Login
Introduction
Videos
Our Team
Blogs
Glossary
HCn3D System
Provider Logic
Patient Centered Network
Patient Journey
Value | Population
Patient Centered Medical Home
CMS AI Health Outcomes Challenge
TennCare
The Pyramid Metaphor
1st Dimension
2nd Dimension
3rd Dimension
Scenarios
Reinforcement
Contact
Introduction
Videos
Our Team
Blogs
Glossary
HCn3D System
Provider Logic
Patient Centered Network
Patient Journey
Value | Population
Patient Centered Medical Home
CMS AI Health Outcomes Challenge
TennCare
The Pyramid Metaphor
1st Dimension
2nd Dimension
3rd Dimension
Scenarios
Reinforcement
Contact
Glossary
Home
|
Introduction
|
Glossary
Healthcare in 3D Terms and Phrases
Dimension
The reference to “dimensions” in the name Healthcare in 3D (3 Dimensions) is not just a catchy marketing phrase, but also a clever reference to the 3 types (or context) of dimensional concepts ubiquitous to HCn3D:
spatial
,
statistical
, and
stratified
.
Dimension (spatial)
In the context of “spatial”
dimension
refers to the sides of the “cube” which is 3 dimensional. Provider Logic, measure, and Time Intervals.
Dimension (statistics)
In the context of “statistics” (data)
dimension
refers to structures or attributes that categorize measurements or metrics. For instance systolic would categorize the "top" number in a patient's blood pressure.
Dimension (strata)
In the context of “strata”
dimension
refers to layers. Most commonly the
multilevel hierarchal layers of the healthcare network
Dimension Reduction
Is the process of reducing the number of variables under consideration to a manageable and useful number of independent
features
. 100s if not 1000s of individual pieces of data (dimensions) may be relevant to a Provider Action at the
Inflection Point
. A very simple example of Dimension reductions can be demonstrated in the categorization of blood pressure. Systolic and diastolic are what data scientists would call
Natural Features
and they could be used directly to
Train
a model, but by
Feature Engineering
a third dimension that categorizes the measurement into a list of 3
Factors
: “Normal”, “hypotensive”, and “hypertensive” the processing and possibly even predictive performance is improved. This is also a good example of Feature Engineering to improve the ability to understand the data, since knowing a patient simply has Hypertension is more intuitive than reading the actual mm of Hg.
High dimensional
Having the property of a higher level of the
network
.
High Dimensionality
In the context of HCn3D analysis, it is simply data instances (records) with the potential to comprise many variables (columns), potentially more variables that there are records being analyzed (e.g. Electronic Medical Record).
Page 1 of 1