Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

A Chatbot System for Supporting Women and

Families during Pregnancy
1
Dr. P. Bhaskar
Professor, Dept. of CSE
QIS College of Engineering and Technology(A),
Ongole, Andhra Pradesh–523001, India

2 3
Tanuja Kunchala Srujana Peddi
Dept. of CSE Dept. of CSE
QIS College of Engineering and Technology(A), QIS College of Engineering and Technology(A),
Ongole, Andhra Pradesh–523001, India Ongole, Andhra Pradesh–523001, India

4 5
Rizwana Syed Pavani Namepalli
Dept. of CSE Dept. of CSE
QIS College of Engineering and Technology(A), QIS College of Engineering and Technology(A),
Ongole, Andhra Pradesh–523001, India Ongole, Andhra Pradesh–523001, India

Abstract:- PregBot is an innovative system that harnesses easily accessible healthcare solutions [1]. This paradigm shift
the power of machine learning (ML) and natural language holds profound implications for specialized areas like maternal
processing (NLP) to provide comprehensive support to and perinatal healthcare, which present unique challenges due
women and families throughout the pregnancy journey. to the intricate nature of care required and the diverse needs of
Recognizing the varying needs and challenges faced by patients [2]. Against this backdrop, our research introduces
expectant mothers, PregBot aims to revolutionize the "PregBot," an innovative system engineered to capitalize on the
maternal healthcare experience by offering personalized combined potential of ML and NLP to provide crucial support
guidance, real-time query resolution, and a virtual to pregnant women and their families [3].
community for support and connection. The system
leverages ML algorithms to analyze user data and tailor Pregnancy, while a time of great joy, is also characterized
responses, while NLP techniques enable natural language by uncertainties and obstacles. The availability of reliable,
interactions, allowing users to communicate with PregBot in personalized health information and emotional support during
a conversational manner. By continuously learning from this period is crucial, yet conventional healthcare systems often
user interactions, PregBot adapts and evolves, ensuring the struggle to meet these demands effectively [4]. Recognizing
delivery of timely and relevant information based on the this critical gap, PregBot is meticulously designed to offer
user's unique circumstances and stage of pregnancy. With comprehensive assistance, spanning from nutritional
its innovative approach to maternal healthcare, PregBot recommendations to mental health support, through the
represents a significant step towards empowering women, utilization of AI technologies to deliver tailored and
promoting positive pregnancy experiences, and contributing contextually relevant guidance [5]. PregBot leverages advanced
to the overall well-being of expectant mothers and their ML algorithms to analyze vast amounts of data, including
families. medical literature, clinical guidelines, and individual health
records, to provide evidence-based recommendations tailored to
Keywords:- Pregnancy, Machine Learning, Natural Language each user's specific needs and circumstances [6]. Through
Processing, Maternal Healthcare, Virtual Assistant. sophisticated NLP techniques, PregBot can engage in natural
language conversations with users, understanding their
I. INTRODUCTION inquiries, concerns, and preferences, and providing empathetic
and informative responses in return [7].
In the landscape of healthcare, the integration of Artificial
Intelligence (AI), particularly Machine Learning (ML) and One of the key strengths of PregBot lies in its ability to
Natural Language Processing (NLP), has emerged as a adapt and evolve alongside the user's journey through
transformative force, ushering in a new era of tailored and pregnancy. By continuously learning from user interactions and

IJISRT24APR697 www.ijisrt.com 522

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

feedback, PregBot can refine its recommendations and support the potential of chatbots to provide effective support in this
strategies, ensuring that they remain relevant and effective context.
throughout the entire pregnancy and postpartum period [8].
This paper outlines the development of PregBot, its A lightweight mobile application aimed at providing
foundational technologies, and the unique approach it takes to maternity guidance, designed by M. Umme Kulsum et al. [8].
cater to the nuanced needs of pregnant women and their They consider the application's functionality and its potential
families [10]. benefits in offering assistance to pregnant mothers.

II. LITERATURE REVIEW The feasibility of a mental health chatbot and its impact on
postpartum health are examined by Suharwardy et al. [9] in
The study by Vaira et al. [1] examines MamaBot, a AJOG Global Reports. Their randomized controlled trial
sophisticated tool built on ML and NLP, designed to support evaluates the effectiveness of chatbot interventions in maternal
pregnant women and their families. This research explores the mental health.
potentials and mechanics of MamaBot in delivering
personalized care and guidance throughout pregnancy. Chung et al. [10] contribute to JMIR Medical Informatics
with their development and evaluation of a chatbot catering to
Mugoye et al. [2] contribute to the discourse on Smart-bot obstetric and mental health care for perinatal women and their
technology, focusing on its potential applications in maternal partners, underscoring the importance of digital tools in
healthcare. Their study examines the pivotal roles that supportive care.
conversational agents, or chatbots, could fulfill in supporting
expectant mothers. Kaneho et al. [11] conducted survey on existing healthcare
chatbots designed for pregnant women. They critically analyze
Montenegro et al. [3] delve into the practical aspects of the chatbots' features and potential improvements, aiming to
chatbot usability within the prenatal context. Their study aims enhance digital healthcare services for expectant mothers.
to evaluate how these digital tools perform in real-world
scenarios, specifically assessing their user-friendliness and Afrizal et al. [12] focus on user-centric design principles
effectiveness in assisting pregnant women. in their research. Their study revolves around evaluating how
Natural Language Processing (NLP) techniques can be tailored
Oprescu et al. [4] conduct a comprehensive scoping to enhance maternal monitoring chatbot systems, with a
review examining the intersection of artificial intelligence (AI) primary emphasis on meeting user needs and improving
and pregnancy. Their research provides an extensive catalog of interaction quality.
diverse AI applications within this domain, offering a
panoramic view of the current landscape and future directions. Sadavarte and Bodanese [13] present an innovative
application of AWS and Alexa technology in the creation of a
In their study, Puspitasari et al. [5] contribute to the pregnancy companion chatbot. Their work underscores the
advancement of prenatal care through the development of a enhanced accessibility and convenience that these technologies
chatbot integrated into Indonesia's Posyandu Application. By bring to prenatal care, allowing pregnant women to access
leveraging decision tree methodology, Puspitasari et al. aim to relevant information and support using voice commands
optimize the functionality and effectiveness of the chatbot in through devices like Amazon Echo.
providing personalized guidance and support to pregnant
women. Arunkumar et al. [14] contribute to the advancement of
women's personal health promotion. They showcase an
Marin and Goga's [6] conference paper investigates the application leveraging machine learning techniques designed to
development of a chatbot specifically designed for counseling empower women in managing their health and wellness
on preeclampsia, a potentially serious pregnancy complication. effectively.
Their research, explores the feasibility and effectiveness of
utilizing such a tool to provide specialized advice and support Tumpa et al. [15] describe Smart Care, an intelligent
to pregnant women facing the challenges associated with assistant tailored for pregnant mothers. Their paper discusses
preeclampsia. the potential of such systems to supplement traditional care
with AI-enhanced support and guidance.
On a related note, the research conducted by R. Wang et
al. [7], delves into the application of supervised machine Olmedo-Requena et al. [16] investigate factors that
learning in chatbots to support mental healthcare during the influence adherence to nutritional guidelines before and during
perinatal period. This research acknowledges the significant pregnancy in Women & Health. Their study provides critical
emotional and psychological challenges that pregnant women insights into dietary behaviors and the implications for maternal
may face, particularly during the perinatal period, and explores and fetal health outcomes.

IJISRT24APR697 www.ijisrt.com 523

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

The creation of an AI chatbot behavior change model is Krishnaveni et al. [29] present a speech recognition
the focus of Zhang et al. [17] in the Journal of Medical Internet module for patient monitoring in smart healthcare applications.
Research. They theorize how AI can be a force for positive Their system aims to capture patient-reported data efficiently
change in promoting physical activity and healthy eating habits and accurately for healthcare use.
among expectant mothers.
Khan et al. [30] review AI approaches for maternal and
The potential of mobile personal health records for neonatal health in low-resource settings in Frontiers in Public
monitoring pregnancy is analyzed by Bachiri et al. [18] in their Health. They assess the viability and impact of AI-driven
Computer Methods and Programs in Biomedicine article. They interventions in environments with limited healthcare
discuss the capabilities and prospective improvements in digital infrastructure.
tools for pregnancy tracking.
III. ABOUT DATASET
A systematic review of mobile app interventions on
maternal behavior and perinatal health outcomes is conducted The dataset utilized in this study comprises a rich
by Daly et al. [19] in JMIR mHealth and uHealth. They provide amalgamation of user interactions, health metrics, and feedback
evidence on the effectiveness of digital interventions in collected through the PregBot system over a period of six
supporting healthy maternal practices. months. It includes data from a diverse cohort of pregnant
women who engaged with PregBot for various support services,
Frid et al. [20] systematically search, evaluate, and such as nutritional guidance, mental health counseling, and
analyze the features of mobile health apps targeted at pregnant general pregnancy-related queries. The dataset is characterized
women. Their research outlines the current landscape of these by its multidimensionality, encompassing textual interactions
apps and their utility in prenatal care. between users and the chatbot, anonymized health information
provided by the users, and user feedback on the utility and
Chung et al. [21] detail their work on a chatbot that assists effectiveness of the support received. This comprehensive
perinatal women and their partners with obstetric and mental dataset allows for an in-depth analysis of user needs, behaviors,
health care in JMIR Medical Informatics. They consider the and preferences, facilitating the continuous improvement of
bot's usability and its potential to improve care during this PregBot's ML and NLP algorithms to better serve its users.
critical life stage.
To ensure the integrity and reliability of the dataset,
The paper by Sharma et al. [22] in Electronics discusses a several data preprocessing steps were undertaken. These
stress detection system using an advanced neural network for included the anonymization of personal information to protect
working pregnant women. They assess the system's accuracy user privacy, the categorization of textual interactions for
and potential to provide support in the workplace. efficient processing, and the normalization of health metrics to
standardized units. Furthermore, sentiment analysis was applied
Raza et al. [23] report on the use of ensemble learning for to textual interactions to gauge user sentiment and emotional
analyzing maternal health during pregnancy in Plos One. Their states, providing valuable insights into the efficacy of PregBot's
research investigates how this approach can lead to more responses and the overall user satisfaction.
precise health risk predictions for expectant mothers.
The analysis of the dataset employed a variety of ML and
Zhou et al. [25] research in Biomedical Engineering NLP techniques to uncover patterns, trends, and insights into
explore the applications of NLP in smart healthcare systems. the usage and impact of PregBot. Machine learning algorithms
They discuss the advancements and challenges in using were used to predict user needs and customize responses, while
language processing to improve healthcare delivery and patient natural language processing facilitated the understanding and
outcomes. generation of natural language interactions. The dataset's
analysis aimed to assess the effectiveness of PregBot in
A systematic review on using ML to predict pregnancy improving maternal health knowledge, enhancing user
complications by Bertini et al. [26] is presented in Frontiers in engagement, and providing emotional support. Through this
Bioengineering and Biotechnology. The paper synthesizes empirical investigation, the study demonstrates the potential of
research on the effectiveness of machine learning models in AI-driven tools like PregBot to offer personalized, responsive
antenatal care. healthcare support, highlighting the critical role of data in
developing and refining AI applications in the healthcare
Tebenkov and Prokhorov [28] discuss machine learning domain.
algorithms for teaching AI chatbots in Procedia Computer
Science. They provide insights into the instructional
frameworks that can enhance chatbots' learning processes.

IJISRT24APR697 www.ijisrt.com 524

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

IV. PROPOSED METHODOLOGY Moreover, the use of Gini Impurity as the splitting
criterion further enhances the algorithm's effectiveness. By
 Overview minimizing impurity, Random Forest maximizes the
The proposed methodology for PregBot encompasses the homogeneity of the resulting nodes, leading to more decisive
integration of Machine Learning (ML) and Natural Language and reliable predictions. This emphasis on purity ensures that
Processing (NLP) techniques to deliver personalized and each decision tree contributes meaningfully to the ensemble,
adaptive support to pregnant women. This approach leverages collectively striving towards optimal predictive outcomes.
user interaction data, health information, and feedback to
continuously refine and enhance the system's capabilities. By In practical terms, PregBot leverages the power of
employing a cyclical development process, PregBot iteratively Random Forest to analyze user data comprehensively and
learns from user engagements to provide more accurate, timely, anticipate their healthcare needs with precision. Whether
and relevant assistance. predicting pregnancy outcomes based on demographic factors,
lifestyle choices, or medical history, the algorithm's ensemble
 Data Collection and Preprocessing approach ensures robustness and reliability across diverse
Data collection is the foundational step in our scenarios.
methodology, involving the accumulation of user interactions,
health metrics, and feedback. Following collection, data  Natural Language Processing (NLP)
preprocessing is performed to ensure quality and consistency. NLP techniques are applied to analyze and generate
This includes anonymization for privacy, normalization of human-like responses to user queries. The core NLP model
health metrics, and text data cleaning for NLP analysis. used is BERT (Bidirectional Encoder Representations from
Sentiment analysis is applied to user feedback and interactions Transformers), which allows PregBot to understand the context
using the formula: Sentiment to understand user sentiment of user inquiries deeply. By processing text data through BERT,
towards PregBot. PregBot can extract user intent and relevant information,
enabling it to provide personalized and contextually appropriate
Score = (Positive Words - Negative Words) / Total Words, responses. The effectiveness of NLP in PregBot is evaluated
using metrics such as accuracy and F1 score, where; balancing
 Machine Learning Algorithms the precision and recall of the model's predictions.
Our methodology utilizes a variety of machine learning
algorithms to analyze user data and predict their needs. Key F1 = 2 * (precision * recall) / (precision + recall),
among these is the Random Forest algorithm, chosen for its
effectiveness in handling diverse datasets and providing reliable Unlike traditional NLP models that process text
predictions. The Random Forest is implemented as follows: for sequentially, BERT's bidirectional architecture enables it to
each decision tree in the forest, a random subset of features is capture nuanced linguistic nuances and dependencies from both
chosen to split the nodes and make decisions, using the preceding and succeeding words in a sentence. This deep
formula: Gini Impurity = 1 - sum(p_i^2) contextual understanding allows PregBot to discern the subtle
intricacies of user inquiries, extracting intent and pertinent
where p_i is the probability of an element being classified information with unparalleled precision.
to a particular class. This ensemble method enhances the
predictive performance and robustness of PregBot. Through the lens of BERT, PregBot engages in a
sophisticated dance of semantic analysis and inference,
Unlike individual decision trees prone to overfitting or unraveling the underlying meaning embedded within user
bias towards certain features, Random Forest mitigates such queries. Whether it's deciphering complex medical jargon,
risks by averaging the predictions of numerous trees, thus understanding colloquial expressions, or discerning subtle
achieving a more generalized and accurate outcome. nuances in language, BERT equips PregBot with the linguistic
prowess to navigate diverse communication styles and contexts
The strength of Random Forest lies not only in its effectively.
predictive performance but also in its inherent ability to handle
high-dimensional data with ease. By randomly selecting subsets Furthermore, the efficacy of PregBot's NLP capabilities is
of features for each tree, the algorithm ensures that no single rigorously evaluated using established metrics such as accuracy
feature dominates the decision-making process. This feature and F1 score. The F1 score, a harmonic mean of precision and
selection strategy not only enhances prediction accuracy but recall, provides a comprehensive assessment of the model's
also guards against the curse of dimensionality, a common performance, striking a delicate balance between the
challenge in machine learning. completeness and correctness of its predictions. By
meticulously tuning the model parameters and fine-tuning on
domain-specific datasets, PregBot strives to optimize its F1

IJISRT24APR697 www.ijisrt.com 525

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

score, ensuring that its responses are not only accurate but also Pregbot's front-end offers a user-friendly interface for
contextually relevant and actionable. pregnant women to interact with the chatbot effortlessly. With
intuitive components like input forms and buttons, users can
The evaluation process entails rigorous testing against navigate and input queries seamlessly, enhancing their
diverse datasets encompassing a wide array of user queries and experience.
scenarios. Through iterative refinement and validation, PregBot
continually hones its NLP prowess, adapting to evolving On the back-end, the server and database components
language patterns and user needs. This relentless pursuit of handle critical tasks such as data processing, user
excellence underscores PregBot's commitment to delivering authentication, and session tracking. This ensures secure data
personalized and contextually appropriate responses, thereby handling and facilitates personalized responses using ML and
fostering trust and confidence among its users. NLP techniques, ensuring reliability and scalability.

Pregbot incorporates various techniques and libraries to Pregbot enriches its functionality by integrating with
create a functional chatbot system. Natural Language external services and APIs, providing access to additional
Processing (NLP) tasks are handled using the NLTK library, resources and enhancing communication between front-end and
including tokenization and lemmatization. The chatbot employs back-end components. Integration with services like Google
a Bag-of-Words (BoW) model to convert text inputs into Maps API enables location-based features, improving user
numerical vectors, which are then used as input to a neural experience.
network model trained with TensorFlow/Keras. This model
predicts the intent of user messages and generates appropriate Pregbot continuously refines its ML models based on user
responses based on predefined intents and responses stored in a feedback, ensuring accuracy, responsiveness, and relevance to
JSON file. SQLAlchemy is utilized for database interaction, user queries. This iterative process drives continuous
enabling the application to store and retrieve user data and improvement, evolving pregbot to meet the changing needs and
preferences. HTML templates and Flask's routing capabilities preferences of its users effectively.
are used for creating different views and handling user
interactions, including form submissions. Additionally, the Support features and community resources play a crucial
application features user authentication, session management, role in pregbot's value proposition, fostering engagement and
and flash messages for providing feedback to users. Overall, the collaboration among users. By offering forums, instant
code demonstrates the integration of NLP, machine learning, messaging, and managing relevant content, pregbot establishes
and web development techniques to build a chatbot-enabled itself as a comprehensive platform for pregnancy-related
web application. information and support, contributing to user satisfaction and
retention.

V. RESULTS AND DISCUSSION

The results from the implementation of PregBot revealed

significant insights into user engagement and satisfaction. The
following figures illustrate the distribution of user responses to
the system's functionality and the overall experience. The
feedback was collected via a comprehensive survey that
encompassed various aspects of the system's impact on prenatal
care.

Fig 1: Block Diagram of Pregbot Architecture and Working

IJISRT24APR697 www.ijisrt.com 526

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

process of entering and updating their health details as a

standout feature.

Fig 2: Survey Response Distribution.

Fig 4: Taking Inputs from the Patient before Interacting with
PregBot

Patient monitoring through PregBot demonstrated a

significant advancement in user engagement with health
management during pregnancy. The platform's ability to track
vital health metrics and provide contextual advice allowed for a
proactive approach to healthcare. Automated reminders for
medication intake, appointments, and personalized health tips
contributed to an enriched patient experience. The system's
monitoring capabilities extended beyond static data collection,
incorporating real-time updates and adaptive responses based
on user inputs. This dynamic monitoring system not only
ensured ongoing patient engagement but also allowed for the
early detection of potential health issues, facilitating timely
intervention.

Fig 3: Mean Scores for User Engagement Metrics.

In our analysis, the PregBot system's user interface (UI)

emerged as a crucial factor in its adoption and efficacy. The UI
was meticulously designed to be user-friendly, ensuring that
even individuals with minimal technical expertise could
navigate it with ease. Initial user feedback indicated that the
form-filling process, an integral component of data entry for
personalized care, was straightforward and non-intimidating.
Features such as drop-down menus for age and pregnancy
month, check boxes for health conditions, and simple yes/no
toggles for service opt-ins were both intuitive and accessible.
This simplicity in design was purposeful, adhering to principles
that reduce cognitive load and enhance user experience. As Fig 5: PregBot Interacting with the Patients.
shown in the survey responses, users reported a high level of
satisfaction with the interface, highlighting the seamless

IJISRT24APR697 www.ijisrt.com 527

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

Interactions with PregBot were augmented by the The project's outcomes encourage ongoing research and
integration of a sophisticated chatbot, powered by NLP development efforts to fine-tune AI applications within
algorithms that simulated empathetic and informative healthcare. With each iteration and improvement, we move
conversations. The chatbot effectively guided users through closer to a world where technology and human expertise
various scenarios, from locating nearby healthcare facilities to combine seamlessly to ensure the best possible care for every
providing emergency instructions. The use of conversational AI patient. PregBot is not just a culmination of this project but a
transformed the monitoring process into an interactive stepping stone towards the future of digital health solutions.
experience, reinforcing users' confidence and trust in the
system. The sentiment analysis of chat logs revealed a strong REFERENCES
positive reception, with users often expressing gratitude for the
chatbot's round-the-clock availability and supportive [1]. Lucia Vaira and colleagues investigate a system called
communication style. These results underscore the value of MamaBot that employs Machine Learning (ML) and
integrating empathetic AI in digital health platforms, paving the Natural Language Processing (NLP) for assisting pregnant
way for systems that not only monitor but also meaningfully women and their families. This work is documented in the
engage with patients. proceedings of the 22nd International Database
Engineering & Applications Symposium, 2018.
The interpretation of survey results underscores the [2]. Kevin Mugoye, Henry Okoyo, and Sylvester Mcoyowo
overwhelmingly positive reception of PregBot, as depicted in delve into the role of conversational agents in maternal
Figure 1, where the majority of users express agreement with its healthcare through "Smart-bot technology," as detailed in
benefits. Figure 2, illustrating mean scores of key performance the proceedings of the 2019 IST-Africa Week Conference,
indicators, reinforces this sentiment by indicating a high level IEEE.
of trust and perceived efficacy in the system's functionality. [3]. A study on the practicality of chatbots during pregnancy,
These findings suggest that the integration of machine learning focusing on user experience, was conducted by Joao Luis
(ML) and natural language processing (NLP) technologies in Zeni Montenegro, Cristiano Andre da Costa, and Luisa
pregnancy support tools can profoundly enhance user Placido Janssen, published in Healthcare Analytics, vol. 2,
experience and engagement. Users' satisfaction and the positive 2022, article number 100072.
perception of PregBot's performance imply its potential to [4]. A comprehensive review on the applications of AI during
revolutionize the way individuals interact with pregnancy- pregnancy is presented by Andreea M. Oprescu and her
related information and support systems. Such advancements team in IEEE Access, volume 8, 2020, pages 181450-
not only reflect technological innovation but also signify a 181484.
significant leap in addressing the nuanced needs and concerns [5]. In "Informatics," vol. 9, no. 4, 2022, Indriana Widya
of expectant individuals, paving the way for more personalized Puspitasari et al. describe the creation of a chatbot for
and effective healthcare solutions. expectant mothers on the Posyandu Application in
Indonesia, adopting a qualitative approach leading to a
VI. CONCLUSION decision tree method.
[6]. Iuliana Marin and Nicolae Goga's contribution to
The integration of artificial intelligence and machine EDULEARN19 Proceedings by IATED, 2019, involves a
learning in maternal healthcare, as exemplified by PregBot, has chatbot offering guidance on preeclampsia to pregnant
shown considerable promise in improving the experiences and women.
outcomes of pregnant women. Throughout the project, PregBot [7]. R. Wang, J. Wang, Y. Liao, and J. Wang explore chatbots
has demonstrated its potential to streamline the monitoring enhanced with Supervised Machine Learning for perinatal
process, provide immediate and personalized support, and mental healthcare at the ICHCI 2020 in Sanya, China, pp.
empower patients with timely information. The user-friendly 378-383, with the DOI of
interface, along with the intelligent chatbot system, has 10.1109/ICHCI51889.2020.00086.
significantly improved user engagement, offering an accessible [8]. As detailed in the 2019 IEEE SPICSCON in Dhaka,
and supportive digital health environment. Bangladesh, M. Umme Kulsum, A. Umara Najib, and M.
E. Hossain present "Design And Development Of MGA
As we look to the future, the insights garnered from this System," a mobile application providing guidance to
project serve as a foundation for further advancements in the expecting mothers, DOI:
field of AI in healthcare. The success of PregBot paves the way 10.1109/SPICSCON48833.2019.9064867.
for the development of more sophisticated systems that can [9]. Sanaa Suharwardy and colleagues assess the feasibility of
cater to an even broader spectrum of healthcare needs. It is a mental health chatbot in postpartum care in "AJOG
anticipated that continued innovation and research will lead to Global Reports," 2023, article 100165.
AI becoming an indispensable element in the delivery of
personalized, empathetic, and efficient healthcare services.

IJISRT24APR697 www.ijisrt.com 528

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR697

[10]. The development and usability evaluation of a chatbot for [21]. Kyungmi Chung, Hee Young Cho, and Jin Young Park
obstetric and mental healthcare for perinatal women and study a chatbot's development for obstetric and mental
their partners are explored by Kyungmi Chung, Hee health care for perinatal women and partners in "JMIR
Young Cho, and Jin Young Park in "JMIR Medical Medical Informatics," vol. 9, no. 3, 2021, article e18607.
Informatics," vol. 9, no. 3, 2021, article e18607. [22]. An improved Deep Recurrent Neural Network for stress
[11]. A 2022 IEEE ICECOCS conference paper by A. E. A. detection in pregnant working women is discussed by
Kaneho, N. Zrira, P. L. Bokonda, and K. Ouazzani- Sameer Dev Sharma et al. in "Electronics," vol. 11, no. 18,
Touhami surveys existing chatbots in the field of 2022, article 2862.
healthcare for pregnant women, Fez, Morocco, pp. 1-6, [23]. A Plos one paper by Ali Raza et al., vol. 17, no. 11, 2022,
DOI: 10.1109/ICECOCS55148.2022.9983388. e0276525, describes an ensemble learning-based approach
[12]. The user-centered design approach for a maternal for analyzing maternal health and predicting risks.
monitoring chatbot system using NLP is discussed by S. [24]. Tarun Lalwani and his team outline the development of a
H. Afrizal, N. Hakiem, A. Erna Permanasari, H. Syaifullah chatbot system using AI and NLP in "IJIRCST," vol. 6,
Albab, G. Yoki Sanjaya, and L. Lazuardi at the 2022 issue 3, 2018.
ICIMCIS in Jakarta, Indonesia, pp. 244-248, DOI: [25]. Binggui Zhou and associates discuss the role of NLP in
10.1109/ICIMCIS56303.2022.10017517. smart healthcare in "IEEE Reviews in Biomedical
[13]. S. S. Sadavarte and E. Bodanese highlight the use of Engineering," 2022.
Amazon Web Services and Alexa in a pregnancy [26]. Ayleen Bertini et al. review machine learning's role in
companion chatbot at the 2019 IEEE PuneCon, Pune, predicting pregnancy complications in "Frontiers in
India, pp. 1-5, DOI: bioengineering and biotechnology," vol. 9, 2022, article
10.1109/PuneCon46936.2019.9105762. 780389.
[14]. P. Arunkumar and associates introduce a machine learning [27]. "Briefings in bioinformatics," vol. 22, no. 5, 2021, article
application to promote personal health for women at the bbaa369, contains Lena Davidson and Mary Regina
2023 ICSSIT in Tirunelveli, India, pp. 908-914, DOI: Boland's review on AI and ML methods for enhancing
10.1109/ICSSIT55814.2023.10061126. [15] Tumpa, pregnancy outcomes.
Sanjida Nasreen, Anika Binte Islam, and Md Tasnim [28]. Evgeny Tebenkov and Igor Prokhorov discuss machine
Manzur Ankon. "Smart care: An intelligent assistant for learning algorithms for AI chatbots in "Procedia Computer
pregnant mothers." In 2017 4th International Conference Science," vol. 190, 2021, pages 735-744.
on Advances in Electrical Engineering (ICAEE), pp. 754- [29]. M. Krishnaveni et al. present a speech recognition module
759. IEEE, 2017. for patient monitoring in "REPS-GIE," 2018, IEEE.
[15]. "Smart care," a system providing intelligent assistance to [30]. A review on AI methods for maternal and neonatal health
expecting mothers, is presented by Sanjida Nasreen in low-resource settings by Misaal Khan et al. is featured
Tumpa, Anika Binte Islam, and Md Tasnim Manzur in "Frontiers in Public Health," vol. 10, 2022, article
Ankon at the 2017 ICAEE, pp. 754-759, IEEE. 880034.
[16]. Rocio Olmedo-Requena and her team examine factors that
influence adherence to nutritional guidelines before and
during pregnancy in "Women & Health," vol. 58, no. 10,
2018, pages 1094-1111.
[17]. Jingwen Zhang and her colleagues discuss the design of
AI chatbots for behavior change in promoting physical
activity and healthy eating in the "Journal of medical
Internet research," vol. 22, no. 9, 2020, article e22845.
[18]. In "Computer methods and programs in biomedicine," vol.
134, 2016, Mariam Bachiri, Ali Idri, José Luis Fernández-
Alemán, and Ambrosio Toval analyze mobile personal
health records for pregnancy monitoring.
[19]. The influence of mobile app interventions on maternal
behavior and perinatal health outcomes is reviewed by
Lisa M. Daly et al. in "JMIR mHealth and uHealth," vol.
6, no. 8, 2018, e10012.
[20]. Gabriela Frid, Kelly Bogaert, and Katherine T. Chen
evaluate mobile health apps for pregnant women in the
"Journal of Medical Internet Research," vol. 23, no. 10,
2021, article e25667.

IJISRT24APR697 www.ijisrt.com 529