China
Africa
Explore chapters and articles related to this topic
Big Data Analytics in Healthcare Data Processing
Published in Punit Gupta, Dinesh Kumar Saini, Rohit Verma, Healthcare Solutions Using Machine Learning and Informatics, 2023
Tanveer Ahmed, Rishav Singh, Ritika Singh
Health insurance firms employ a variety of approaches to uncover fraudulent activities and develop different types of methods to prevent medical fraud. Companies use Hadoop apps that use data from earlier health claims, earnings, and demographics, among other data sources, to identify fraudsters. According to the authors [65], different technologies are also used in healthcare to detect fraud such as data mining and machine learning technologies.
Clinical and Business Intelligence
Published in Salvatore Volpe, Health Informatics, 2022
In data mining, the analyst asks a question, creates a profile for the mining tool based on that question, and then the software digs through the data looking for patterns that might provide clues to answer the question. Here is an example of how data mining might be applied in a healthcare setting. Consider a scenario in which there is a certain disease condition with divergent outcomes. One group of patients responds to treatment and recovers while another group seems to languish and not improve. Data mining could be used to profile the two sets of patients and find commonalities in the treatment regime, test results, patient history, genetics, or other elements that may represent hidden factors causing the difference in outcomes. If the sampling size is big enough, this technology can look at treatment sequencing and when interventions during the patient’s course of care are most effective.
Introduction
Published in Rui Nunes, Healthcare as a Universal Human Right, 2022
However, the existence of large collections of data, both structured and unstructured, also termed big data or even lake data, questions society and public authorities about the destiny of this large amount of information and how it can be used to make better decisions for mankind. Data mining, machine learning (including deep and reinforcement learning), machine reasoning, and robotics (integration of different techniques into cyber-physical systems) have been developed to extract information and transform these data into valuable use in the healthcare system. Artificial intelligence also has a profound impact in modern systems because of the capacity to integrate all this big data and propel it to promote new treatment modalities, besides preventing life-threatening diseases. Indeed, it may help physicians in the diagnosis and treatment of many diseases, and in the real-time monitoring of patients, sometimes at long distances.
Ontology-Based decision tree model for prediction of fatty liver diseases
Published in Computer Methods in Biomechanics and Biomedical Engineering, 2023
Seyed Yashar Banihashem, Saman Shishehchi
A large number of electronic medical data has not been evaluated using classification algorithms and ontology together. Usually, in related healthcare systems, researchers are employing either a data mining algorithm (Palaniappan and Awang 2008; Islam et al. 2020; Shuja et al. 2020) or ontology models. This study aims to achieve this objective by applying an ontology-based decision tree algorithm for 957 patients. Another reason for using ontology is that ontology-based clustering is more efficient and flexible than a typical decision tree clustering and also ontology based system can be reused in other system. For large volumes of data, ontology can deal with it (Mathew et al. 2014). Decision Tree plays the most crucial role in prediction by using the classification method. The primary output of the decision tree is if-then rules obtained from the dataset. Ontology will use these rules to check whether the patient suffers from fatty liver or not. The result of the ontology emphasized decision tree results. Additionally, having all rules in hand by the decision tree model help us to create a more accurate ontology.
Discovering hidden patterns among medicines prescribed to patients using Association Rule Mining Technique
Published in International Journal of Healthcare Management, 2023
The findings of our study warrant further application and development of data mining techniques for medical informatics. Despite positive trends observed, several topics need to be adjusted as well as special considerations may be taken, such as sample size, temporal changes in prescribed medicines, automatic associations with clinical outcomes, algorithms for extracting association rules in less computational time, incorporation of physicians’ judgement, and many others [43]. Future studies may focus on prescription data mining considering dynamics in the patient conditions and physicians’ prescribing behavior. We can investigate how different physicians may prescribe medicines for similar patient treatment cases. Similarly, patients with similar conditions may be prescribed differently by the same physicians. Apart from that, different drugs have different dosage strengths and multiple pharmaceutical suppliers and manufacturers. Future research studies can consider these factors.
Applications of monitoring and tracing the evolution of clustering solutions in dynamic datasets
Published in Journal of Applied Statistics, 2023
Muhammad Atif, Muhammad Shafiq, Friedrich Leisch
The clustering technique is a potent field in data mining and has wide applications in medical diagnostics, social paradigms, image processing, pattern recognition, and psychological sciences. It is an unsupervised learning technique that deals with the identification of hidden structures in unlabeled datasets. A cluster is a collection of data items having a high similarity index with each other, while preferably dissimilar to objects of other clusters. However, many real-world applications in recent years produce dynamic datasets that evolve over time, and consequently, the hidden structure is not stationary. As a result, the clustering solution over this stream is not stationary and changes with the influx of new data items. In this paper, a detailed literature review about various algorithms incorporated for monitoring and tracking the evolution of clustering solutions over time is presented. The applications of monitoring clustering evolution in ‘media use and trust issues’, ‘air quality of Bowen’, and ‘crime against women in India’ are highlighted with the help of real-world published datasets. It is quite evident from these examples that monitoring and tracking changes in a dynamic environment plays a crucial role in future prediction and policy-making. Of course, this has far-reaching applications; however, only a few are discussed in this paper.