Secure Health Information System with Blockchain Technology

Abstract

This paper focuses on highlighting the problems that are associated with the absence of privacy and security of medical records in a healthcare system. It seeks to bridge the gap between the currently used security protocols in the management of health information, and encryption algorithms that should be used. Extant health information systems have always been developed with conventional databases. With all the privileges to read, write and execute assigned to the administrator, who has centralised control over all medical records, there is the likelihood of the misuse, distortion and loss of such records in the event that the administrator becomes compromised or inadvertent system failure. To solve this problem, the use of decentralised and distributed databases becomes paramount. Blockchain technology has recently received much attention due to its ability to permit a peer-to-peer network with distributed databases that can be stored locally on each node in the network. Subsequently, all updates on records in a database are communicated to all participating parties, hence addressing the problem of centralised control. In this paper, we propose a health information system on a blockchain to create a trust-free system for both health personnel and patients. From the results obtained, we achieved the decentralisation of the medical records’ database to enhance the security and privacy of data on the modeled peer-to-peer network.

Enhancing the Random Forest Model via Synthetic Minority Oversampling Technique for Credit-Card Fraud Detection

“…FS is a pre-processing step that reduces the dimensionality of a dataset by removing irrelevant and docile feats or parameters [6], [67] -leading to an improvement in the model classification performance [68]- [70]. It also yields streamlined data collection in model training for scenarios where cost is critical (e.g., target design in gene therapy).…”

Strategic Feature Selection for Enhanced Scorch Prediction in Flexible Polyurethane Form Manufacturing

“…FS is a method to reduce dimensionality by removing irrelevant features or parameters [33], [34]. As a pre-processing step in machine learning tasks, FS has been successfully used in a variety of applications as its usefulness are numerous to include (and not limited to) overcoming the burden of parameter dimensionality and eliminating irrelevant cum docile features [35]- [37]; Thereby, leading to enhanced performance of the machine learning heuristics for both regression or classification task. FS becomes critical in domains where cost and the measure of attributes are of utmost importance.…”

Hybrid Genetic Algorithm Trained Bayesian Ensemble for Short Messages Spam Detection

“…In this study, accuracy, recall, error rate (ER) and specificity are used to evaluate the performance of the detection models. The formulas of the above criteria are calculated as follows (Ibor et al, 2023;Ojugo, Aghware, et al, 2015b;Ojugo, Oyemade, et al, 2015). To measure effectiveness and accuracy, we measure their rate of misclassification and corresponding improvement percentages in both training and test data sets as summarized in Tables 3.…”