Estimation of Intracranial Pressure Using Non-invasive Fundus Videos

Intracranial pressure (ICP) is based on the simple principle of Monroe-Kellie Doctrine. Since the volume of the cranial cavity is constant and the brain, CSF and blood can only compensate for a certain change in volume, pressure will increase at a critical increase in volume. An increase in pressure caused by an increase in intracranial volume is distributed throughout the entire brain and spinal canal equally. The normal range of ICP varies with age. Normal values are less than 10-15 mmHg for adults and adolescents, 3-7 mmHg for younger children and 1.5-6 mmHg for newborns even sub-atmospheric ICP has been shown.

The threshold for intracranial hypertension varies in the literature, which is reflected in the clinical use. Generally, ICP above 20 mmHg is defined intracranial hypertension. ICP higher than 20-25 may require treatment, and continuous ICP above 40 mmHg indicates severe, lethal intracranial hypertension.

Some forms of measurements, e.g. 1. Intraparenchymal strain gauge or fiber optics or 2. Intraventricular ICP monitoring via external ventricular drain (EVD). EVD is considered golden standard, but in the daily clinic there is often no or limited difference between 1. and 2. The issue regarding invasive methods are the same as for any surgical procedure, e.g. bleeding and infection. Risk of bleeding is 2-10% for EVD and 2.5% for intraparenchymal monitors. The consequences of such a bleed can be severe and result in disability. EVD has a higher risk of infection (5-20%) compared to intraparenchymal monitors (0-0.1%).

Non-invasive methods such as transcranial doppler ultrasound pulsation index and optic nerve sheath diameter has been used in studies, which seemed promising, but has not yet been validated for routine clinical use. Also, in non-invasive methods there are significant inter observational variation.

Studies of the eye's fundus have shown a correlation between ICP and ratio between artery and vein diameter (A/V ratio, unpublished data) Currently, there are no published work that has considered the correlation between A/V ratio and ICP, however there is a study with a theoretical prediction of such a correlation (Babbs CF 2016. Weldon School of Biomedical Engineering Faculty Working Papers. Paper 1 http://docs.lib.purdue.edu/bmewp/1).