Over 11 years of my career was spent in the field of Healthcare, especially in Radiology to implement, support and to manage RIS (Radiology Information System) and PACS (Picture Archiving and Communication System) solution. During my time I have seen the evolution of RIS and PACS from a just being treated as a means of storing healthcare digital images of various healthcare studies such as CT, MR, Mammo, general x-ray and ultrasound, to processing these images for diagnostic purposes locally within the institute to sharing the digital images across institutes within country for diagnosing, further consulting and improving patient care within constraints of regulations.
In the previous articles we have already discussed what is Big data, the advantages and applications. In this article we specifically attempt to help you understand how Big Data and Analytics is currently being used in the field of Radiology and how institutes who have not embarked in Big Data can embrace the potential that Big Data offers.
Out of the 5 V’s of Big Data Character, let us consider the below four.
Data collected in radiology is of the highest in terms of storage space taken in a hospital. Radiology when collected data over years has many data sets of a similar category (e.g: CT Abdomen, CT Pelvis, X-Ray Chest).
The speed in which the data is generated in an institute. With the very high cost of medical equipment such as CT or PET/CT or MR, institutes maximizes by ensuring this equipment are used to their highest potentials. Some institution have patient scheduled for up to 6 months with no slots for any additional scans.
The data generated from radiology come from various modality types, such as CT, PET-CT, MR, Angio, Ultrasound, Mammo, general x-ray and etc.
This related to the quality of data that comes into the dataset. Most of the radiology department have radiographers and senior radiographers who will vet through the study and send in the study which are required for the study. If for any reason a series of the study found corrupted or incomplete or there are some artefacts found, the images or series are deleted and re done. Thus ensuring a high veracity and low noise in the dataset.
Radiology is also the best suitable department to get into big data analysis, because all the patient demographics, related data such as radiological images, medical conditions and others are all integrated between each other via RIS (Radiology Information System) and PACS (Picture Archiving and Communication System). These data are integrated with HIS (Hospital Information System) where other data such as Anaesthetic, Cardiology, Laboratory, Pharmacy and other data are integrated. With all these data available in a hospital, big data classifies them into structured and unstructured data.
Data Acquisition and storage stage:
In regard to radiology data, structured data are non-radiological images such as patient demographics, study details, details from HL7 message, Lab test reports, details on patient visits and etc. Unstructured data are radiological modalities which can generate in a healthcare standard DICOM format and non-DICOM formats such as PDF, MPEG, JPEG, PNG, avi, XLM and many others. These radiology images are stored and shared across the organization using a suitable XDS based infrastructure.
Data Pre-processing (Validate and Cleansing) Stage:
The data acquired in earlier stage must be checked for consistency and validated against a set of predetermined conditions and rules in order to ensure the data is valid. The data has also to be ensured the details are stored in the specific category and format. Data may be spread across multiple datasets, requiring that dataset be joined together to conduct the actual analysis. To ensure only the correct data will be analyzed in the next stage, it might be necessary to integrate multiple datasets. Such as the details in patient demographics.
Data Analysis and Interpretation Stage:
The Data Analysis stage of the Big Data Lifecycle stage is dedicated to carrying out the actual analysis task. A query can be simple query like for e.g.: What is mortality rate in a particular region? or complex query as how many patients with diabetes are likely to develop heart related problems in next 5 years? Depending on the complexity of the query, the data analyst must choose appropriate platform and analysis tools.
Benefits of using Big Data Analytics in Healthcare
For Individuals and Patients
Improved predictive analytics to support clinical decision for patient diagnosis and treatment by corelating and cross referencing the patient’s medical history, medical condition and diagnosis to the same patient’s previous diagnosis and also with other patients with similar medical history, diagnosis and conditions.
Based on the above pointer, correct and effective treatment and also predicting patient’s criticality by continuous health monitoring using wearable devices, which helps to increase the life expectancy and quality of life of patients.
Better prediction considering all the internal and external data to build a better predictive system, to predict number of patients visit to a hospital per day which is used to schedule the number of doctors, nurse and other administrative staff. This forecast reduces patient waiting time and better patient care.
Better administrative decisions for hospital management, predicting hospitalization patterns and answering various questions regarding disease treatments.
Other than the above two benefits seen, there are various other benefits realized by Insurance companies, Pharmaceutical and government organizations.
Big Data Analytics for healthcare has its own set of challenges. But big data analytics strength is high and has the potential to transform the way healthcare organization provide treatment from traditional ways to a more data driven, gaining insights from tools and technologies for medical professionals to make constructive informed decisions.
There are various innovative ideas and solutions which are being developed using the possibility of Big Data Analytics in the field of Healthcare and in the near future we can be confident that we will see a rapid, widespread implementation and use of big data analytics across healthcare organizations and industry.