الفهرس 
A Proposed Data Extraction Technique for Electronic Medical Records
AbstractOnly 14 pages are availabe for public view
 |  | from | 140 |  |  |
| | | from | 140 | | |
Abstract
Medical records are vital for providing appropriate medical care services to any patient. While many countries already have electronic medical records (EMRs), other countries do not have paper or electronic medical records. Therefore, healthcare organizations store the information of their patients in their private healthcare information systems. If patients make visits to a new healthcare service provider, they will have to provide their medical information from scratch.
Recently, some governments have been working more seriously than ever before to adopt the use of electronic medical records (EMR). An EMR is an integral part of electronic health (e-health) applications because it contains any patients’ health information. In fact, there is a positive relationship between the existence of EMRs and the accessibility and the healthcare service usage efficiency. After comparing EMRs to paper records, EMRs are believed to result in better medical decisions and increased patient safety. Therefore, the necessity for medical data extraction and transformation into structured form to be added to EMRs is growing for encouraging healthcare workers to derive benefits from patients’ clinical information.
Extracting and transforming medical data into structured form brings several benefits. The benefits not only include the use for diagnosing diseases and monitoring healthcare conditions but also include research benefits. On a large scale, it helps with enhancing reference intervals, raising quality standards, and supporting organizational decisions. Electronic decision support systems in healthcare are considered a productive way that decreases diagnostic dispensable testing, and these systems are usually in need of structured data to offer high-quality output. On a small scale, extracting medical data contributes to rationalizing financial resources, reducing workload, conducting statistical operations, and improving patient experiences in healthcare service usage.
In this thesis, a technique was proposed to enable the participation of patients in the creation of their own personal health records. This is done by uploading their medical laboratory test reports and scanning their medications images to be added along with other inputs to an EMR system as structured data in a format that allows further analysis. In addition, patients can allow access to any of their registered healthcare providers to their healthcare data for a better medical care experience.
There are two main contributions of the thesis. Firstly, it introduces an approach that facilitates extracting data from heterogeneous unstructured or semi-structured medical laboratory test reports. Secondly, it extracts data to provide the medication commercial name, type, dose, and pack size from package images that are captured in uncontrolled capturing environments. This structured data output can be used to track patient health progress in countries that do not have EMRs. Both approaches do not require any data entry so they are more flexible to an extent that they are likely to work with no issues on extracting and transforming medical laboratory test reports of other structures and images of new medications packs into a structured data.