
describes the root cause of disease. The history of
Lack of this evidence, people are diverting towards
Ayurveda can be traced back to various periods, the
Allopathy Treatment, which is fast in relief, well-
earliest being the Veda Purana in the ancient era,
equipped technologies that are used in the diagnosis
when the Aryan’s compiled the four Veda’s, with the
of many diseases which makes the decision faster to
most references in the Rigveda and Atharvaveda [6].
start the treatment. And well documentation is been
In Atharvaveda anatomical structure of the human
maintained regarding the treatment of many
body, metabolism, digestion, blood circulation,
diseases.
diseases, and their cause, mineral and surgical
Examining the patient by Ayurveda vaidhya’s is not
techniques, different types of worms and diseases
just by examining on a particular disease ,the
caused by them along with the diagnosis and
diagnosis process will be to both mind, body and
treatment have been described in the Atharvaveda.
soul [8]. The prediction of diseases and treatment
Classification of plants, their uses in the treatment of
process in our Traditional Medicine is from the root
particular diseases is all described in Atharvaveda.
cause by predicting imbalances in the Tridosha such
It is believed that Ayurveda, or Traditional Medicine,
as Vata, pitta, Kapha. The bodily processes, anatomy,
was transformed to sages from Gods , who then
analysis, medication, and treatments are all based
passed it on to others. Ayurveda is thought to have
on the examination of imbalance in Tridosha. Every
originated from heaven by Lord Brahma, who later
dosha is thought to have inherent characteristics
passed the knowledge on to Dakshaprajapati and
that manifest themselves in an individual's physical,
then to AshwiniKumaras. Later on, on Earth, when
psychological, and physiological characteristics [9].
diseases were at their peak, Maharishi Bharadwaj
The authentic Ayurvedic manuscript by Charaka’s
learned Ayurveda from Lord Indra. From then on,
and Sushruta’s Samhita explicitly explains how to
Sage Atreya received Traditional Medicine
identify dosha properties through signs and
knowledge and passed on to Agnivesha. Agnivesha
symptoms that make human being vikriti. Tridosha
wrote the Agnivesha Tantra, also known as the
analysis assists in prioritizing any nurturing,
Charaka Samhita (fifth Century BC). Kashiraj

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

preventative, and curative programme specific to a diagnostic method is examined by both types of
person. As a result, a tridosha-based prescription Ayurveda. Cohen’s weighted statistics with the
can aid to improve a regimen's therapeutic impact hypothesis of homogeneous diagnosis is been used
while lowering the drug's unpleasant side effects. to measure performance measure of both doctors
within the same diagnosis in same hospital and also
3. Allopathy /Modern Medicine between doctors from different hospitals. According
to doctors among those who observed concludes
Modern Medicine is also known as Allopathy, is an
that better training and experience may lead to more
evidence-based system that has the proof and
efficiency reliability. The author says that the
successful clinical trials [10] on the medications they
standardization of the pulse examination method
use. The term Allopathy was coined by Samuel
must be improved while using the integrated way of
Hahnemann in the year 1810. The diagnosis method
diagnosis, adding to this the author says that
of modern medicine physicians is disease-oriented.
integrating the new technology will better display
A physician checks the symptoms and writes the
the pulse pattern along with the doctor’s diagnosis
prescription. Allopathy follows the three main steps
for more efficient results.
that include Hypothesis, Experimentation,
Marques, Oge. [12]. this paper gives the importance
Observation, and Conclusion [10]. Thus we can say
of using computer science techniques and
that Allopathy treatment is a symptotic one i.e. it
algorithms, information science and technology,
treats the symptoms but not the root cause.
modern engineering in the field of Ayurveda. The
Allopathy approaches like body as a machine. Body
author explains several factors on why Ayurveda is
parts are treated as isolation. Allopathic medicines
becoming more popular in India. The author with
unlike the Ayurvedic medicines are made of
reference to different studies makes clear how the
chemicals and some of the plants used. The
use of technology in the Ayurveda diagnosis will
Allopathic system is systematic in treatment where
make the easy in decision making during diagnosis
the doctor looks at the symptoms and gives the
and also suggest that restructuring of Ayurveda is
treatment accordingly. The primary concept of 687
necessary in order to meet the demands of cyber
allopathy is to provide immediate relief to the
society.
patient by destroying the causative organisms of the
Wallace, Robert Keith. [13]. their study explains how
disease. The advantages of Allopathy are, gives fast
Ayurveda and genomics can both benefit from one
relief, is good in treating life-threatening diseases.
another and evidence-based systems and modern
Allopathy medicine uses modern technologies and
science can assist both Ayurveda and modern
diagnosis machines efficiently to understand the
medicine with preventive approaches in their
patient’s symptoms.
respective fields. The author explains how the four
P’s are predictive; preventive; personalized; and
4. Amalgamation of Ayurveda and Allopathy/Modern participatory are very efficient in modern medical
Medicine Diagnosis Approach diagnosis. Also in this study author focus on a new
The medical field is nothing but providing treatment approach in Ayurgenomics, which uses big data and
to a patient to overcome health-related issues. machine learning technologies can improve the
which can be used to overcome many health-related diagnosis approach and in fast prediction and
problems. An amalgamation of best practices of treatment of a disease.
Ayurveda and Modern medicine is being used as a Sharma, Rohit, and Pradeep Kumar Prajapati. [14],
best diagnosis methodology that can be applied to their study explains how Ayurvedic treatment is
manage and reduce the risk factors of diseases. The well suited with f predictive, preventive, and
review of the previous works related to integration personalized medicine along with
of Ayurvedic and modern medicine is given in the pharmacogenomics. The author explores this area
next session. that may aid researchers in identifying genetic
factors that may support classification level changes
5. Literature Review in ordinary persons, potentially leading to various
Kurande, Vrinda, et al. [11] in their study predict the disease Predictions. Further research into
pulse examination and give the test case report of amalgamation or supplementing genetics and
intrarater and interrater Ayurvedic doctors. This modern homeopathic discipline with Prakriti is
study gives a clear picture of how the pulse highly anticipated, as it may provide new insights
toward the individual healthcare aspiration.

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

Ruchika Nandha, Harpal Singh [15] in their study utility against the coronavirus that causes severe
explain its necessity for the integration of Ayurvedic acute respiratory syndrome (SARS).
and Allopathy. The author explores the integration Megas et al are a group of people who work together
of Ayurvedic therapies with the modern medical to solve problems. [20] Combining traditional
system that must be done without disrespecting any therapy with Anthroposophic therapies, according
of their diagnosis methodologies. The author to the researchers, could be helpful and
concludes that this integration may be useful in advantageous for individuals undergoing plastic
disease prediction, diseases prevention with the surgery. In addition to functional and physical
best results. approaches, this strategy stresses mental state,
Nisula, Tapio[16] in their study have done a survey creativity, and self-determination.
on how Ayurvedic practitioners have gained Anthroposophical massage (rhythmical and
popularity by integrating Ayurvedic with streaming massage), breathing treatment,
biomedicine during their practice. The author also ergotherapy, eurhythmy therapy, hyperthermia,
explores the drawbacks of the old diagnosis painting therapy, clay modelling therapy, music
methodology of Ayurveda. The authors also address therapy, physiotherapy, and psychotherapy are
how diagnosis tools and technology will have an some of the complementary approaches. In addition
impact on the medical integration diagnosis in to these mind-body techniques, natural goods are
Ayurveda during prescription. given.
Kataria, Sushila, et.al.[17]. in their study explores the Shaq, Abid, et, al. [21] in their study identify
outcome of the Ayurvedic treatment for the mild significant features and efficient Data mining
attack of COVID-19. In this study, the authors techniques that can improve the predictability of
address the Ayurvedic formulation of Tinopora cardiovascular patient survival. To predict patient
cadifalia which is also known as Guduchi and Pippali survival in this study, nine classification models are
integrated with modern medicine. The author used: To address the imbalanced class issue, the
predicts the impact of the integrated treatment for Synthetic Minority Oversampling Technique is used
688
COVID-19 by considering the parameters like (SMOTE). In addition, RF selects the highest-ranked
clinical recovery, interaction with Ayurvedic features for training machine learning models. The
doctors, safety once the patient got discharged after analysis are matched to those provided by AI
three months. The author further addresses that algorithms that use the entire set of features.
such integration medicine needs proper evidence to Experiment results show that ETC outperforms
provide effective and safe treatment for COVID-19 other models in prediction, achieving 0.934
disease. accuracies with SMOTE.
Seetharaman, Mahadevan et. al. [18] in their study Shaohui Wang et.al. [31] in their review elaborates
explain how the integration of convention medicine on the diagnosis and treatment of Rheumatoid
with Ayurvedic medicine will also be beneficial in arthritis (RA) by the integrated diagnosis of western
the prevention and treatment of communicable and and traditional medicine with an application of AI
chronic diseases COVID-19. The author explains in along with deep learning, cloud computation. In this
order to avoid future pandemics, a holistic approach review, the author predicts effective ethnic drugs
to healthcare that emphasizes prevention and can be prescribed with the help of AI technology.
immunity building is required. The traditional Aggarwal, Bharat B et.al. [32] in their study focuses
approach of diagnosis methods such as Ayurveda, on cancer treatment using traditional medicine
Chinese Traditional Medicine, and Yoga, would be along with modern medicine. This integrated
included in such a healthcare approach, as would treatment for cancer is very effective with the best
modern medical education and practice. results. The review in this article also presents
Kshirsagar and Rao (2021) [19] present recent the evidence that the Ayurvedic medicine not
research on Artemisia derivatives, which are widely only prevents cancer but also gives better results
used in many traditional medicines for their in cancer treatment. Since the Ayurvedic
properties such as antiviral, antifungal, medicine is fewer side effects can be used along
antimicrobial, insecticidal, hepatoprotective, and with the chemotherapeutic treatment which
neuroprotective [34]. Artemisinin, a well-known enhances therapeutic effects and reduces
phytochemicals derived from Artemisia, has been chemotherapy-induced toxicity. Saikat Sen and Raja
shown to have potent antiviral properties as well as Chakraborty [32] in this study predict how the
integration of traditional herbal medicine in clinical

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

practices helps in the achievement of the “Health for while increasing the distance along with both
ALL” concept. Also, author gives reviews on classes and reducing the classification errors [34].
challenges faced in the integrated medicine. The middle class distance is the distance between
the decision hyperplane and the immediate event a
Methodology class member is a part of. In SVM, each data point is
The amalgamation of centralized data using data first organized as a point in a n-dimensional space
mining, and using the AI-based machine learning with features such as a specific integration. For
algorithms is of great practical significance in the differentiation, a hyperplane with a large margin of
integrated diagnosis process, which helps in better difference between the two categories is available.
treatment with less cost. The integrated treatment
of both Ayurveda and Modern Medicine is proving
the best treatment for the diseases like diabetes,
cough, cold, arthritis, liver diseases, cancer, piles,
etc.
Machine Learning techniques are based on a set of
well defined rules which includes sets of
mathematical methods that describes the
relationship between the variables while predicting
the results. The model proposed in this predicts the
particular disease depending on the dataset
extracted from the best diagnostic features of both Figure 5. SVM illustration to identify hyperplane
Indian Traditional medicine and Evidence based
medicine diagnosis process. The primary concern of
using the ML algorithms is to predict the result in
more accuracy. The ML algorithms is divide into 689
three categories, Supervised Machine Learning,
Unsupervised Machine Learning and Supervised
Machine Learning
Supervised ML[22] refer to the technique in which
the model is trained on the different features or
input data along with the known output. In Medical Figure 6. Illustration of DT with elements and rules
field this is related with the symptoms of particular
diseases. So once the model is trained successfully, it Decision Tree
will be capable of predicting the particular diseases
on the given input.. The prediction made by using Decision tree is among the supervised machine
Supervised ML can be in the form of discrete e.g. learning algorithms used in regression and
positive or negative, or prediction can be continuous classification problems. In DT, nodes represent the
e.g. using the score from 0 to 100. Thus the data and the leaves of a tree represent the final
prediction made from discrete is referred as result. The decision tree is created by using training
Classification algorithms. In our proposed model data during the training process. The DT manages
different classification algorithm are applied to both categorical and numerical data set. In DT no
predict the result. pre-processing of data is required. A few
applications of decision tree are used in medical
Support Vector Machine Supervised ML Algorithm field to detect breast cancer, in which leaf nodes of
(SVM) the tree are split into two groups (Benign or
Malignant). The DT tree has different levels with the
The SVM algorithm is among the categories of nodes at each level is data and the first node
aligned machine learning algorithms used for represents the root node of a tree. The level of the
different line and non-line layouts. The SVM DT varies depending on the prediction problem
algorithm starts to map each data set in the statement. DT is more efficient to infer, analyze and
N-dimensional features space where the "N" is the fast to analyze in the medical diagnostic protocol. An
sum of the number of features. It later detects a illustration of DT with its elements and rules is
hyperplane that divides the data into two classes shown in Figure 6.

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

Random Forest Boosting

Random forest [35] is an ensemble classifier and The boosting algorithm constructs base classifiers in
consists of many decision trees similar to the way a sequential fashion. The basic idea behind sequence
the forest is represented as a collection of many is that on the training set, a weak classifier is built
trees. The efficiency and performance of the first. If the sample is correctly classified, it will be
Ensemble algorithm can be improved by building assigned a small weight in the training set,
base learners from the various approaches. otherwise it will be assigned a relatively large
number of weights based on the classification result
Bagging of the previous classifier. In order to achieve the best
Bagging [36] is one of the most widely used classification performance, this process will be
ensemble-building techniques. Bootstrap sampling repeated several times with several weak classifiers.
and model aggregation are two crucial steps in The boosting algorithm produces a final model that
bagging. Bootstrap sampling selects N samples from is a linear combination of many base classifiers
the data set using sampling with replacement, weighted by their own results. Freund and
ensuring the independence of the various sampling Schapire's AdaBoost is one of the most effective
training sets. Furthermore, the major voting method boosting algorithms. Figure 9 represents the
is used, which accepts the classification results with AdaBoost algorithm.
the most occurrences as the final classification
results if multiple base learners are used. Figure 8
depicts the algorithm [36] for bagging.

690

Figure 8. Bagging Algorithm

Figure 9. AdaBoost Algorithm

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

XGBoost (eXtreme Gradient Boosting) as extraction of medical-related rules in the medical

Chen and Guestrin [37] proposed XGboost in 2016. database [30], to analyze the diseases using
It is a more advanced version of the boosting clustering and K means clustering techniques.
algorithms, and it has been recognised as an
advanced estimator with extremely high Data Collection
performance in classification and regression. The The proposed model integrates the best features of
XGboost expression is given below to prevent both Ayurvedic and Modern Medicine and is
overfitting during analysis. analyzed using the machine learning algorithms.
The dataset related to chronic diseases is collected
from Ayurvedic Vaidya’s, PubMed, Google Scholar,
To deal with overfitting issues, XGBoost employs the
Scopus, and Science Direct. The dataset and
learning rate, boosting number, maximum tree
information on the medicine and diagnosis method
depth, and subsampling.
are extracted using the keywords such as Ayurvedic
Machine Learning algorithms is becoming most
Diagnosis, Modern Medicine Diagnosis, and
popular in the field of the Medical sector in wide
Diagnostic tools. Figure 11 explains the integrated
research areas such as analysis and prediction of
method analysis of different diseases using AI
diseases using supervised classification [22], cancer
model. The detailed description of the proposed
prediction with risk valuation after the surgery [23],
model is explained in Figure 12. Data is collected
frequent disease data mining by applying Apriori
from both the diagnosis methods of Ayurveda and
algorithm [24], predicting breast cancer
Modern Medicine from different resources as
survivability [25], prediction of drug dosage
mentioned earlier. The same is shown in Figure 12.
[27 – 29]. In the healthcare sector unsupervised
learning algorithms are used in different areas such

691

Figure 11. Integrated Diagnosis process of Disease using AI-Trained Model

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

692

Figure 12. Block Diagram of Proposed Model for analysis of diseases using AI and ML techniques in the integrated medicine approach

Preprocessing of Data and Data Refinement whoch are used to train the model and evaluate the
Preprocessing the data and Data refinement is performance respectively. 80% of the samples are
carried out to select most discriminative variables. taken as training data set and 20% of the samples
The variables with the unique values, irrelevant are taken as testing data set. Diagnosis ability and
variables with label are filtered, and only data performance of different algorithms are determined
pertinent to model training is retained. The data set by evaluation matrices such as confusion matrix and
is divided into training data set and testing data set Receiver Operating Characteristics (ROC).

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

Figure 12. [a] The basic Framework of confusion matrix [b] Presentation of ROC Curve

In ML technique, the evaution matrix called The ROC curve is created by plotting the true
Confusion Matrix is used for anaysing the positive rate against the false positive rate at various
performance of the algorithms. In this, True-Positive threshold settings [38]. It is one of the fundamental
(TP) are actual results correctly identified by tools used in diagnostic test evaluation. The area
different algorithms. Similarly, True-Negative (TN) under the ROC curve can also be used to assess the
are undesirable cases correctly identified by predictability of various machine learning
different algorithms. False-Positive (FP) are classifiers.
undesirable cases where the algorithms classify
erroneously as positive. False Negative (FN) are Result Analysis
positive cases where different algorithms false
The main objective is to study is to analyze the
classification as negative. The perfomance
comparitive performance evaluation among 693
measures based on the confusion matrix are
different supervised machine learning algorithm
analyzed for different machine learning algorithms
such as SVM, Decision Tree, Ensembled Algorithm
in this proposed model are given by
Random Forest with Bagging and Boosting, KNN,
a. Accuracy: The accuracy measure is applied to
Naive Bayes, Logistic Regression. The comparison is
recognize correctly predicted values among all
analysed based on the evaluation matrix namely,
the other values in a data set. Accuracy from the
accuracy, precision, F1, Sensitivity and specificity.
confusion matrix is evaluated as:
𝑇𝑃 + 𝑇𝑁 Experimental results predicts that the ensemble
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 = learing algorithmgives best performance compared
TP + TN + FP + FN to other algorithms. May be because the proposed
b. F1 Score: Is taken as the harmonic mean of
model with other algorithms is overfitting the
precision and sensitivity. In F1 score True
trainaing data.
Negative values are not considered.
2 X TP
𝐹1 𝑠𝑐𝑜𝑟𝑒 = Conclusion
2 𝑋 𝑇𝑃 + 𝐹𝑁 + 𝐹𝑃
c. Precision: This predicts the true diseased The purpose of this study is to analyze the
patient from the total number of data set taken performance of different machine learning
for analysis. algorithms used in the integrated diagnosis method
TP of Ayurvedic and Mosern Medicine.The best analysis
Precision =
TP + FP and common features from both the diagnosis
d. Sensitivity: is used for categorizing the patients methology i.e. from Ayurvedic and Modern Medicine
who are having disease from patient dataset. is extracted fron the existing medical health record.
TP
Sensitivity = Recall = 𝑇𝑟𝑢𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑟𝑎𝑡𝑒 = TP+FN This integrated dioagnosis model using best
Specificity : Is used for predicting the patients who Machine Learning algorithm is proven to serve the
are not having the disease from the patient dataset. fast diagnosis process for may choronic diseases
TN with cost effective treatment. Thus with the use of
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑖𝑡𝑦 = AI-ML techniques will invent new therepeutic
TN + FP
𝐹𝑃 methods by integrating diagnosis model with fast
𝐹𝑎𝑙𝑠𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑟𝑎𝑡𝑒 = and most accurate disease prediction and treatment.
FP + TN

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches

Again integration is not the blind fold acceptance. clinical study of an add-on Ayurvedic formulation
Integration must be done with the best treatment containing Tinospora cordifolia and Piper longum in mild to
moderate COVID-19." Journal of Ayurveda and Integrative
both from Ayurvedic and Modern Medicine. So Medicine 13, no. 2 (2022): 100454.
AI-ML techniques can play very important role in the Seetharaman, Mahadevan, Geetha Krishnan, and Robert H.
decision making during diagnosis. Schneider. "The Future of Medicine: Frontiers in Integrative
Health and Medicine." Medicina 57, no. 12 (2021): 1303.
References Kshirsagar, Suhas G., and Rammohan V. Rao. "Antiviral and
immunomodulation effects of Artemisia." Medicina 57, no. 3
Bheemavarapu, Lakshmi, and K. Usha Rani. "A Review on Role (2021): 217.
of Data Science in Ayurveda Based Disease Diagnosis Using
Megas, Ioannis-Fivos, Dascha Sophie Tolzmann, Jacqueline
Prakriti Type in Trividha Pariksha." Information Technology
Bastiaanse, Paul Christian Fuchs, Bong-Sung Kim, Matthias
in Industry, 9, no. 3 (2021): 1038-1048.
Kröz, Friedemann Schad, Harald Matthes, and Gerrit Grieb.
Wikipedia Data: https://en.wikipedia.org/wiki/Data "Integrative Medicine and Plastic Surgery: A Synergy—Not
World Economic Forum: Desjardins, Jeff. "How much data is an Antonym." Medicina 57, no. 4 (2021): 326.
generated each day." In World economic forum, vol. 17. 2019. Shaq, Abid, Saima Sadiq, Muhammad Umer, Saleem Ullah,
https://www.lowyinstitute.org/publications/big-data-and- Seyedali Mirjalili, Vaibhav Rupapara, and Michele Nappi.
national-security-guide-australian-policymakers "Improving the prediction of heart failure patients’ survival
Fayyad, Usama, Gregory Piatetsky-Shapiro, and Padhraic Smyth. using SMOTE and effective data mining techniques." IEEE
"From data mining to knowledge discovery in databases." AI access 9 (2021): 39707-39716.
magazine 17, no. 3 (1996): 37-37. W. Dai, T.S. Brisimi, W.G. Adams, T. Mela, V. Saligrama and I.C.
Kumar, Shailendra, and Sanghamitra Choudhury. "Ancient Vedic Paschalidis. “Prediction of hospitalization due to heart
Literature and Human Rights: Resonances and diseases by supervised learning methods,”International
Dissonances." Cogent Social Sciences 7(1) (2021): 1858562. Journal of Medical Informatics, Elsevier,vol.84, no.3, pp.189 –
Khwaja, Ayub Zickriya. "Significance of Ancient Ayurveda in 197, Mar 2015.
Soul-Body Relationship." Annals of Ayurvedic Medicine, 9, T. Oliveira, E. Barbosa, S. Martins, A. Goulart, J. Neves and P.
4(2020): 313-319. Novais,“A Prognosis System for Colorectal Cancer”,
V.J. Thakar, “Diagnostic Methods In Ayurved”, Ancient Science of Proceedings of 26th International symposium on Computer
Life, Vol No. I No.3 January 1982, Pages 139 – 145. based Medical systems, Jun 2013, Porto, Portugal, IEEE,
481-484. 694
Vrinda Kurande, Anders Ellern Bilgrau,Rasmus
Waagepetersen,Egon Toft, and Ramjee Prasad “Interrater M. Ilayaraja and T. Meyyappan. “Mining medical data to identify
Reliability of Diagnostic Methods in Traditional Indian frequent diseases using Apriori algorithm,” Proceedings of
Ayurvedic Medicine”, Volume 2013 |Article International Conference on Pattern Recognition, Informatics
ID 658275. https://doi.org/10.1155/2013/658275 and Medical Engineering (PRIME), Feb 2013, Salem, India,
Evidence-Based Complementary and Alternative IEEE, pp. 1-6.
Medicine / 2013 / Article D. Delen, G. Walker and A. Kadam. “Predicting breast cancer
https://www.jeevana.in/benefits-ayurveda-allopathy survivability: a comparison of three data mining methods,”
Artificial intelligence in medicine, 34(2), 113–127, Jun 2005.
Kurande, Vrinda, Rasmus Waagepetersen, Egon Toft, and
Ramjee Prasad. "Intrarater and interrater reliability of pulse P.J. Hu, T.H. Cheng, C.P. Wei, C.H. Yu, L.F. Chan, and H. Y.Wang.
examination in traditional Indian Ayurvedic medicine." “Managing clinical use of high-alert drugs: A supervised
Integrative medicine research 2(3) (2013): 89-98. learning approach to pharmacokinetic data analysis,”
Systems Man and Cybernetics, Part A: Systems and Humans
Marques, Oge. "Integrating contemporary technologies with
IEEE Transactions, vol. 37, no. 4, pp. 481–492, Aug 2007.
Ayurveda: Examples, challenges, and opportunities."
In 2015 International Conference on Advances in Computing, Y.H. Hu, F. Wu, CL. Lo and C.T. Tai. “Predicting warfarin dosage
Communications and Informatics (ICACCI), 1399-1407. IEEE, fromclinical data: A supervised learning approach,” Artificial
2015. Intelligence in Medicine, vol.56, no.1, pp.27–34, Sep 2012.
Wallace, Robert Keith. "Ayurgenomics and modern medicine." Y. Wenqi, N. Widmer and G.D. Micheli, “Example-based support
Medicina 56, no. 12 (2020): 661. vector machine for drug concentration analysis,”
Proceedings of thirty third Annual Conference Engineering in
Sharma, Rohit, and Pradeep Kumar Prajapati. "Predictive,
Medicine and Biology Society, EMBC, Aug-Sep 2011, Boston,
preventive and personalized medicine: Leads from
USA, IEEE 2011, 153–157.
Ayurvedic concept of Prakriti (human constitution)."
Current Pharmacology Reports 6, no. 6 (2020): 441-450. R. Shirwaikar, N. Mago, D. Acharya, K. Makkithaya, G. Hegde. L.
Lewis and Shivakumar “Partition and Hierarchical Based
Nandha, Ruchika, and Harpal Singh. "Amalgamation of ayurveda
Clustering Techniques for Analysis of Neonatal Data,”
with allopathy: A synergistic approach for healthy society."
Proceedings of International Conference on Cognition and
International Journal of Green Pharmacy (IJGP), 7(3): (2013).
Recognition, D. Guru, T. Vasudev, H. Chethan. Y. Kumar Eds.
Nisula, Tapio. "In the presence of biomedicine: Ayurveda, Lecture Notes in Networks and Systems, Singapore, 14
medical integration and health seeking in Mysore, South Springer, 2018.
India." Anthropology & Medicine, 13(3) (2006): 207-224.
R. Paul and H. Latiful Abu, “Clustering medical data to predict
Kataria, Sushila, Pooja Sharma, Jai Prakash Ram, Vikas Deswal, the likelihood of diseases,” Proceedings of 5th International
Manish Singh, Rakesh Rana, Richa Singhal, Arunabh Conference on Digital Information Management, ICDIM, Jul
Tripathi, Kuldeep Kumar, and Naresh Trehan. "A pilot

eISSN 1303-5150 www.neuroquantology.com

NeuroQuantology | May 2022 | Volume 20 | Issue 5 | Page 684-695 | doi: 10.14704/nq.2022.20.5.NQ22224
H.M. Manjula et al / An Amalgamation of Ayurveda and Evidence-based Medicines with Artificial Intelligence and Machine Learning: A Synergistic Approach
for Less Expensive and Effective Diagnosis Approaches
2010, Lakehead University, Thunder Bay, Canada, IEEE,
44-49.
Wang, S., Hou, Y., Li, X., Meng, X., Zhang, Y., & Wang, X. (2021).
Practical Implementation of Artificial Intelligence-Based
Deep Learning and Cloud Computing on the Application of
Traditional Medicine and Western Medicine in the Diagnosis
and Treatment of Rheumatoid Arthritis. Frontiers in
pharmacology, 12, 765435-765435.
Aggarwal, Bharat B., Haruyo Ichikawa, Prachi Garodia, Priya
Weerasinghe, Gautam Sethi, Indra D. Bhatt, Manoj K. Pandey,
Shishir Shishodia, and Muraleedharan G. Nair. "From
traditional Ayurvedic medicine to modern medicine:
identification of therapeutic targets for suppression of
inflammation and cancer." Expert opinion on therapeutic
targets, 10(1) (2006): 87-118.
Sen, Saikat, and Raja Chakraborty. "Revival, modernization and
integration of Indian traditional herbal medicine in clinical
practice: Importance, challenges and future." Journal of
traditional and complementary medicine, 7(2) (2017):
234-244.
Uddin, Shahadat, Arif Khan, Md Ekramul Hossain, and
Mohammad Ali Moni. "Comparing different supervised
machine learning algorithms for disease prediction." BMC
medical informatics and decision making, 19(1) (2019): 1-
16.
Denisko, Danielle, and Michael M. Hoffman. "Classification and
interaction in random forests." Proceedings of the National
Academy of Sciences, 115, no. 8 (2018): 1690-1692.
Breiman, Leo. "Bagging predictors." Machine learning, 24(2)
(1996): 123-140. 695
Chen, T.; Guestrin, C. Xgboost: A scalable tree boosting system.
In Proceedings of the 22nd ACM SIGKDD International
Conference on Knowledge Discovery and Data Mining, San
Francisco, CA, USA, 13–17 August 2016; 785–794.
Fawcett, Tom. "An introduction to ROC analysis." Pattern
recognition letters, 27(8): (2006): 861-874.

eISSN 1303-5150 www.neuroquantology.com