Date of Award

Fall 2021

Degree Type

Restricted to Claremont Colleges Dissertation

Degree Name

Philosophy, PhD

Program

Center for Information Systems and Technology

Advisor/Supervisor/Committee Chair

Samir Chatterjee

Dissertation or Thesis Committee Member

Jessica Clague DeHart

Dissertation or Thesis Committee Member

Wallace Chipidza

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2021 Balakrishnan Mullachery

Keywords

AI, Deep Learning, GIS, Healthcare, Machine Learning, NLP

Abstract

Improving efficiency in care management was one of the most discussed subjects during the Covid-19 pandemic. Many transdisciplinary studies have been progressing in distinct aspects of care management. For example, telemedicine and remote patient monitoring systems for chronic patients. Most of these studies concentrate on a care-centric remote monitoring approach for active ailing chronic patients rather than a patient-centric approach for patients who are in self-care leading an Active Daily Life (ADL). The above studies' primary research data sources were hospitals, treatment-related stakeholders, and telemedicine infrastructure. The number of chronic disease cases are increasing rapidly, and their life expectancy also increasing due to medical advancement. Chronic patients require lifelong care services to manage their disease and it is critically important to have better overall mental and physical health. Affordable healthcare facilities, however, have insufficient resources. This study focused on the environment where chronic cancer and cardiovascular (CVD) disease patients live, considered geography a significant construct in studying chronic patients, and identified the significant factors that positively or negatively influence their ADL. These ADL measuring factors help to develop a system that can keep and monitor a chronic patient at home and reduce care center utilization to ease care management. The primary data of patients are from home, and these temporal data can speak more about the ADL, well-being, and risks of chronic patients, hence the care management. This study discusses two domain areas: The problem space of healthcare, more specifically the chronic patients care and disease management, and the solution space of Information and Communication Technology. Researchers are attempting to find an acceptable solution for this growing problem of frequent medical service requirements for chronic patients. One solution for this growing problem is to keep chronic patients at home by providing prolonged and efficient disease management, either through caregivers or self-management. However, we must identify how to provide sustainable care and security to these Chronic Patients at home and avoid hospitalization or unnecessary medical services. This approach can help the overall care management process. Unlocking the power of mobile internet helps connect people to people and people to the industries for new opportunities and services. Advancement in communications technology such as GIS, IoT, Social media, AI, and pervasiveness of online connections enable people to access information and services to leverage shared information and enrich their knowledge. This transdisciplinary study is designed to learn and understand chronic cancer and CVD patients' activities of daily life during their self-care at home and find patterns that are useful to determine their well-being status using technologies. Any risk associated with a patient can be analyzed from their home-based sequenced daily living data and provide necessary guidance to manage the disease at home. During this exploratory study, a software prototype was developed to create a non-clinical dataset using technologies such as IoT, mobile app, and GIS that are related to cancer and CVD patients during their self-care at home. These temporal data were analyzed using multiple AI methods including multiple linear regression, deep learning, NLP, clustering, and classifications. This analysis provides a lot of insights in terms of features that influence a chronic patient from their ADL and are documented in the dissertation. This study provides foundational knowledge on how a simplified patient-centric disease management system can help avoid unnecessary medical services and keep patients at home with safety and security. The novelty and implication of this research are manifold and are discussed in detail in the discussion and contribution chapters. Integrating sensor data (IoT), patient self-reported non-clinical data, patient geolocation, and environmental spatial data to analyze chronic patients' health using AI has never been done. This study provides insights into and validated the usefulness of AI and text mining models in the healthcare industry for studying chronic patients at home using temporal non-clinical data. The AI models developed in this study are a contribution to the body of knowledge in this under-explored topic. GIS and location analytics are widely used in scientific research, engineering, business, real estate, and health informatics. However, there is limited research conducted in medical diagnoses or behavioral health studies using GIS. This study contributes to the medical science body of knowledge concerning how GIS can be used to analyze chronic patients’ health conditions (medical diagnostics) and well-being (health behavior) at home during self-care. The other major implications are the importance of the locality construct of health geography in designing any ICT platform for care and disease management and rethinking the physical hospital space.

ISBN

9798759991694

Download

Share

COinS