Diagnosis and surgical decision-making in hepatic hemangioma: From venous malformation overlap to artificial intelligence-assisted multimodality imaging

Table 1 Terminology and entity distinctions relevant to a diagnosis-focused review of hepatic hemangioma

Entity	Typical setting	Key diagnostic clue	Management implication
Adult hepatic hemangioma (routine radiology term)	Incidental lesion in adults; usually asymptomatic	Marked T2 hyperintensity and progressive nodular fill-in on dynamic imaging	Most lesions can be observed; biopsy is usually unnecessary when imaging is classic
Hepatic venous malformation (vascular anomaly framing of many adult ‘cavernous hemangiomas’)	Symptomatic or giant adult lesion; slow-flow ectatic vascular spaces	Overlap of imaging/pathology with adult HH; recent GJA4-linked studies support VM biology	Introduces percutaneous sclerotherapy and vascular-anomaly-based terminology into management discussions
Infantile hepatic hemangioma	Infancy; may be focal, multifocal, or diffuse	Biologically proliferative vascular tumor with pediatric presentation	Different natural history and treatment pathway from adult incidental HH
Sclerosed/hyalinized hemangioma	Older or involuted lesion; often atypical imaging	Reduced T2 brightness, incomplete fill-in, retraction or rim-like enhancement	Major mimic of malignancy; often drives escalation to MDT review or surgery if unresolved.

HH: Hepatic hemangioma; VM: Venous malformation; MDT: Multi-disciplinary team.

Table 2 Multimodality imaging clues for typical and atypical hepatic hemangioma

Lesion	Typical ultrasound/CEUS pattern	Typical CT/MRI pattern	Main diagnostic pitfall
Typical HH	Well-defined hyperechoic lesion; CEUS shows peripheral nodular enhancement with progressive fill-in	T1 hypointense, markedly T2 hyperintense; peripheral nodular arterial enhancement with delayed centripetal fill-in	Usually, few pitfalls when the imaging pattern and clinical context are both classic
Flash-filling HH	May enhance rapidly and almost homogeneous on CEUS	Uniform arterial hyperenhancement, sometimes with less obvious delayed fill-in	Can mimic FNH, HCC, or hypervascular metastasis if delayed phases and T2 signal are underweighted
Sclerosed/hyalinized hemangioma	Often nonspecific or heterogeneous	Reduced T2 hyperintensity, incomplete enhancement, fibrosis, retraction	Frequently mistaken for cholangiocarcinoma or metastasis

HH: Hepatic hemangioma; CT: Computed tomography; MRI: Magnetic resonance imaging; CEUS: Contrast-enhanced ultrasound; FNH: Focal nodular hyperplasia; HCC: Hepatocellular carcinoma.

Table 3 Common mimics that should not be prematurely labeled as hepatic hemangioma

Mimic	Typical ultrasound/CEUS pattern	Typical CT/MRI pattern	Clue arguing against classic HH
Focal nodular hyperplasia	May be iso- or mildly hyperechoic; spoke-wheel vascularity can be present	Homogeneous arterial enhancement; often iso/hyperintense on hepatobiliary phase	Presence of functioning hepatocytes and hepatobiliary phase uptake argues against HH
Hepatocellular carcinoma	Variable; often not classically hyperechoic in cirrhosis	Arterial phase hyperenhancement with washout/capsule; diffusion restriction more common	Cirrhotic background, washout, growth, and capsule should override a hemangioma default assumption
Hypervascular metastasis	Often multiple; CEUS may show rim or targetoid enhancement	Ring-like or targetoid arterial enhancement, lower T2 signal, restricted diffusion, no progressive blood-pool matching	Known malignancy, multiplicity, interval growth, and ring enhancement make HH less likely

HH: Hepatic hemangioma; CT: Computed tomography; MRI: Magnetic resonance imaging; CEUS: Contrast-enhanced ultrasound.

Table 4 Representative artificial intelligence and radiomics studies relevant to hepatic hemangioma differential diagnosis

Ref.	Task and modality	Main message	Key limitation
Popa et al[20], 2023	Systematic review of automated focal liver lesion diagnosis across US/CT/MRI	Automated diagnosis is feasible, but no modality or model family clearly dominates	Heterogeneous studies; many older single-center datasets
Arribas Anta et al[50], 2025	Narrative review of AI for focal hepatic lesion detection/characterization	AI is strongest as decision support for lesion characterization and workflow assistance	Mostly retrospective evidence; limited real-world implementation data
Pomohaci et al[51], 2025	Systematic review of AI in liver imaging	Liver lesions are the commonest AI target, but external validation and code sharing remain uncommon	Broad liver scope; not restricted to HH-specific questions
Park et al[21], 2026	T2-weighted MRI radiomics for small metastasis vs hemangioma	Excellent discrimination, including visually indeterminate lesions in validation sets	Retrospective study; validation still scanner- and center-dependent
Ozaki et al[22], 2026	Deep-learning reconstruction for DWI differentiation of metastasis vs hemangioma	Improved image quality and lesion conspicuity; diagnostic accuracy similar to standard reconstruction	Small sample size; single-vendor technical ecosystem
Starmans et al[23], 2024/Yang et al[52], 2025	T2-based radiomics for benign vs malignant liver lesions and blood-rich lesion classification	Quantitative T2 features can complement conventional reading when blood-rich benign lesions mimic malignancy	Broader lesion groups rather than pure HH-focused cohorts

HH: Hepatic hemangioma; AI: Artificial intelligence; US: Ultrasound; CT: Computed tomography; MRI: Magnetic resonance imaging; DWI: Diffusion weighted imaging.

Table 5 How diagnostic classification influences interventional and surgical decision-making

Comparison	Evidence summary	Practical implication for a diagnosis-focused review
Observation vs intervention	Many asymptomatic HHs remain stable or enlarge slowly, and lesion size alone is a poor isolated indication for treatment	Observation is generally preferred when diagnostic certainty is high and symptoms are absent or not clearly attributable to the lesion. Intervention is considered when symptoms, compression, complications, or diagnostic uncertainty affect management
Percutaneous sclerotherapy vs transarterial embolization/chemoembolization	Recent pooled analyses suggest similarly high technical success and comparable clinical or radiological response, with low major complication rates	If giant adult lesions are framed as low-flow venous malformations, sclerotherapy becomes a rational alternative rather than a marginal option
Percutaneous sclerotherapy vs surgical resection	Retrospective data suggest both can be effective for selected large lesions, with different trade-offs in efficacy, invasiveness, and resource use	Atypical or indeterminate lesions still require secure diagnosis before non-surgical treatment is chosen
Laparoscopic vs open resection	Meta-analysis and propensity-matched cohorts generally favor laparoscopy in selected peripheral lesions due to reduced blood loss, shorter stay, and faster recovery	Minimally invasive surgery is an anatomic consequence of diagnosis; it is not a universal answer for every giant or central lesion
Robotic vs laparoscopic/open resection	Robotic approaches may shorten postoperative recovery in selected patients, but cost, access, and lesion complexity remain major determinants	Robotics is most persuasive when preoperative imaging defines anatomy precisely and malignancy has been reasonably excluded

HH: Hepatic hemangioma.