Spatiality in Medicine

By Jonathan D. Gold, MD MHA MSc FAMIA FHIMSS

May 23, 2025

Challenge yourself:

1. What is the importance of capturing spatiality in medicine?

2. Why do standards like ICD-11 and SNOMED-CT include so much depth in defining spatiality?

3. Beyond anatomy, what is the role of spatiality in healthcare and health policy?

ICD-11 and SNOMED-CT provide rich anatomic taxonomies that allow us to codify systems, organs, tissues, etc. Additionally, descriptors can be used to specify findings on the root object (e.g., abscess, lesion). These two standards also include many non-specific geographic locations (e.g., home, hospital, school). [1]

So what is the importance of capturing spatiality, and, why do these standards include so much depth in defining spatiality?

On the one hand, understanding the anatomical location for a unique patient's disease is fundamental in defining and treating it. Beyond that though, pointing to the geographical spread and determinants of disease are essential in understanding disease distribution and outbreaks, optimizing healthcare access and resource allocation, enhancing public health interventions, and improving clinical practice and medical research.

Essentially, defining and analyzing spatiality allows us to move beyond treating individuals in isolation and to consider the broader context in which health and disease occur. Spatiality, like temporality, is integral for effective prevention, diagnosis, and treatment strategies for the patient and for the broader community. It is fundamental for both precision medicine and public health.

Spatial Perspectives for Events 

Problem and procedure entities often have inherent spatiality associated with a state or action (otitis media [middle ear], myocardial infarction [heart myocardium], pneumonia [lung]; lymphadenectomy [lymph node], appendectomy [appendix]). Additional spatial modifiers increase the specificity (left lower lobe pneumonia, right otitis media, fracture of the left distal ulna; hysterectomy-vaginal approach, nasal polypectomy). Intrinsic and anatomic sites, as well as those locations on a non-corporeal object are referred to as structural spaces. Geospatial settings refer to non-structural locales (e.g., ‘Evergreen Medical Center’, ‘Duluth, MN’, and ‘school’). Positional spatiality captures the relative direction and distance or gap between two objects, structures or points. These three spatial views help define spatiality much like the three temporal views define time-based relationships.

Spatial Perspectives

Below are examples of different spatial perspectives:

1)      Structural: Location in relationship to the body, a body system, organ, structure, tissue, cell, etc.

a.       Anatomic

i.      “cardiovascular”

ii.      “aorta”

iii.      “artery”

iv.      “endothelium” 

v.      “Penetrating wound 2 cm proximal to abdominal aorta bifurcation”

b.       Non-corporeal object

i.      “ECMO”

ii.     “dialysis machine”

iii.    “pacemaker”

iv.    “catheter”

v.     “metal fragments”

2)      Geospatial: Location in relationship to the external world

a.       Geographic site

i.      Region, country, state/province, city, address, landmark, etc.

b.       Non-geographic site (descriptive definition)

i.      “home”

ii.     “hospital”

iii.    “school”

   iv.  Topography, geographic features, elevation

3)      Positional: Location of external object or another person/group/anatomical structure in relation to an object or person under review.

i.      “Penetrating wound 2 cm proximal to abdominal aorta bifurcation”

ii.     “Radiation exposure 50 kilometers south of Chernobyl nuclear site.”

Spatiality and Natural Language Processing

Not all non-anatomic spatial references are critical. Often these provide background information that aids the clinician in understanding context ("dehydrated while hiking in the desert") but do not alter the diagnosis or treatment. In certain instances, like exposures or type and location of non-corporeal objects, spatiality is critical to correct clinical interpretation ("lives in two room apartment with father who has active TB").

Construction of Plottable Spatiality

The Spatial domain contains three subdomains: 1) Parts of Speech, 2) Pre-Coordination, and 3) Foundation.

Spatial phrases can be readily interpretable, like pre-coordinated phrases, or deconstructed into parts of speech that are connected using rules of syntax to produce an interpretable meaning. Not all interpretable phrases can be plotted spatially, though many significant ones can. Semantic validity requires that parsing the spatial phrase result in a plottable location.

Spatiality Overview

Parts of Speech Subdomain

Structure

Structure defines the anatomic or non-corporeal object being referenced.

Anatomic

Non-corporeal object

Value

Value represents the number or the position in a sequence. It also includes comparative terms or phrases (called “modifiers”).

Cardinal number [“13”][2]

Ordinal number [“sixth”]

Modifier [“more than”]

Unit of Measurement[3]

Measurement serves as the type of unit associated with the value. It may include a size, volume, weight or density.

Setting

Setting provides a narrative of an anatomical object or spatial description of a procedure.

Topographic Descriptor [“lesion”, “mole”]

Shape [“annular”]

Laterality [“left”]

Location modifier [“proximal”]

Position [“supine”]

Disposition [“dilated”, “hypertrophic”]

Procedural Approach [“right paramedian”]

Mode

Mode contains prepositions and conjunctions that serve to define the context of a phrase.

Preposition [“above”, “by”, “from”]

Conjunction ["or", "and"]

Structure'

A secondary structure. Structure defines the anatomic or non-corporeal object being referenced.

Anatomic

Non-corporeal object

Contiguity

Contiguity refers to whether or not a structure, lesion, object, etc., is connected or fills a single space or is disconnected.

Non-contiguous ["discrete"]

Evenly spaced

Variably spaced

Contiguous [“uniform”]

Distribution

Distribution defines the number of occurrences per defined space or units per space.

Occurrence fraction ["4 out of 6"]

Unit fraction [“3 per LPF”]

Inexact [“approximately”]

Certainty

Certainty describes the likelihood that an event occurred at an identifiable location or anatomical structure.

Ambiguous [“possible”]

Probable/Definite [“likely”]

 

Pre-Coordination Subdomain

Pre-coordinated phrases may link value + measurement, value + measurement + non-geographic (or anatomic), or other common combinations.

 

Measurement

Length [“{x} inch(es)”]

Area [“{x} cm2”]

Volume [“{x} cc’s”]

Weight/Mass/Substance [“{x} kg”]

Density [“{x} g/cm3”]

Positional 

[“proximal to tibial tuberosity”]

Anatomic 

[“cardiovascular system”, “chest”, “heart”, “left ventricle”]

Geospatial 

Geographic position [“Pittsburgh”, “89927”, “Avista Adventist Hospital”]

Non-geographic location [“home”, “hospital”, “school”, “office”]

Observable Narrative 

[“Address:”]

Foundation Subdomain

Additional supporting groups of concepts, like ‘Calculation Concepts’ that provide mathematical expressions, points in space, delimiters, and syntactic context, and are not parts of speech, help convert text to spatial points.

The 4D Medical Record and Spatiality in Public Health

4D Medical Record

The incorporation of spatiality and its visual representation in the medical record allow for rapid understanding about many aspects of the individual patient. Beyond that though, capture of spatiality, in a larger sense can aid our understanding of public health concerns. The COVID-19 Dashboard constructed and maintained by Johns Hopkins University from the early days of the COVID crisis through the beginning of March 2023 became an essential tool for healthcare professionals and informing health policy.

Johns Hopkins University COVID-19 Dashboard, March 10, 2023, the final update. [4]

Conclusion

The approach presented above takes into account the three different spatial perspectives--structural, geospatial and positional--and provides a method for parsing locations that do not match pre-coordinated concepts.

Capturing, defining, codifying and analyzing spatiality can be integral both for precision medicine (the individual) and incorporation of critical data for public health, medical research, and policy decisions. Robust anatomical taxonomies already exist in ICD-11 and SNOMED-CT. Inclusion of relevant geospatial data expands our ability to contextualize the environment in which health events occur and our understanding of determinants of health.

---

[1] SNOMED-CT: Spatial and relational concepts (qualifier value) SCTID: 309825002

ICD-11: Topology: XK--, Anatomy: XA----, Histopathology: XH----, Dimensions of Injury: XJ---, Dimensions of External Causes: XE---, Health Devices: XD----

[2] Include concept-to-concept mappings to exact value, lower delimiter, upper delimiter, and value delimiter

[3] Standard measurements use the metric system, such that inches are converted to centimeters, feet and yards to meters, miles to kilometers, ounces to cc’s, pounds to kilograms.

[4] https://www.esri.com/en-us/industries/blog/articles/a-thank-you-to-johns-hopkins-university. Accessed: 5/25/2025