Special Issue(24-19): Predictive AI Models for Patient Readmission Risk and Resource Allocation

Predictive AI Models for Patient Readmission Risk and Resource Allocation

One of the most expensive issue confronting healthcare systems is hospital readmissions, which are usually reliably predicted by traditional models. A lot of the approaches that are currently in use only employ expensive, dataset-specific manually-engineered parameters. The goal was to create and assess algorithms that use equipment for algorithms learning that autonomously derived characteristics from temporal data in order to anticipate hospital readmissions. Neural network approaches to acquire the vectors of features from everybody's cumulative healthcare record. These records span a minimum of several generations before the patient's initial hospitalisation and are sourced from a variety of sources. Each patient has a significantly different quantity of these information. By allocating time to the appropriate tasks, a team with a fair resource utilisation rate makes sure that the endeavour as a whole move ahead. Because it provides you with a more comprehensive view of your workers' burden, resource utilisation is particularly crucial if your staff members are engaged on numerous assignments. By starting to schedule individuals for the mission with utilisation understanding, you may prevent employee exhaustion and excessive utilisation by having a real-time insight of assignments, processes, and resource allocation.

It is possible to effectively target healthcare efforts to lower the risk of readmission by predicting if an arriving patient is at high risk. In the era of massive data, we can identify the risk of rehabilitation by using machine learning to evaluate a cohort of factors.  The lack of simplicity is one of the barriers preventing artificial intelligence from being widely used in healthcare. In addition to that, it offers many ways to comprehend the rationale behind the model's assumptions, ranging from the overall collection at global scale to the level of individual patient sightings. These justifications raise trust in the model's ability to make decisions. An algorithm for scheduling is a collection of guidelines that establishes that has to be done when. Numerous transmission rescheduling methods have been presented in the literature; however, their design is hindered by several factors such as technical intricacy, integrity, and the necessity to handle varying service levels. In communication networks, planning and resource allocation are crucial elements. The issue of predicting which users will be online within a specific time window is referred to as scheduling. The issue of distributing tangible layer resources, such lighting and bandwidth, among several active users is recognised as resource allocation.

Papers are invited that consider, but are not limited to, the following themes:

• Challenges to developing a synthetic knowledge model for impulsive readmissions.
• Predicting readmissions using deeply contextualised clinical concept embedding.
• Probability of postoperative Intensive Care hospital readmission using neural networks.
• Immediate care predictive modelling: an analysis of machine learning techniques.
• Integrating electronic health information to enhance hospital readmission predictions using predictive analytics.
• Predicting ICU returning by data mining with clinical physiology on evacuation.
• Using several models to forecast patient readmissions in critical care units.
• Estimating patient readmissions in medical facilities using multiple models.
• Predicting readmissions to hospitals using persistent mobile data sources.
• Monitoring hospital readmissions with assistance vector machines powered by smart swarms.
• A modified multiple-phase classification approach to evaluate the risk of patient readmission.
• Advantages and obstacles in implementing predictive modelling in computerised health care.

Guest Editor Information:

Dr. Engr. Abdul Kadar Muhammad Masum

Professor,

Department of Software Engineering,

Faculty of Science and Information Technology,

Daffodil International University, 

Birulia, Savar, Dhaka –1216, Bangladesh 

Official Email: [email protected] 

Personal Webpage: https://faculty.daffodilvarsity.edu.bd/profile/swe/kadar.html 

Scholar Webpage: https://scholar.google.com/citations?user=SOZ8uAEAAAAJ&hl=en 

Research Areas: Machine Learning, Business Informatics, Health Informatics, ICT applications in Business, Education and Governance

Dr. Md Zia Uddin

Senior Research Scientist,

Sustainable Communication Technologies Department,

SINTEF Digital, Oslo, Norway

Official Email: [email protected] 

Personal Webpage: https://sites.google.com/site/webpagezia/ 

Scholar Webpage: https://scholar.google.com/citations?user=YFu7iB8AAAAJ&hl=en 

Research Areas: Image Processing, Sensing, Robotics, Computer Vision, Smart Home Healthcare, Artificial Intelligence, Machine Learning, Pattern Recognition

Dr. Md Abdus Samad

Professor,

Department of Information and Communication Engineering, 

Yeungnam University, 

Gyeongsan 38541, South Korea

Official Email: [email protected] 

Personal Webpage: https://www.drabdus.info/ 

Scholar Webpage: https://scholar.google.com/citations?user=NmUWfQIAAAAJ&hl=en 

Research Areas: Data Science, Wireless communication, RF propagation, Signal Processing, Machine Learning

Tentative Timeline: 

• Deadline for manuscript submissions will be February  20, 2025. (Revised)
• Expected publication date will be based on Journal decision