E-Healthcare and Chatbots in Preliminary Diagnosis: A Survey

Gokul Dudani, Tanu Shree, Kajal Singh, Anushka Singh Chauhan

Department of Computer Science & Engineering,

Galgotias College of Engineering & Technology, Greater Noida, India

Keywords: E-Health, Health Chatbot, Preliminary Diagnosis, Smart Communication, Conversational Agents (CA)

Abstract: E-healthcare, also known as Mobile Health is a healthcare practice that incorporates electronics & other

technologies in public health domain and is proposed to overcome geographical, temporal and organizational

barriers of the traditional healthcare architecture. A Health Chatbot is essentially an integral part of E-

Health. It is an Artificial Intelligence (AI) program capable of establishing a smart communication with the

user via auditory or textural methods regarding healthcare issues. These chatbots can be utilized for

providing preliminary diagnosis, as a consequence of which the current human workforce crisis in the

medical field may also be significantly reduced. This paper presents a survey of various characteristics of

chatbots and their existing applications in healthcare. It also includes a systematic review of existing

research on related domains and attempts to outline the essential research findings that are crucial to be

addressed. It provides insight into the desirable features of a chatbot for acceptance by both the doctor and

the patient. A sincere attempt has also been made to study the existing challenges and E-Health and

Chatbots future scope in providing a preliminary diagnosis.

1 INTRODUCTION

According to the World Health Organization

(1948) Health is defined as – "A state of complete

physical, mental and social well-being and not

merely the presence of disease or infirmity".

Since the dawn of human civilization,

accessibility to healthcare facilities has been a

fundamental right of a human being. In a world of

almost 7.5 billion people comprising vast diversity,

it poses an enormous challenge to the healthcare

delivery system. Moreover, due to the workforce's

uneven distribution globally, many people residing

in remote locations have minimal or no accessibility

to the healthcare infrastructure. The traditional

healthcare facilities are expensive and are usually

high-priced by the socially and economically

underprivileged and marginalized families.

E-Healthcare is a relatively recent healthcare

practice that incorporates electronics & other public

health technologies to overcome geographical,

temporal and organizational barriers of the

traditional healthcare architecture. Mobile Health

(m-Health) also aims to introduce smartphones and

other technological gadgets in the healthcare system.

It supports two-way communication; it engages user

interactively through a mechanism of proper

feedback and assists the doctors in suggesting

medical advice to the patients.

Health Chatbot is essentially the pillar on which E-

Healthcare system depends. Formally, a health

chatbot can be defined as- "An artificial intelligent

(AI) program that is capable of establishing a

smart communication with the user via auditory or

textural methods regarding healthcare issues".

There are a variety of chatbots used in the

healthcare domain. For example, Ada (Ada Health

GmbH 2018) enables its users to respond to

several diagnostic suggestions and identify over

1500 clinical pictures and 200 rare diseases (Flick

2018). Ever since the concept of health chatbots

has come into the limelight, its capabilities and

applications have increased. Its applications

include: -

 Reducing healthcare spending (increasing

affordability).

 Allow early detection of disease (preliminary

disease diagnosis).

 Improve clinical outcomes validated in a real

clinical-context.

Shortly, health chatbots will become the primary

contact for disease diagnosis and clinical treatment.

Dudani, G., Shree, T., Singh, K. and Chauhan, A.

E-Healthcare and Chatbots in Preliminary Diagnosis: A Survey.

DOI: 10.5220/0010562200003161

In Proceedings of the 3rd International Conference on Advanced Computing and Software Engineering (ICACSE 2021), pages 39-44

ISBN: 978-989-758-544-9

Table 1.

S.No. Reference Keywords Research Outcomes

Chin-Yan-Huang,

Ming-Chan Yang,

Chin-Yu Huang,

Yu- Jui Chen,

Meng-Lin Wu,

Kai-Wen Chen

Overweight,obese, smartphone,

mobile app, public health, artificial

intelligence, chatbot

The paper proposes a chatbot supported Smart

Wireless Interactive Health Care System (SWITCH-

es) with an objective of weight control and

spreading health awareness. The system can

establish smart communication and has a wide-

reaching approach for users who require medical

consultation.

Keng Siau, Weiyu

Wang, 2018

Health Chatbots, healthcare, trust,

artificial intelligence

The paper discusses the factors that affect the trust-

building process and develops a trust model that

depicts the building process between the user and

the health chatbots.

Arvind Kasthuri,

2018

Awareness,Accessibility,

Workforce Crisis, Affordability,

Accountability

The paper presents an overview of the existing

challenges to the healthcare system in India. The

limitations are expressed in five A's: Awareness,

Accessibility, Absence of Workforce or Human

Power Crisis, Affordability and Accountability.

Sven Laumer,

Christian Maier,

Fabian Tobias

Gubler, 2019

Conversational agents, disease

diagnosis, adoption, chatbot,

mHealth, UTAUT

The research focuses on developing a research

model that can be used as a basis for the acceptance

of health chatbots or conversational agents (CA) for

disease diagnosis in the medical domain. The model

comprises the UTAUT2 (Unified Theory of

Acceptance and Use of Technology) factors, newly

identified factors, and re-defining other factors to

etter fit a consumer context.

Ahmed Fadhil,

Gianluca Schiavo

Design Patterns, Bots,

Conversational UIs, Dialogue

Systems, Health and Wellbeing

The paper introduces CA-CUI (Conversational User

Interface) for healthcare. It also describes several

design principles and other complex elements for

building emotionally intelligent health chatbots for

interaction with humans. The design principles are

presented after a careful literature review of relevant

research works.

Flora Amato,

Stefano Marrone,

Vincenzo Moscato,

Gabriele Piantadosi,

Antonio Picariello,

and Carlo Sansone

eHealth, Big Data, Deep

Learning,Watson Decision Support

System, Prevention Pathways

The objective of the research was to study the

effectiveness of the classic human-machine

interaction for eHealth applications. Proposes an

approach of a chatbot that can interact with patients

as a medical professional validated in a real clinical

context to overcome the cons of restricted and

biased interaction between the human and the

computer software. It also proposes HOLMes

(Health Online Medical Suggestion), which is an

eHealth recommendation system.

Nivedita Bhirud,

Subhash Tatalle,

Sayali Randive,

Shubham Nayar,

2019

Chatbot, Healthcare Domain, ML

(Machine Learning), NLG (Natural

Language Generation), NLU

(Natural Language Understanding),

Smart Communication, Virtual

Communicatin

Frien

The paper presents a discussion about the various

NLU, NLG and ML techniques which must be

incorporated in health chatbots to overcome the

monotonous interaction between the user and the

chatbot and make the m-Health system more smart

and communicative.

Krishnendu Rarhi,

Abhishek

Bhattacharya,

Abhishek Mishra,

Krishna Mandal,

2017

Chatbot, Health

The research objective is to provide a design for

health chatbot that can provide disease diagnosis and

treatment through the user's symptoms to the

system. It replaces many existing chatbots and gives

a realistic experience to the user during the

interaction with the chatbot.

Divya S, Indumathi

V, Ishwarya S,

Priyasankari M,

kalpana Devi S, 2018

ArtificialIntelligence,

Prediction,Pattern matching,

Disease, Query processing

The paper emphasizes developing a chatbot with

improved symptom recognition and disease

diagnosis features with AI's help (Artificial

Intelligence).

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering

Table 1 (Cont.).

S.No. Reference Keywords Research Outcomes

Rashmi Dharwadkar,

Neeta A. Deshpande,

2018

Medical Chatbot Natural

LanguageProcessing,

PorterStemmer Algorithm, Word

Order SimilarityBetween

Sentences

The paper proposes a system through which users

can consult about the dosage of drugs through voice.

The system can effectively process the query and

can generate and display all medicine names. The

paper recommends incorporating various data

anal

tic techni

ues for buildin

the s

stem.

Rohan Jagtap, Kshitij

Phulare , Mrunal

Kurhade, Kiran

Shrikant Gawande

Teacher Forcing,DL(Deep

Learning), Inference Model,

Tensor Flow,Feed-Forward

NeuralNetwork,

SoftMax, Word-Embedding,

Conversation Context, GRU,

Diagnosis, Categorical Cross-

Entro

The idea presented in the study is to develop an AI-

powered chatbot which can extract the keywords

from the symptoms provided by the user and try to

diagnose the disease accordingly. It takes into

consideration the user's conversational and replies

accordingly to better fit the consumer context.

2 LITERATURE SURVEY

This section presents a literature survey of existing

researches. For each paper, a sincere attempt has

been made to summarize its main characteristics,

salient features and contributions. An effort has also

been made to prepare a draft of common jargons and

recurrent patterns. Moreover, it includes crucial

research observations and inferences that should be

addressed.

3 CHARACTERISTICS OF A

DESIRABLE HEATH

CHATBOT

This section focuses on the desirable characteristics

of health chatbots for eHealth applications. This

investigation aims to develop a human-machine

interaction mechanism with an approach that

leverages a human being validated in a real clinical

context.

There are certain factors which affect the adoption

of health chatbots in a real environment:

 Performance Expectations: the degree to

which chatbot can benefit the customer in providing

intended health and other related services.

 Effort Expectations: the extent of simplicity

associated while interacting with a chatbot.

 Social Influences: the willingness of an

individual to seek medical assistance from a health

chatbot.

 Facilitating Conditions: the availability of

resources or support mechanisms to effectively

interact with a conversational agent (CA) for health-

related FAQs.

 Hedonic Motivations: refers to the degree of

customer satisfaction after interaction with health

chatbot.

 Price Values: the cognitive trade-off between

the applications' perceived benefits and the

monetary cost of using them.

There are several other characteristics which should

be present in a desirable health chatbot. An ideal

health chatbot should be user compatible; it should

be practical and accurate in giving health

suggestions and providing disease diagnosis.

Interactivity with the user must be high. It should

also maintain integrity, privacy and ensure the

security of user's data. The trust-building process

with the user is also one of the critical features of a

desirable health chatbot. Health chatbots must also

be reliable, transparent, and accountable for any

recommendations and diagnosis provided by it.

Few other traits must be kept in mind while

designing a chatbot for medical applications. A

health chatbot designed for specific applications that

suit the needs of the customer should consider:

 User Demographics

 Chatbot Application Domain

 Data Interaction

 Dialogue Structure and so forth

Unfortunately, there has not been much discussion of

the critical design elements related to health

chatbots. Moreover, recent studies lack highlighting

crucial issues and challenges faced during the

development of realistic conversational agents.

When taken into consideration, these characteristics

can make the system more capable of establishing

smart communication with the user and can prove

fruitful for counselling and providing other health

E-Healthcare and Chatbots in Preliminary Diagnosis: A Survey

recommendations. These chatbots can be carefully

deployed in the medical industry to predict diseases

and provide a preliminary diagnosis to the patients

who can play a crucial role in saving human lives.

4 CHALLENGES TO

HEALTHCARE SYSTEM

There has been a glorious tradition of public

healthcare since the beginning of human civilization.

It can be seen in references to the study of the Indus

Valley Civilisation (5500-1300 BC) where the

concept of "Arogya" or holistic well-being has been

illustrated.

In the current scenario, the population of the world

has been multiplying. The population spectrum

presents enormous diversity and poses an enormous

challenge to the existing healthcare system.

Moreover, since the population distribution is

uneven, it becomes challenging for people living in

remote locations to have accessibility to the

healthcare facilities. There have also been

challenges in delivering healthcare to the so-called

"everyone" which includes the socially

marginalized, disadvantaged and the economically

deprived people of the society.

While there are many challenges, this section briefly

presents some (five A's) for the consideration: -

 Awareness: can be described as a general

understanding of an individual regarding health

issues. Lack of awareness is one of the critical

reasons for challenges in the healthcare domain.

Low educational status, poor literacy rate and low

priority for health in the people are the key reasons

for lack of awareness.

 Accessibility: can be defined as the

opportunity for availing any facility or service.

Physical access is one of the determining factors of

accessibility is defined as "the ability to enter a

healthcare facility within 5km from the place of

residence or work". Periodic surveys have revealed

that people living in remote locations have minimal

or no healthcare infrastructure access due to

temporal & geographical limitations.

 Absence or Workforce Crisis: a 2011 study

estimated that approximately 20 healthcare workers

are over a population of 10,000 people in India.

Moreover, there is a non-uniform distribution of

healthcare workers over the region. It has been

observed that the concentration of healthcare

resources in bare areas is dense as compared to

mountainous or other remote regions.

 Affordability or Cost of Healthcare: can be

thought of as the ability to avail healthcare

resources. The public sector usually offers

healthcare services at very cheap or no cost but is

often unreliable and inefficient. On the other hand,

the private sector offers services at a very high cost

and is generally the seeker's first choice unless

he/she cannot afford it. Thus, the economically

deprived section of the society is forced to make

unreliable and inefficient public healthcare services.

Thus, customer choice and satisfaction are often

neglected, which is a big challenge to the healthcare

system.

 Accountability or Risk of it: it is the set of

procedures by which one takes responsibility for

actions and its consequences. Generally, the private

sector is often held more accountable as compared to

the public health domain. The lack of trust and

communication gap between the seeker and the

public health sector poses a challenge to the existing

system.

There are numerous other factors than the five

mentioned above that present a challenge to

traditional healthcare architecture. As we move into

the future, there is an urgent need of looking out for

healthcare alternatives that should rule out the

present limitations and ensure universal health and

well-being.

An emphasis has to be made on the significance

of "Preliminary Diagnosis" in the medical field that

can be crucial in saving patients' lives and accurate

health delivery, which is currently missing from the

existing framework. E-heath and health chatbots can

be utilized to fulfil the purpose and provide medical

professionals with accurate results.

5 FUTURE SCOPE OF E-HEALTH

AND CHATBOTS

E-Health is a term used to describe the introduction

of electronics & other technologies in the traditional

healthcare infrastructure. Also, known as Mobile

Health (mHealth), it aims at overcoming the

geographical, temporal and organizational barriers

of the existing healthcare system. Health chatbots are

an integral part of mHealth. They are an Artificial

Intelligence (AI) program that provides users with

accurate disease diagnosis and their cure based on

the system's symptoms. It supports two-way

communication; it engages user interactively

through a mechanism of proper feedback and assists

the doctors in suggesting medical advice to the

patients.

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering

Figure 1: The Operational Concept of Health Chatbots

There are many chatbots which are currently

providing their services in the healthcare domain.

For example, Endurance is a chatbot which helps

users suffering from a disease known as Dementia.

The current limitations in the existing healthcare

delivery system can be overcome by deploying

health chatbots in the medical industry and

transitioning from health to e-health. The future

scope of e-health and health chatbots is discussed

below:

 E-health and related conversational agents

(CA) can provide general healthcare FAQs that can

spread awareness among the public. For example,

MedWhat is a question-answer chatbot that answers

basic healthcare FAQs and provides information

related to various diseases and their symptoms.

 Specific mobile applications can be used in

public health to keep track of nutrition, diet

planning, physical, mental & social activities

deemed more accurate than relying on one's analysis.

 Recent advancements in technology have led

health chatbots to overcome the disadvantages of

classic interaction between doctors and patients, thus

removing bias and allowing the patient to a freer and

comfortable interaction paradigm.

 Moreover, there are plans to create advanced

versions of chatbots that will emulate medical

professionals on one side of the communication in a

real clinical context and focus on improving the

health chatbots' human-like behavior.

Various data processing and analytics algorithms are

being used to transform healthcare, aiming to

provide modern digital healthcare solutions to the

patients. The whole e-healthcare delivery process is

more accurate and efficient, less expensive and of

high quality.

 Health chatbots are highly reliable and are

also capable of establishing smart communication

with the user. Moreover, the recommendation

building procedure is entirely transparent and

explainable, contributing to the mHealth

infrastructure's accountability. It also helps in the

development of the trust-building process between

the health chatbots and the users.

 Medical consultation involves user's data and

thus data protection, privacy and security become a

matter of prime importance which is effectively

maintained by e-health and health chatbots.

With the ever-growing advancements in technology,

health chatbots' capabilities have been growing

significantly and are likely to become the primary

point of contact for medical consultation and

diagnosis. Medical professionals would be using

health chatbots as a supporting tool for preliminary

disease diagnosis and medical consultation. On the

other hand, patients would use chatbots and mHealth

facilities for healthcare- related FAQs and other

recommendations.

6 SCOPE OF HEALTH

CHATBOTS IN PRELIMINARY

DIAGNOSIS

Diagnosis is the process of identifying a disease,

condition or injury from its sign and symptoms.

Preliminary diagnosis is an initial stage of medical

diagnosis that occurs before any or little symptoms.

Preliminary diagnosis is essential as it allows timely

detection of diseases and can play a crucial role in

saving the patient's life. Health chatbots can play an

essential role in providing a preliminary diagnosis to

the patients. It would enable accurate diagnosis as

well as treatment of diseases which will eventually

improve clinical outcomes.

According to research, approximately 60% of the

visits made to a medical practitioner are for small-

scale and straightforward diseases. These include

common cold, infections, headache, abdominal pains

and allergies etc. These diseases can be easily cured

at home using traditional home remedies. If ignored,

these diseases can act as base for several deadly

diseases like tuberculosis, cardiac arrest, brain

strokes, etc. Here preliminary diagnosis comes into

practical application.

There are a variety of chatbots which provide

services in the medical domain. But the majority of

these chatbots provide answers to general healthcare

FAQs only.

The future scope of these chatbots would be to

accurately predict the disease from the patient’s

early symptoms so that timely treatment could be

E-Healthcare and Chatbots in Preliminary Diagnosis: A Survey

provided, which would be crucial in saving lives.

Moreover, treatment of many small-scale diseases

could be directly provided by the chatbots, which

would reduce the current crisis of medical

professionals in the health sector.

7 CONCLUSION

The paper attempts to present a discussion on

eHealth and Health Chatbots in providing a

preliminary diagnosis to the patients. It also

describes various flaws existing in the current

architecture of the healthcare system and provides an

insight into the future scope of eHealth and utilizing

health chatbots to ensure universal health. The

research study then discusses the characteristics

present in an ideal health chatbot for acceptance by

both the doctor and the patient. The paper then

concludes with a systematic literature review of the

existing researches related to the field and sincere

efforts were put to outline the vital research findings

which were considered essential to be addressed.

REFERENCES

Ahmed Fadhil, Gianluca Schiavo, "Designing for Health

Chatbots", University of Trento, Italy

Arvind Kasthuri, "Challenges to Healthcare in India - The

Five A's", Indian J Community Med. 2018 Jul-Sep;

43(3): 141–143.

Chin-Yan-Huang, Ming-Chan Yang, Chin-Yu Huang, Yu-

Jui Chen, Meng-Lin Wu, Kai-Wen Chen "A Chatbot-

supported Smart Wireless Interactive Healthcare

System for Weight Control & Health Promotion",

Taiwan

Divya S, Indumathi V, Ishwarya S, Priyasankari M,

Kalpana Devi S, 2018, "A Self-Diagnosis Medical

Chatbot Using Artificial Intelligence", 2018

Flora Amato, Stefano Marrone, Vincenzo Moscato,

Gabriele Piantadosi, Antonio Picariello, and Carlo

Sansone, "Chatbots meets eHealth: automatizing

healthcare", DIETI - University of Naples Federico II,

via Claudio 21, 80125 Napoli (Italy)

Keng Siau, Weiyu Wang, "Trust in Health Chatbots",

December 2018

Krishnendu Rarhi, Abhishek Bhattacharya, Abhishek

Mishra, Krishna Mandal, "Automated Medical

Chatbots in Healthcare Industry”, January 2017

Montana SP, Magure T, Kandawasvika G. Geographical

access, transport and referral systems. In: Hussein J,

McCaw- Binns A, Webber R, editors. I Maternal and

Perinatal Health in Developing Countries. CAB

International e-books; 2012. pp. 139–54.

NationalCancerInstitute,https://cancer.gov.in/publications/

dictionaries/cancer/terms/def/diagnosis

Nivedita Bhirud, Subhash Tatalle, Sayali Randive,

Shubham Nayar, "A Literature Review on Chatbots in

Healthcare Domain", August 2019

Rao M, Rao KD, Shiva Kumar AK, Chatterjee M,

Sundararaman T., "Human resources for health in

India", The Lancet. 2011; 377:587–98.

Rashmi Dharwadkar, Neeta A. Deshpande, "A Medical

ChatBot", June 2018

Rohan Jagtap, Kshitij Phulare, Mrunal Kurhade, Kiran

Shrikant Gawande, "Health Conversational Chatbot for

Medical Diagnosis"

Roy S., "Primary health care in India", Health Popul

Perspect Issues. 1985; 8:135–67.

Sven Laumer, Christian Maier, Fabian Tobias Gubler,

"Chatbot Acceptance in Healthcare: Explaining User

Adoption of Conversational Agents for Disease

Diagnosis", May 2019

Valentine, D.B. and Powers, TL (2013), "Generation y

values and lifestyle segments", Journal of Consumer

Marketing, Vol. 30 No. 7, pp. 597-606.

Venkatesh, V., Thong, J. Y. L., and Xu, X. 2012.

"Consumer acceptance and use of information

technology: extending the unified theory of acceptance

and use of technology," MIS Quarterly (36:1), pp.

157–178.

World Health Organization (WHO), no.2, p. 100, 1948

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering