Comparative analysis of intracranial pressure prediction models using optic nerve sheath diameter and additional ultrasound metrics

Optic nerve sheath diameter (ONSD) assessment of elevated intracranial pressure is increasingly emphasised as a noninvasive modality. Nevertheless, quantitative assessment remains limited, and study criteria vary. We constructed four prediction models based on ONSD and compared their accuracy in predicting intracranial pressure (ICP). Changes in the ONSD during the normalization of intracranial pressure in patients with increased intracranial pressure were investigated. We recruited 220 patients who completed lumbar puncture (LP) (model groups, n = 154; test groups, n = 66). Sixty-six patients with increased intracranial pressure were successfully followed up. Four mathematical equations were established in the model group (n = 154): 1. the ONSD combined OND model: predicted ICP = 10.236 × ONSD–7.986 × OND–20.151; 2. ONSD model: predicted ICP = 7.966 × ONSD – 29.62; 3. ΔONSD model: predicted ICP = 10.32 × ΔONSD-14.45; 4. ONSD/ETD model: predicted ICP = 119.5 × ONSD/ETD-14.48. In the test group (n = 66), the ONSD combined OND model achieved the largest ICC (0.89), showing better predictive consistency. During follow-up, in patients with intracranial pressure exceeding 20 mmHg, ROC analysis indicated ONSD > 5.83 mm and ΔONSD > 3.21 mm could act as potential screening cut-offs for ICP > 20 mmHg. Bedside ultrasound ONSD combined with OND may serve as a supplementary non-invasive ICP assessment method. The ONSD combined OND model and the ΔONSD models outperformed single ONSD prediction, which may suggest ICP fluctuation influences the optic nerve sheath more prominently than the optic nerve. It should be noted that when the ONSD combined OND indicators to evaluate the therapeutic effect in patients with intracranial hypertension,the higher the intracranial pressure, the slower the recovery of OND, and this asynchronous phenomenon must be considered.