International Conference on Innovative Data Communication Technologies and Application (ICIDCA-2023)

IEEE Xplore Part Number: CFP23CR5-ART; ISBN: 979-8-3503-9720-8

Framework For Predicting Suicidal Attempts Using

Healthcare Data and Artificial Intelligence
2023 International Conference on Innovative Data Communication Technologies and Application (ICIDCA) | 979-8-3503-9720-8/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICIDCA56705.2023.10099967

1st Pradumn Saxena 2nd Sandeep Prabhu

Symbiosis Institute of Digital and Telecom Management) Symbiosis Institute of Digital and Telecom Management)
of Aff. Symbiosis International (Deemed University) of Aff. Symbiosis International (Deemed University)
Pune, India Pune, India
pradumn.saxena2123@[Link] sprabhu@[Link]

Abstract—The main objective of this research paper is to data on paper in healthcare organizations [4]. By properly
analyse and define a framework of Artificial Intelligence (AI) for applying machine learning algorithms on patients data, health-
healthcare. The AI enabled framework based on hospital health care professionals can quickly and easily go to some decisions
data is used to identify hospital’s current or past patients who
are likely to attempt suicide. The study is important for its social and make future projections on basis of pattern recognition.
aspect. AI on the healthcare has not received much attention in Artificial intelligence therefore plays a significant part in the
research. This research will assist healthcare providers, govern- planning and administration of healthcare. For example, auto-
ment entities in developing social welfare solutions. Since AI is matically reminding patients to renew their prescriptions help-
viewed as the technology of the future, extensive and methodical ing them in remembering to take their medications. Despite
research is necessary. In order to fill this gap, this study proposes
a framework of artificial intelligence based on healthcare data. the significant rise of artificial intelligence in healthcare there
This framework application may lead to the development of are few limitations related to skilled AI talent, cost-intensive
creative solutions for the healthcare sector. Thus, the objective of implications, patient data management and privacy [5]. There
the paper is to highlight several AI applications for healthcare are few countries where access to proper healthcare system
to predict suicidal attempts. is still lacking which has been a major issue for healthcare
Index Terms—Healthcare professionals, Diseases, Healthcare,
Artificial Intelligence, Suicidal attempts services. So, in order to avoid or overcome these issues,
there is a requirement on having a better understanding of the
different Artificial Intelligence applications in the healthcare
I. I NTRODUCTION
area. There are many literatures which shows how important
Artificial Intelligence refers to intelligence shown by ma- is Artificial intelligence in healthcare. This research paper
chines to do human tasks in day-to-day life. The term “Ar- focuses on showcasing artificial intelligence applications that
tificial Intelligence” was created by John McCarthy which can benefit the healthcare sector in various ways, ultimately
defines AI as the engineering and science of building creative raising service quality and proposing a framework by utilizing
devices [1]. AI is coming out to be an interesting area in AI and healthcare data to predict suicidal attempts.
computer technology as it has amplified human existence
in various sectors. It has been applied in many sectors like II. L ITERATURE R EVIEW
education, business, and engineering. It is the capability of The most recent method for dealing with the compli-
machine learning algorithms to provide nearest conclusions cated problems encountered in the medical field is artificial
based on input data. With the recent digitalization and lot of intelligence-based medical support. AI has more intelligent
data acquisition, artificial intelligence is touching those areas methods for handling medical problems [6]. Looking at a
which were untouched by the human experts and one of those number of academic papers that have been published may help
areas is healthcare where it has shown a lot of capabilities in we acquire a better understanding the usefulness of artificial
prevention, diagnosing, and detecting wide range of diseases intelligence in the healthcare sector [7]. India is working
[2]. It has been discovered that there is a significant need towards boosting an AI platform for integration of data to
to expand the medical services in rural areas. Rural areas achieve desired insights. Due to indeterminate evidence in their
continue to struggle to access high-quality medical facilities. routine practice, physicians must continue to make frequent
One technology that can help solve these issues is artificial judgements. Cognitive bias, or the difficulty of evaluating a
intelligence, which is shown excellent growth by offering number of possibilities under a certain circumstance, another
high-quality facilities all around the world. The market size problem in the context of medical decision-making noted by
of artificial intelligence in healthcare is expected to reach Redelmeier and Shafir [8]. IBM Watson, an AI tool, is em-
USD 187.95 billion by 2030 [3]. The art of taking quick and ployed in oncology and is assisting the medical community in
good decision in Healthcare has become simple with digitally the treatment of diseases like cancer. Patients with cancer are
storing data and Artificial intelligence is used to analyse the benefiting from tailored treatments with help of IBM Watson.
extracted data in contrast to the traditional method of storing With its array of oncology solutions and its ability to support

979-8-3503-9720-8/23/$31.00 ©2023 IEEE 1085

Authorized licensed use limited to: VTU Consortium. Downloaded on November 23,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 International Conference on Innovative Data Communication Technologies and Application (ICIDCA-2023)
IEEE Xplore Part Number: CFP23CR5-ART; ISBN: 979-8-3503-9720-8

informed clinical decision-making with regard to cancer, IBM

Watson has had an impact on the healthcare sector [9]. AI
frequently uses different algorithms to understand features and
then delivers insights to help experts or medical professionals
make decisions. To help physicians provide effective patient
care, an AI system has been utilised to provide current medical
knowledge from journals, textbooks, and clinical practices.
AI may self-correct mistakes and keep learning based on
user feedback. AI is employed for a variety of functions,
including treatment, outcome, and recognition [10]. Lopez
and colleagues investigate the use of artificial intelligence
in medical applications in their study focused on web-based
medical diagnosis and prediction [11]. In their work, Langlotz
and co-authors talk about how AI is being applied in medical
imaging to enhance patient outcomes [12]. In this research
paper, we’ll examine many applications of AI in the healthcare
sector and how they impact it. Various use cases will be
examined with an emphasis on decision-making and improved
patient service.

III. R ESEARCH M ETHODOLOGY

This research is a qualitative study using expert in-depth

interviews. Five computer science experts in the field of
Artificial Intelligence and Machine Learning (AI-ML), five
hospital management experts, and 15 customers of hospital Fig. 1. Research Methodology
were interviewed. The transcript creation followed by content
analysis was carried out to understand issues and gaps in
expected hospital services. Based on the content analysis a IV. D ISCUSSION AND A NALYSIS
framework was created. The final framework is then validated A. Artificial Intelligence Applications in Healthcare
by 3 independent AI experts. At second stage framework is
validation based on a case study methodology, in which ex- This segment focuses on various artificial intelligence appli-
tensive details of numerous instances are used with a primary cations in healthcare sector and AI has shown remarkable im-
focus on the healthcare sector. provements in this sector and continues to show improvement.
Emergence of AI will show promising signs for rural areas
The objective of second stage of validation was to determine
in near future. There have been studies on AI applications
the function of artificial intelligence in healthcare and its
those have raised service standards in the healthcare industry.
effect on service quality. The data in this research study is
In many healthcare-related disciplines, proper analysis is still
collated from various sources which includes reports, articles,
required.
and research papers. The case study approach, according to
Swanborn, is a detailed examination and analysis to look 1) Chatbots: Many hospitals and healthcare organizations
into numerous outcomes and predict various future behaviors, have begun to use Chatbots powered by artificial intelligence
to identify upcoming trends and to address any lingering for patient care through the use of websites. These chatbots
problems in order to resolve them and arrive at a clearer, help in making critical health information accessible to the
conclusive knowledge [13]. right authorities at right time. A survey by Salesforce Inc.,
shows 86
The following research questions are addressed in this study:
• Regulate Updates provided to patients: Chatbot is very
• What type of data analysis applications are used in the essential and beneficial for patients who tend to send
medical industry to improve service quality? regular messages and mails regarding appointments.
• What are capabilities of Artificial intelligence in the • Management of Patient Appointments: Chatbots can send
medical field? automated reminder to patients regarding medicine sched-
• How can Artificial intelligence applications deployed in ules and collect updates about crucial health information.
medical industry set-up? • Support in other medical investigations: Chatbots can aid
• How Artificial intelligence and healthcare data can be in the prediction of disease, billing of disease in question,
deployed to predict suicidal attempts prediction? as well as retrieve and suggest insurance information.

979-8-3503-9720-8/23/$31.00 ©2023 IEEE 1086

Authorized licensed use limited to: VTU Consortium. Downloaded on November 23,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 International Conference on Innovative Data Communication Technologies and Application (ICIDCA-2023)
IEEE Xplore Part Number: CFP23CR5-ART; ISBN: 979-8-3503-9720-8

• Engage Patient interactions: A common problem faced by B. AI Use Cases in Healthcare

patients is that many therapists require past documents Numerous medical applications of AI that are previously
during patient visits during different appointments. There addressed are discussed in reference to a use case. This section
can be times when patient might forget to carry the explains how these initiatives have been effective and how they
reports. In such situations chatbots can be useful as they are resulting in a more effective healthcare system.
maintain patient data through the conversation or the last 1) Chatbot in Healthcare: Applications employing chatbots
calls. have been created by numerous healthcare start-ups and orga-
• Handle Frequent Queries: It has often observed that nizations to help patients in various ways.
medical support team is too busy in query handling. At
• Cancer Chatbot: The goal of CSource’s Cancer Chatbot is
times the medical team might miss out on some patients.
Under such situation chatbots can be help by saving time to assist patients in seeing early indications of the disease
of the patient and the support team [15]. in order to treat it effectively. Patients can access the
support they need from their mobile devices thanks to
Chatbots powered by AI can serve as an agent to deal with a wide database of cancer treatment organizations and
mental health difficulties. It can serve as a coach, helping those services [18].
who are lonely and find it difficult to understand and navigate • Synapse Medicine: This company takes help of chatbot
the world. AI has ability to have cognitive behavioral therapy in providing drug prescription. It also provides essential
(CBT), a type of psychological care. Using CBT patients are information related to healthcare professionals, hospitals,
given some questionnaires, and the questionnaire is analyzed healthcare facilities [19].
to determine the patients’ levels of stress and anxiety under • Woebot Health: This company provides a chatbot for
different situations. mental health., adults, maternal mental health. The bot
2) Early Diagnosis and Test Analysis: Various medical enables the users to manage stress and anxiety as well as
scans, including CT scans and X-rays, are being read by depression, loneliness, procrastination [20].
applications based on AI. These algorithms are used to detect • Dr. Lal Path Labs Virtual Assistant: The chatbot from
many diseases at an early stage. AI has the ability to digitize the path labs help customers with covid-19 related infor-
data obtained from various samples that pathologists have mation, track reports etc. It also tells the various tests
taken. Additionally, testing and scanning may benefit. In order pricing.
to deliver a diagnosis based on the test results or scan results,
many AI techniques are being applied. AI decision-making C. Early Diagnosis and Test Analysis in Healthcare
will assist medical professionals in moving on with their Many organizations are using artificial intelligence exten-
procedures. Since early cancer detection can help millions sively to make an early diagnosis for conditions such as cancer
of individuals every year, cancer screening is a crucial area by analysing a wide range of scans and tests associated to
where AI has a lot to offer. The machine learning component many diseases using AI.
of artificial intelligence is utilized to help pathologists make • Oncostem Diagnostics: This company employs AI to
accurate diagnoses. It is simple to determine whether any find trends related to the breast cancer risk. To produce
genes can cause cancer by analyzing biological data such in-depth understandings and knowledge for the doctors,
as RNA sequencing and DNA methylation using AI-based analyze historical data and feed it into the system [21].
technologies [16]. • Royal Bolton Hospital: Since the Covid pandemic, this
3) Robotic Surgeries: In the form of cooperative robots, AI hospital has been using an AI-based X-ray system called
technology has revolutionized the surgical field. The revolution qXR. Utilizing AI to interpret X-rays could help save
is usually noted when creating minor incisions in terms adequate time by providing early diagnosis, improving
of its depth and rapidity. Surgery outcomes, particularly a patient care, and assisting doctors in making decisions.
replacement or complicated procedure, can vary according on • Medicircle: Covid testing using artificial intelligence
the surgeons’ ability. AI utilization can reduce case-to-case called Spectral Instant Test was created by an Indian start-
variances while simultaneously increasing the effectiveness up (SpectraLIT). To assess whether the virus’s spectral
of even the most expert surgeons. Because AI machines pattern as seen in a nasal swab and SARS-CoV-2 are
are so exact, there is no chance that they will wobble or comparable, spectral analysis is used.
make any unintentional motions while performing surgery.
For example, AI-controlled robots can perform fundamental D. Robotic Surgeries
tasks such as precise cutting and sewing while also working • Narayana Health: Hospital based out in India has paved
with greater accuracy and miniaturization. Surgeons at the the way for advanced healthcare in terms of surgery
Maastricht University center in the Netherlands used robotic in India to provide accuracy and flexibility making the
surgery with AI assistance to close blood arteries those were process hassle-free. It is done by three robot arms and
only.03 to.08 mm wide. Additionally, AI systems can use a video camera inserted in the body through small cuts.
information from previous procedures to plan new surgical The surgeon then sits in front of computer and control
techniques [17]. the robotic arms.

979-8-3503-9720-8/23/$31.00 ©2023 IEEE 1087

Authorized licensed use limited to: VTU Consortium. Downloaded on November 23,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 International Conference on Innovative Data Communication Technologies and Application (ICIDCA-2023)
IEEE Xplore Part Number: CFP23CR5-ART; ISBN: 979-8-3503-9720-8

• Myomo: This company uses Myopro to encounter loss of family data is to be collected and updated subsequently. AI-
motion or weakness in the hands and arms due to many ML algorithms such as Gradient Boosting Classification or
conditions. The device will read nerve signals from skin Deep Neural Network (DNN) is to be used for prediction of
surface and at those points activate or encourage natural future suicide attempts. The decision to be validated by outputs
arm or hand movements from chatbot discussions. Recurrent Neural Network (RNN)
the type of algorithm used for ChatGPT will validate the Deep
E. Conceptualization Neural Network outcome. Based on the validation the patient
Based on the earlier discussions, we can deduce a number to be connected and counselled.
of advantages that artificial intelligence implementation can
provide in the healthcare. These are:
• Reduced waiting time
• Easier scheduling of appointments
• Hassle free discharging
• Faster payment of bills
• Improved prognosis
• Faster therapies and surgery
• Economical treatment
• Good user-based experience
• Detection of disease outbreaks

AI implementation benefits not only the general public but

also health care practitioners, and hospitals. The information
gathered will aid them in producing results in quickly.

V. F RAMEWORK TO P REDICT S UICIDAL ATTEMPTS USING

AI AND H EALTHCARE DATA
Based on the content analysis of in-depth interviews of
experts in AI and medical domain and past hospital patients
with experience of hospitalization a framework is created.
The framework has the ability to predict suicidal attempts
where the first step will be acquiring the data from various
hospitals, health organizations, social media etc. The second
step is integrating the data and cleaning it so that data is in
structured format and can give effective results. Data analysis
will be the next step where the framework studies basic
trends from the data like which city has most number of
suicidal attempts, which age group committed most suicides,
which zone has most number of suicides, which is the most Fig. 2. Framework
frequently occurring cause of the suicide. The interactions and
suicide cases in the past can be analyzed by AI algorithms
like Random Forest, or Gradient Boosting Classifier. Even B. Limitations
Deep Neural Networks can be used for prediction of possible In spite of benefits and advancements which AI is pro-
suicide attempts. In addition for text analytics sequential neural viding to the healthcare industry there are still few issues
network models based on Recurrent Neural Network (RNN) or challenges which needs to be addressed such as big
are suggested. Data collected will be used to training some part data being generated on day-to-day basis around the world.
of dataset and once training is done, based on accuracy best-fit Another important aspect is data incomplete, noisy and can
model will be chosen for prediction of suicidal attempts and have compatibility issues with various other platforms due to
finally deploying that model in the real world for testing on interoperability. There are only 36
real scenarios.
C. Implications
A. Role of Artificial Intelligence In order to redesign and develop more effective and efficient
Considering data, collection at different stages of patient healthcare solutions, academics will be able to use artificial
interactions with the hospital network, structured data is to be intelligence technologies. The research will also help medical
maintained in unified data center. Past data of patients’ suicide institutions and hospitals learn more about AI technologies
attempts is also to be recorded by the hospitals. Admissions that can be used to improve services to the public. The
data, visits data, medication data, treatment data, financial data, globe has already made strides in the use of technology that

979-8-3503-9720-8/23/$31.00 ©2023 IEEE 1088

Authorized licensed use limited to: VTU Consortium. Downloaded on November 23,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 International Conference on Innovative Data Communication Technologies and Application (ICIDCA-2023)
IEEE Xplore Part Number: CFP23CR5-ART; ISBN: 979-8-3503-9720-8

can simplify the lives of individuals and, at the same time, [7] Accenture, “Rewire for Growth,”Business Journal for India., pp. 1–98,
lessen the absence of medical facilities for most, if not all, 2019, [Online]. Available: [Link] acnmedia/PDF-
112/[Link].
of them. Nearly everywhere in the world, high-speed internet [8] D A Redelmeier , and E Shafir, ” Medical decision making in situations
connectivity has improved with time, enabling individuals to that offer multiple alternatives,” JAMA: The Journal of the American
receive remote consultations, accelerate the diagnosis of any Medical Association, vol. 273, no. 4, January 1995.
[9] Joe Gagnon, “Watson Health is committed to using AI to tackle
illness, and prevent a number of diseases [23]. An intriguing major healthcare challenges - Watson Health Perspectives,” 2022.
technology with a bright future, AI has always fascinated [Link]
people. Researchers from nations with insufficient AI deploy- healths-challenges-tell-us-more-about-healthcare-data-than-it-does-
about-ai/?sh=705f39245b48 (accessed Aug. 20, 2022).
ment can continue their work to support the adoption of AI [10] S. Reddy, “Use of Artificial Intelligence in Healthcare Delivery,”
technology in the healthcare industry[24]. Future investigation eHealth - Mak. Heal. Care Smarter, no. July, 2018, doi: 10.5772/in-
is possible on how with help of AI doctors can predict kidney techopen.74714.
[11] F. Amato, A. López, E. M. Peña-Méndez, P. Vaňhara, A. Hampl, and
diseases, how AI can contribute to the research in cancer J. Havel, “Artificial neural networks in medical diagnosis,” J. Appl.
treatment mainly in radiation therapy, how AI can be used to Biomed., vol. 11, no. 2, pp. 47–58, 2013, doi: 10.2478/v10136-012-
accelerate the development and discovery of genetic disease 0031-x.
[12] C. P. Langlotz et al., “A roadmap for foundational research on artificial
[25]. Furthermore, future research can discuss how feasible intelligence in medical imaging: From the 2018 NIH/RSNA/ACR/The
these technologies are and how they could be game changers Academy workshop,” Radiology, vol. 291, no. 3, pp. 781–791, 2019,
in the healthcare sector [26]. doi: 10.1148/radiol.2019190613.
[13] P. Swanborn, “EBL Reader - Case Study Research: What, Why and
How?,” 2010. [Link]
VI. C ONCLUSION what-why-how.
[14] S. Bansal, “Healthcare Chatbot – Key Advantages and Use
AI in healthcare has various challenges in the form of no Cases,” 2022. [Link]
trust in the case of results delivered with the help of AI, but cases-of-healthcare-chatbot/
by improving the workflows and the operations, assist users [15] M. Bryant, “Hospitals turn to chatbots, AI for care — Health-
care Dive,” 2018. [Link]
in getting answers in quick time and development of new and healthcare/516047/
innovative treatments and therapies. [16] P. P. Ippolito, “Understanding Cancer using Machine Learning
The case study mentioned above satisfies the fact that it — by Pier Paolo Ippolito — Towards Data Science.”
[Link]
offers patients and the general public a number of advantages, learning-84087258ee18
including quicker appointments, convenience, rapid bill pay- [17] Compunnel Digital, “AI in Healthcare: Top 5 Real-World Ex-
ment leading to time savings, and amazing service quality. amples.” [Link]
real-world-examples/
However, it aids medical professionals in making decisions. [18] Y. Paul et al., Artificial Intelligence in the Healthcare Indus-
We can also draw the conclusion that it benefits hospi- try in India. 2018. [Online]. Available: [Link]
tals and other relevant institutions, as well as individuals, governance/files/ai-and-healtchare-report
[19] I. B. Abbasov, “Artificial intelligence in medical imaging,” J. Phys. Conf.
by enabling them to enjoy reduced prices, quicker payment Ser., vol. 2094, no. 3, 2021, doi: 10.1088/1742-6596/2094/3/032008.
alternatives, shorter wait times, error-free support, and higher [20] A. Esteva et al., “Deep learning-enabled medical computer vision,” npj
patient satisfaction. Digit. Med., vol. 4, no. 1, pp. 1–9, 2021, doi: 10.1038/s41746-020-
00376-2.
The framework proposed will help in predicting suicidal [21] N. Aayog, “Responsible Ai # Aiforall Part 2,” Niti Aayog, no. February,
attempts eventually saving lives and will help the health sector p. 8, 2021.
and healthcare professionals especially psychiatrist to make [22] K. Mirin, “Implementation of AI in Healthcare: Challenges and Poten-
tial,” 2021. [Link]
advancement in their domain by using artificial intelligence challenges-and-potential/
for suicidal attempts prediction. [23] H. Alami et al., “Organizational readiness for artificial intelligence in
health care: insights for decision-making and practice,” J. Health Organ.
R EFERENCES Manag., vol. 35, no. 1, pp. 106–114, 2021, doi: 10.1108/JHOM-03-2020-
0074.
[1] Topics, “Learn about popular topics in technology from IBM ex- [24] D. W. Bates et al., “The potential of artificial intelligence to improve
perts,” IBM. [Link] patient safety: a scoping review,” npj Digit. Med., vol. 4, no. 1, pp. 1–8,
intelligence#toc-types-of-a-q56lfpGa (accessed Mar 7, 2022). 2021, doi: 10.1038/s41746-021-00423-6.
[2] S. Agrebi and A. Larbi, “Use of artificial intelligence in infectious [25] R. V. Rodriguez, S. Sinha, and S. Tripathi, “Impact of Artificial
diseases,” Artificial Intelligence in Precision Health, pp. 415-438, 2020. Intelligence on the health protection scheme in India,” Public Adm.
[3] Topics, ” Artificial Intelligence (AI) in Healthcare Market,” Policy, vol. 23, no. 3, pp. 273–281, 2020, doi: 10.1108/PAP-03-2020-
precedenceresearch. [Link] 0019.
intelligence-in-healthcare-market (accessed Mar 7, 2023). [26] T. Sarwar et al., “The Secondary Use of Electronic Health Records
[4] S. R. Dutta, M. Sneha, S. Bagalkot, and M. Kolukuluri, “Artificial for Data Mining: Data Characteristics and Challenges,” ACM Comput.
Intelligence and Big Data Analytics in Healthcare for Predictive Analysis Surv., vol. 55, no. 2, 2023, doi: 10.1145/3490234.
and Future Prospects,” Oeconomia Copernicana, vol. 12, no. 4,pp. 30-45, [27] S. Syed and I. Amin, “Prediction of Suicide Causes in India using
2021. Machine Learning,” J. Indep. Stud. Res. - Comput., vol. 15, no. 2, 2017,
[5] F. Jiang et al., “Artificial intelligence in healthcare: Past, present and doi: 10.31645/jisrc/(2017).15.2.01.
future,” Stroke Vasc. Neurol., vol. 2, no. 4, pp. 230–243, 2017, doi:
10.1136/svn-2017-000101.
[6] Accenture, “Artificial Intelligence: Healthcare’s New Nervous
System,” Accent. Rep., pp. 1–8, 2017, [Online]. Available:
[Link]
growth.

979-8-3503-9720-8/23/$31.00 ©2023 IEEE 1089

Authorized licensed use limited to: VTU Consortium. Downloaded on November 23,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.