mathematical modelling of the carotid–ophthalmic Arterial network from Anatomage Virtual Dissection: proof-of-concept | Koren Shapiro
Modelling Carotid–Ophthalmic Arterial Network
Authors: Koren Shapiro (1), Anton Morgenstern (1), Ethan Kol (1), Bacem Othman (1,2)
Supervisor: Omid Moztarzadeh (1)
(1) Department of Anatomy, Faculty of Medicine in Pilsen, Charles University
(2) Bioptical Laboratory, Ltd, Pilsen, Czech Republic
State-of-the-Art: Ophthalmic artery pathology arises from atherosclerotic stenosis, embolic
occlusion, vasospasm, inflammatory arteritis, compressive mass, traumatic dissection, fibro
muscular dysplasia, or iatrogenic injury. While mathematical models of intracranial vascular
networks exist for stroke and aneurysm management, quantitative analysis of full extracrani
al-to-intracranial collateral supply remains lacking. Virtual dissection enables precise geometric
extraction of vessel radii, lengths, and bifurcation angles across the entire arterial tree. Our
Anatomage-based mathematical model focuses on intracranial-extracranial segmentations,
with a particular emphasis on orbital revascularization.
Objective: We aim to develop a cadaver-adaptable mathematical network model of the cere
brovascular tree, derived from Anatomage virtual dissection, to rank extracranial routes sugges
tive of ophthalmic artery pressure relief.
Material and Methods: Geometric measurements were extracted from virtual dissection of 65
arterial segments across cerebrovascular, vertebrobasilar, craniofacial, and cervical vessels bi
laterally using Anatomage Table data, assuming steady-state, rigid-wall Newtonian flow. Unlike
synthetic Lindenmayer-system or constructive constrained optimization methods, we assessed
every bifurcation using directly measured Anatomage morphometry, which was computation
ally enhanced by processing the VPF segment. Hemodynamic resistance was modeled propor
tionally using the simplified Poiseuille relationship. The network was assembled as a Kirchhoff
admittance matrix and solved across 65 nodes via Gaussian elimination. Possible stenotic foci
were imposed as a resistance penalty scaling to simulate plausible branch occlusion.
Results & Discussion: Our model identified the inferolateral trunk, the meningohypophyseal
trunk branches, and the sphenopalatine artery as collateral steal conduits that reduce ophthal
mic perfusion, highlighting extracranial therapeutic targets. Among 28 tested extracranial by
pass configurations, angular artery-to-dorsal nasal anastomosis ranked first (18–24% pressure
reduction), followed by superficial temporal-to-supraorbital and infraorbital-to-lacrimal anas
tomoses. Internal carotid artery cavernous segment stenting produced the largest single-vessel
conductance gain. Murray‘s law deviation concentrated at the ophthalmic origin and caroti
cotympanic bifurcation, predilection sites for turbulence and embolic lodgment. A user-friendly
interface for the model is shown in Fig 1.
Conclusion: Mathematical modelling of the full cerebrovascular network from Anatomage vir
tual dissection morphometry may provide quantifiable alternative scenarios for cerebrovascu
lar diseases, potentially obviating the need for a craniotomy
Anatomage-Derived morphometric model for coronary Arterial reconstruction | Charlotte Schild
Morphometric Model for Coronary Grafting
Authors: Charlotte Schild (1), Andrew Kyle Stewart Groser (1)
Supervisor: Bacem Othman (1,2)
(1) Department of Anatomy, Faculty of Medicine in Pilsen, Charles University, Czech Republic
(2) Bioptical Laboratory, Ltd., Pilsen, Czech Republic
State-of-the-Art: Coronary heart disease is often arising secondary to the build-up of plaque
in the endothelial lumen rendering it stenotic. This morbid vascular dysfunction is highly preva
lent worldwide, especially in geriatric population. Contemporary computational models of coro
nary arterial anatomy rely predominantly on L-system branching algorithms, three-dimensional
Navier–Stokes solvers, synthetic tree generators, and constrained constructive optimisation.
These models flatten segmental tortuosity, bifurcation angles, and collateral topology. Our
study therefore characterise the broader permutational range of coronary anastomotic config
urations using Anatomage data.
Objective: Using virtual cadaveric dissections, we aim to characterise the luminal radius, taper
rate, cross-sectional area, segment length, relative hydraulic resistance, and branching hierar
chy of coronary arteries to improve grafting strategies of tortuous coronary arteries with max
imal efficiency.
Material and Methods: The right coronary artery (RCA), left coronary artery (LCA), left ante
rior descending artery (LAD), left circumflex artery (LCX) and their branches were digitally dis
sected and precisely measured using the Anatomage measurement tools. For each segment, the
following parameters were measured: proximal radius (Ru), distal radius (Rd), segment length,
mean radius, cross-sectional area, taper rate, and relative hydraulic resistance. Branching hier
archy and parent-segment relationships were charted across 37 discrete arterial segments from
the aortic arch down to terminal coronary vessels. Tortuosity indices and ramification patterns
were assessed at each bifurcation point.
Results & Discussion: Morphometric data were derived from four virtual cadavers (2F/2M;
Asian and Caucasian; adult and geriatric). Proximal radii ranged from 84.6 µm (RCA proximal) to
3.3 µm (AV nodal branch, right), with taper rates spanning 0.057 (LCX proximal) to 1.008 (Acute
Marginal 1). Relative hydraulic resistance varied by four orders of magnitude, from 3.3 × 10-6(RCA
proximal) to 4.99 × 10-2 (AV nodal branch, right). Septal perforators, ciliary branches, and nodal
vessels were resolved at sub-millimetre radii. The dataset further identified deviations from Mur
ray ’s law at multiple bifurcation nodes, with area ratios and asymmetry indices incompatible with
synthetic tree assumptions.
Conclusion: The Anatomage virtual dissection platform provided calibration data of sufficient
fidelity to resolve the visualisation of individual septal perforators, ciliary branches, and nod
al vessels at sub-millimetre resolution. Limitations included undisclosed medical histories and
variable segmentation details among the cadavers
Lymphatic segmentation and regional Architectural patterns using Anatomage: A pilot study | Leen Abo Jeesh
Lymphatic Segmentation and Architectural Patterns
Authors: Leen Abo Jeesh (1), Bacem Othman (1,2)
Supervisor: Omid Moztarzadeh (1)
(1) Department of Anatomy, Faculty of Medicine in Pilsen, Charles University
(2) Bioptical Laboratory, Ltd., Pilsen, Czech Republic
State-of-the-Art: Nodal size and shape have been considered determinants of neoplastic me
tastasis, with open questions raised on lymphotropic routing and anatomic predilection sites for
neoplastic progression. Moreover, inadequate attention has been paid to anatomical morpho
metric characteristics of lymphatic vessels, nodal clustering and geometric architecture. This
study postulates that lymphatic vessel segmentation and regional architectural patterns may
test theoretically grounded insights.
Objective: To segment lymphatic vessels in Anatomage and characterize lymphatic architecture
patterns—including vessel continuity, regional node clustering, and vessel-node spatial rela
tionships—in cancer, non-cancer, and morphologically indeterminate cases.
Material and Methods: Anatomage Table lymphatic data retrieved from three representative
cadavers was used with color-coded segmentation. They included a female with confirmed GIT
malignancy (Cancer_1), a cancer-free female (No_Cancer_2), and a male with unknown clinical
status (Unknown_3). Lymphatic vessels and lymph nodes in head/neck, thorax, abdomen/pel
vis, and extremities were labelled with different channel dominance. Manual and automatic
measurements assessed regional node count, node density, vessel density, mean node size,
lymphotropic routing and vessel tapering gradient. Manual measures for all nodes was per
formed using Anatomage‘s toolkit. The automatically caliberated paramenters were assessed
computationally after exporting VPF files for the lymphatic maps and corresponding imaging.
Results & Discussion: GIT Cancer cadaver demonstrated higher overall node burden ( 39.93%
concentrated in abdomen and thorax) and elevated vessel density. In contrast, Cancer-free ca
daver showed inverse predominance, with abdominal/pelvic nodes comprising 31.18%. Male
cadaver with unknown data presented a generalized node distribution accounting for 71.07%
of all nodes, with minimal extremity involvement (3.21%). Exceeding even Cancer_1, node pixel
concentration and lymphotropic routing in Unknown_3 were similarly skewed toward the thorax
and neck with whose sizes nodes exceeded 3 cm in largest diameter across multiple regions
(Fig. 1). All cadavers showed free-end saddles of lymphatic vessels in the head/scalp, inferring
tapering connection with the glymphatic system.
Conclusion: Lymphatic architectural analysis can be used as a complementary tool in the mor
phometric classification of nodal disease and mapping functional lymphatic system. Limitations
of the study include a few representative cadavers for sampling.
Funding: None