Abstract:
The aim of this thesis was to investigate anterior related eye conditions in the in-vivo human
lens due to eye conditions related to the writer using computational fluid dynamics (CFD)
methodology.
A three-dimensional model of the in-vitro rodent lens was created in a commercially renowned
CFD software – COMSOL Multiphysics. Generally speaking, this showed good agreement to
available experimental data and previous work by Vaghefi [1].
This in-vitro Rodent (mouse) model was then used as a baseline to extend onto a novel threedimensional
in-vivo human lens model that studies the fluid dynamics, along with the inclusion
of the iris and anterior/ posterior chambers of the eye to create a more realistic in-vivo human
lens. This model was fine tuned by investigating the differences between a 4 partition setup and
a 2 partition setup, it was found that the 2 partition setup seems to match more closely with
available data in the in-vivo human lens.
This model was then used to determine a range of anterior eye related conditions such as effects
of sutural changes with ageing, effects of iris-lens channel closure and effects of vitrectomy, as
well as an interest case of motionless vitreous study.
To study the effects of sutural changes with ageing, aged lenses of different shape and sizes
(7-year old, 28-year old and 46-year old) with increasing complex sutural patterns were used.
It was found that anatomical changes of sutures alone with ageing increase its surface area, to
enhance fluid movement between extracellular and intracellular domains of the lens. If the lens
sutures did not ‘grow’ with ageing, lens microcirculation could weaken by a factor of 4 at age 47.
Furthermore, lens microcirculation system using the whole anterior eye component has shown
that it does weaken with ageing, with decreased fluid velocities and intraocular pressure, but
this effect is not as dramatic as what we observed from analysing the sutures alone.
To study the effects of iris-lens channel closure, we modelled the normal iris-lens channel opening
(7 μm) and three different levels of iris-lens channel closure (5, 3 and 0 μm). It was shown that
as long as the iris-lens channel is open (i.e. > 0 μm), the microcirculation system appears to be unaffected. When the iris-lens channel is fully closed, the pressure inside the lens increases.
To study the effects of vitreous fluid property changes to mimic vitrectomy surgery/ retinal
detachment procedures, we modelled one case in saline and another in silicone oil. It was shown
that replacing the vitreous with saline increases the posterior portion of the microcirulation
system, bringing more fluid from the posterior cavity into the lens whilst still having an anterior
flow effect. In contrast, replacement of vitreous with silicone oil lead to a full domination of
posterior flow into the lens tissue.
As an interest case, we also studied the impractical but still interesting case of motionless
vitreous. We observed that this condition led to almost zero posterior lens microcirculation (i.e.
fully driven by anterior chamber flow) and an appropriately disturbed lens pressure gradient.
In conclusion, the microcirculation system has been studied computationally in the in-vivo
human lens for the first time, and although there are limitations with this model as it was not
able to calculate hydrostatic pressure as per recent discoveries. It aligns in fluidic pattern mostly
with literature data. Further improvements of the model to include new features to correct for
this hydrostatic pressure and inclusion of mechanical modelling will be of great value to the lens
research community.