Understanding the ABC2 Formula
The ABC2 method simplifies intracranial hemorrhage quantification by reducing a 3D bleeding mass to its principal axial dimensions. Rather than requiring manual planimetry or specialised software, clinicians measure two perpendicular axes on the largest bleeding slice, note the slice thickness from the DICOM header, and count participating slices weighted by coverage percentage.
The formula accounts for shape factor—a correction that distinguishes compact ellipsoid bleeds (shape = 2) from irregular, fragmented, or multilobulated haemorrhages (shape = 3). This distinction reflects the actual geometric packing of blood within the parenchyma. An ellipsoid shape factor yields smaller volume estimates and correlates better with actual volumetric measurements in rounded bleeds.
The ABC2 approach trades some precision for speed and reproducibility, making it ideal for acute settings where rapid triage and communication of bleed size between teams are essential.
ABC2 Volumetric Calculation
The calculator combines two key calculations: first, a weighted slice count that adjusts for partial involvement of individual slices; second, the core volumetric formula that incorporates axial dimensions, slice thickness, and geometric correction.
Weighted slice count = (slices ≥75%) × 1 + (slices 25–75%) × 0.5 + (slices <25%) × 0
Volume (mL) = Length × Width × Slice thickness × Weighted slices / Shape factor
Length (A)— Maximum anteroposterior or mediolateral dimension of bleeding on the largest slice, in centimetres.Width (B)— Perpendicular dimension of bleeding on the same slice, measured orthogonally to Length.Slice thickness (C)— DICOM reconstructed slice thickness or spacing, typically 2–5 mm; convert to centimetres for consistent units.Weighted slices— Sum of slice counts, where complete coverage counts as 1, 25–75% coverage as 0.5, and <25% coverage as 0.Shape factor— 2 for round or ellipsoid haemorrhage; 3 for irregular, separated, or multinodular bleeding patterns.
Common Causes of Parenchymal Haemorrhage
Uncontrolled hypertension remains the leading cause of spontaneous intracerebral bleeding, accounting for roughly 50–60% of cases. Chronic high blood pressure weakens small penetrating arteries in the basal ganglia, pons, and cerebellum, eventually causing rupture.
Anticoagulant and antiplatelet medications—including warfarin, dabigatran, rivaroxaban, and dual antiplatelet therapy—substantially raise bleeding risk in susceptible patients. Other significant contributors include:
- Cerebral amyloid angiopathy: abnormal protein deposits in cortical vessel walls, particularly in older adults, leading to lobar haemorrhages.
- Vasculitis and arteriovenous malformations: structural vascular lesions prone to rupture.
- Stimulant abuse: cocaine and amphetamines can acutely elevate blood pressure and trigger haemorrhage.
- Coagulopathy: severe liver disease, disseminated intravascular coagulation, or inherited bleeding disorders.
- Haemorrhagic transformation: secondary bleeding into an ischaemic stroke.
Clinical Interpretation and ICH Severity
Volume alone does not determine outcome, but it is a powerful predictor. A bleed <30 mL often has a favourable prognosis if brainstem and cerebellar structures are spared. Volumes of 30–60 mL represent moderate to large bleeds requiring intensive monitoring and possible evacuation. Haemorrhages exceeding 60 mL, or those occupying >40% of hemispheric volume, carry substantially higher mortality and morbidity.
Location matters critically: deep haemorrhages (basal ganglia, thalamus) and brainstem bleeds are far more dangerous than lobar or cerebellar ones, even at identical volumes. Intraventricular extension, mass effect causing midline shift, and hydrocephalus from ventricular obstruction all worsen prognosis independent of raw volume figures. The ICH Score, which incorporates volume, location, intraventricular haemorrhage status, infratentorial location, and Glasgow Coma Scale score, provides a more nuanced mortality estimate.
Practical Measurement Tips
Accurate dimensions and slice classification directly impact the volume estimate and clinical communication.
- Standardise measurement planes — Use the same CT slice where both maximal length and perpendicular width are visible. Do not measure length on one slice and width on another; this introduces systematic error. Most PACS software provides electronic callipers; use them consistently and document your measurements in the clinical note.
- Handle partial slices carefully — Slices with haemorrhage occupying 25–75% of the area represent a substantial proportion of clinical cases. If uncertain whether a slice is 25% or 75%, lean toward the lower category (0.5 weight) rather than full weight. Consistent bias is preferable to random assignment.
- Distinguish shape factors accurately — A haemorrhage confined to one lobe with smooth margins is ellipsoid (factor = 2). Multiple separate foci, a burst distribution pattern, or extension into the ventricular system warrant the irregular category (factor = 3). When in doubt, using shape = 3 yields a more conservative (larger) estimate.
- Convert units before calculation — Ensure length and width are in centimetres and slice thickness in centimetres (5 mm = 0.5 cm). Mixing millimetres and centimetres is the most common hand-calculation error and will yield volumes off by an order of magnitude.
Limitations and When to Seek Specialist Input
The ABC2 formula is a validated but inherently approximate tool. It performs less accurately on very small (<5 mL) or extremely large (>100 mL) bleeds, where the ellipsoid or irregular model diverges further from true geometry. Bleed shape deformation by surrounding oedema, mass effect, or ventricular compression can reduce measurement reproducibility.
Semi-automated or fully automated volumetric analysis using imaging software provides superior precision and should be considered for borderline clinical decisions—such as the threshold for neurosurgical evacuation—or in research settings. Additionally, the ABC2 method does not account for perihematomal oedema, which peaks days after ictus and significantly affects mass effect and clinical deterioration despite stable bleed volume.
Serial CT imaging at 24 hours post-ictus is recommended to detect haemorrhage expansion, defined as growth >6 mL or >33% of baseline volume, as expansion substantially worsens prognosis and may trigger more aggressive intervention.