Data analytics in medical data

processing

Prof. Naveen T.S.

Assistant Professor
Department of Medical Electronics Engineering
Dayananda Sagar College of Engineering
Bangalore
 Data Analytics
Data analytics is the process of analyzing raw data for the
purpose of determining trends and enabling better decision
making. It is relevant to all types of organizations, especially
health care organizations.

Data analytics in health care is vital. It helps health care

organizations to evaluate and develop practitioners, detect
anomalies in scans and predict outbreaks in illness.

Data analytics can also lower costs for health care organizations
and boost business intelligence. Most importantly, it helps health
care companies to make better care decisions for patients.
 Data Processing in Healthcare
• Raw data is not helpful for any organization. To make the data useful or in a
readable format use graphs, charts, and texts, we can use it.
• Data processing directs to collecting raw data and converting it into usable
information.
• The raw data is processed, gathered, filtered, sorted, analyzed, and stored before
being presented in a readable format.
• A genuinely data-driven medical analytics solution should be able to integrate all
the data sources and analyze both structured and unstructured data in real-time.
Your insights aren't as successful as they may be unless you include all the patient
data, including diagnostic information, doctor observations, and real-time data
from medical equipment.

https://www.xenonstack.com/blog/data-processing-in-healthcare
Increase outcomes
By delivering real-time patient information and decreasing the risk of human error and incorrect diagnosis, it can
improve the quality of patient care and safety levels.

Increase operational efficiencies

Can cut costs, decrease waste, and streamline your operations by combining data from numerous sources. It will also
provide a better understanding of resource use, inventory levels, and procurement.

Medical Equipment model development for recommendations

Big data analysis help in the development of new medical equipment in it. Its specialists can detect potential strengths
and weaknesses in trials of the development of medical equipment by combining historical, real-time, predictive
information and a cohesive blend of data visualization tools.

Enhanced Medical research on chronic diseases

The healthcare sector can perform intensive research on chronic diseases by drilling down insights such as medication
type, symptoms, transmission methods, and prevention methods from various patients' data.

Patients' Disease analysis over time

Patient data can be cleaned, aggregated, and stored in a database every visit. The stored data can help analyze the
patient's disease pattern and medications taken over time for better health care.

Finance and accounting should be simplified

Big data analytics in healthcare generates insights that can help you swiftly discover and correct financial mistakes,
allowing you to align your KPIs to achieve your financial objectives.
 Ways to use data analytics in healthcare
Data analytics and visualisation can increase a patient’s access to services and lower costs. Here are the
major applications of data analytics in healthcare:

• Manage patient health records: Hospitals need to digitise medical records for substantial cost
savings and extracting data from administrative and diagnostic systems to update them in real time.
• Forecast operating room demands: Operating rooms are expensive to build and maintain. Data
analytics helps hospitals optimise the cost without impacting patient care. This demands a
thorough understanding of the relationships between operating room variables for better
scheduling.
• Optimise staff: Accurate staffing is important to save costs and time. Data analytics help hospitals
predict and deal with staff challenges based on local weather trends, holidays, and seasonal
infections in advance.
• Prevent 30-day hospital readmissions: Data analytics prevents unnecessary readmissions in a 30-
day window to reduce costs and ensures availability for patients who need immediate care.
• Predict no-show appointments: Patient no-shows have financial ramifications and throw doctors
off their schedule. Data analytics can predict no-shows to improve staffing and reduce wait times.
• Manage supply chain costs: Hospitals rely on massive supply chains, and delivery times are critical.
Data analytics maintains the efficiency by tracking supply chain metrics, thus saving lives and costs.
• Enhance security and prevents fraud: Cybercriminals attack hospital databases resulting in the loss
of revenue and leakage in confidential information about patients. Data analytics identifies patterns
and changes in network traffic to detect suspicious online behaviour.
• Reduce medical errors: Medical errors in surgeries, diagnosis, and medication affect over 400,000
patients annually. This happens due to a lack of information or negligence by the staff. Data
analytics reduces such errors by flagging anything that seems out of place.
https://online.maryville.edu/blog/data-analytics-in-healthcare/
• Descriptive analytics: Descriptive analytics in health care uses historical patient
data to glean insights into benchmarks and trends.
• Prescriptive analytics: Prescriptive analytics relies on machine learning to propose
a strategy.
• Predictive analytics: In health care, predictive analytics uses both forecasting and
modeling to predict what will probably happen in the future.
• Discovery analytics: Like prescriptive analytics, discovery analytics also uses
machine learning. The difference is, it utilizes machine learning to examine clinical
data for the purpose of determining patterns that provide actionable insights.

https://www.comptia.org/content/articles/how-is-data-analytics-used-in-health-care
 Analyzing clinical data to improve medical research
Data analytics techniques are being applied to improve research efforts in many health-related areas by
gathering and analyzing clinical data from various sources. Among the most useful sources of clinical
information are EHRs, electronic medical records, personal health records, and public health records.

EHRs:
Combine a patient’s X-rays and other medical images, diagnoses, treatment plans, allergies, and test
results in standard digital formats. This makes the information easy to share but introduces privacy and
regulatory compliance requirements that limit how the data may be used.
Electronic medical records:

Similar to EHRs but include only information from the patient’s paper charts created in medical offices,
clinics, and hospitals. They are used primarily for diagnosis and treatment; their main value is in tracking
a patient’s healthcare over years of visits and screenings.
Personal health records:

Maintain a history of the patient’s health treatment that the patient keeps rather than healthcare
providers. The records are intended to assist in the patient’s own health management and don’t legally
replace the medical records that healthcare providers maintain.
Public health records

Among the most promising sources of health data for medical research. For example, the National
Cancer Institute’s Cancer Research Data (CRDC) serves as a cloud-based data science platform that links
data analytics tools with data repositories storing genomic, proteomic, comparative oncology, imaging,
and other data types.
 Data points in the healthcare industry

Hospitals operate and manage a wide range of clinical and operational information
systems. While this is not an exhaustive list, here are some of the data points:

• Electronic health records (EHRs): Clinical documentation, reporting test results,

patient medical history, and patient orders
• Laboratory information system (LIS): The laboratory information system that
interfaces with the EHRs
• Diagnostic and monitoring: The vast inventory of diagnostic and monitoring
equipment for magnetic resonance imaging, vital signs, and test result
interpretation, which may or may not interface with the EHRs
• Insurance claims and billing: Information on a patient’s treatment, the costs, and
expected payment as well as information in the EHR that determines the level of
service
• Pharmacy: Data on medicines and other supplies sourced from the hospital
pharmacy
• Human resources and supply chain: Data about employees and their role within
the medical institution, which also includes the location and utilization of medical
supplies
 The data processing strategies in the cloud
AWS (Amazon Web Services)
Amazon EMR:
A cloud big data platform that uses open-source analytics frameworks to conduct large-
scale distributed its jobs, interactive SQL queries, and machine learning(ML)applications.
AWS Glue:
A server less data integration tool for finding, preparing, and combining data for analytics,
machine learning, and application development.
Amazon Kinesis:
Gathers, process, and analyze real-time and streaming data.
Amazon S3:
An object storage service with industry-leading scalability, data availability, security, and
performance.
GCP (Google Cloud Platform)
Dataproc:
A purpose-built, fully managed, highly-scalable cluster.
Cloud Data Fusion:
A fully managed, cloud-native data integration tool at any scale.
Dataflow:
A serverless streaming analytics service.
Cloud Storage:
An object storage service in google cloud.
Azure
HDInsight:
A serverless cloud Hadoop and Spark service with a 99.9% SLA.
Data Factory:
Data transfer and transformation can be orchestrated and automated with the help of
a data integration service.
Azure Synapse Analytics:
Unlimited information analytics service.

On-Premises
Hadoop:
Open-source, distributed framework.
Spark:
Distributed, open-source data processing engine that processes data using random
access memory (RAM).
Data Lake:
A centralized data storage, processing, and security solution for massive amounts of
organized, semistructured, and unstructured data.
Hive:
Open-source, distributed data warehousing database operating on Hadoop Distributed
File System.

https://www.xenonstack.com/blog/data-processing-in-healthcare