Topographical Mapping for Precise Disease Phenotyping

This direction is merged into another direction: anatomical lesion localization. Its last version stays readable and citable here.

Version 1 of 152 papers61 conceptsfrozen 2026-10-04

In clinical medicine, a diagnosis often answers the question "What is wrong?" However, a growing body of research suggests that the more critical question is "Where exactly is it happening?" By shifting from global assessments—like a single "brain age" score or a binary "stroke" label—to high-resolution spatial maps, researchers are uncovering distinct disease subtypes, predicting survival rates, and identifying the root causes of complex conditions.

This approach, known as topographical phenotyping, treats the spatial distribution of lesions not as random noise, but as a structured signature of the underlying pathology.

From Global Scores to Spatial Maps

Traditional diagnostic models often compress complex anatomical data into a single number. For instance, "brain age" is typically calculated as a global metric of neurodegeneration. However, Ben Glocker and colleagues in Local Brain-Age: A U-Net Model argue that this obscures vital information. By generating high-resolution spatial maps of aging, they allow clinicians to cluster patients based on which regions are aging fastest. This spatial profiling can distinguish between normal aging and specific neurodegenerative trajectories that a global score would miss.

Similarly, in the study of radiation-induced lung damage, Bin Liang and Hui Yan introduced "dosiomics." Instead of using a one-dimensional summary of radiation dose, they analyze the 3D spatial distribution of energy across anatomical sub-regions. This multi-region approach captures localized effects that are far more predictive of radiation pneumonitis than standard clinical models (Dosiomics: Extracting 3D Spatial Features From Dose Distribution to Predict Incidence of Radiation Pneumonitis).

Anatomical Partitioning as a Diagnostic Filter

The core intuition behind topographical mapping is that different diseases have "territorial preferences." By partitioning an organ into discrete zones, researchers can use the location of a lesion to infer its etiology.

In stroke research, Mattias Drake and Markus D. Schirmer demonstrate that mapping ischemic lesions to specific vascular territories is essential for identifying the source of the blockage. Multiple lesions scattered across different arterial zones typically point to a systemic embolic source, such as the heart, whereas a single isolated lesion suggests localized small-vessel disease (Diffusion-Weighted Imaging, MR Angiography, and Baseline Data in a Systematic Multicenter Analysis of 3,301 MRI Scans of Ischemic Stroke Patients).

This logic extends to microvascular health. Bryce Wei Quan Tan and Christopher Li Hsian Chen use anatomical rating scales to categorize cerebral microbleeds. Microbleeds in the "lobar" (outer) regions of the brain often signal cerebral amyloid angiopathy, while those in "deep" regions are more indicative of hypertensive damage (Haemoglobin, magnetic resonance imaging markers and cognition: a subsample of population-based study).

Automated Localization and Semantic Enrichment

The bottleneck for topographical mapping has historically been the time required for manual anatomical labeling. Recent work has focused on automating this process using deep learning architectures like the U-Net.

Thomas Weikert and Paul F. Jaeger developed a pipeline that combines lesion detection with automated region segmentation. By calculating the spatial overlap between a detected lesion and a segmented organ, the system can automatically label a mass as a "primary tumor" or a "metastasis" based on its anatomical context (Automated lung cancer assessment on 18F-PET/CT using Retina U-Net and anatomical region segmentation).

To make these automated systems even more precise, Francesco Cremonesi and Olivier Humbert proposed "semantic enrichment." By mapping AI-generated segmentations to standardized medical ontologies like the Foundational Model of Anatomy (FMA), they transform raw spatial coordinates into clinically meaningful data, allowing for the discovery of survival-linked patterns in metastatic lung cancer (Knowledge-based semantic enrichment of medical imaging data for automatic phenotyping and pattern discovery in metastatic lung cancer).

Spatial Patterns as Mechanistic Signatures

Beyond simple localization, the morphology and fragmentation of lesions provide clues about disease progression. Anna K. Bonkhoff found that in certain brain regions, the number of discrete lesion fragments—rather than the total volume of damage—is the primary driver of clinical deficits (Deep profiling of multiple ischemic lesions in a large, multi-center cohort).

In oncology, the "seed and soil" hypothesis is being validated through spatial statistics. Nobuyuki Izutsu and Manabu Kinoshita used Chi-square frameworks to prove that breast cancer subtypes have specific "preferences" for brain regions: Luminal subtypes favor the cerebellum, while HER2-positive tumors target the basal ganglia (Cerebellar preference of luminal A and B type and basal ganglial preference of HER2‑positive type breast cancer‑derived brain metastases).

Prognostic Boundaries and Barriers

Finally, topographical mapping identifies critical physiological thresholds. Bin Zhang and colleagues demonstrated that for nasopharyngeal carcinoma, the diaphragm acts as a prognostic boundary; once cancer crosses this landmark into the subphrenic lymph nodes, survival rates drop precipitously (Subphrenic Lymph Node Metastasis Predicts Poorer Prognosis for Nasopharyngeal Carcinoma Patients With Metachronous Metastasis).

Whether it is the growth plate in pediatric bone infections (haiting jia) or the vascular territories of the brain, these spatial landmarks provide a structural framework for understanding how diseases move, hide, and ultimately respond to treatment.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_topographical_mapping_for_precise_diseas_12aa522d,
  title = {Topographical Mapping for Precise Disease Phenotyping},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/topographical-mapping-for-precise-disease-phenotyping/rdr_12aa522daad54f8f8159017c56a1c0e9},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_topographical_mapping_for_precise_diseas_12aa522d_v1,
  title = {Topographical Mapping for Precise Disease Phenotyping},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/topographical-mapping-for-precise-disease-phenotyping/rdr_12aa522daad54f8f8159017c56a1c0e9/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}