Background

Liver fibrosis is the scarring that builds up when the liver is injured over a long period. To study it in rodents, researchers give the animals a substance that damages the liver, then wait.

The problem is in the waiting. The dose and the duration are usually fixed in advance and applied to every animal alike, but animals differ in how they respond, and sex makes a difference too. Some reach the disease stage the study needs long before the schedule says so, and carry on being dosed. Others never get there. Some die before the endpoint. Without a way of looking inside a living animal, there is no way to tell which is which.

Ultrasound offers one. It is routine in human medicine and Eric Felli’s group at the University of Bern, led by Annalisa Berzigotti, is internationally known for human liver ultrasound. Applying it to animals as small as a mouse is a different technical proposition, and there are no standard protocols for it in this field. The alternatives are worse: a biopsy is invasive and cannot be repeated often, blood markers are unreliable, and MRI is expensive and hard to access.

The Zurich Integrative Rodent Physiology group at the University of Zurich specialises in exactly this kind of imaging in rodents. This grant funded three days there, working on cadaveric mice and rats, to learn the anatomy and build a repeatable scanning protocol before any live animal was involved.

Aims

• Learn high-frequency ultrasound imaging of the liver in mice and rats.
• Map the anatomy on cadaveric specimens, identifying the liver lobes, the major blood vessels and the gallbladder, in both males and females.
• Establish which settings and probe positions produce consistent images, and record them as a standard protocol.
• Build the anatomical and technical foundation for later measurements of liver stiffness in living animals.
• Bring the protocol back to Bern for use in the group’s ongoing liver fibrosis work, and share it with collaborators.

Outcome

Over three days at ZIRP, Eric Felli and PhD student Lars Abgottspon worked through the technique from first principles to a complete scanning protocol. The training was built around cadaveric mice and rats, which is what makes the anatomy learnable without any live animal being scanned during the learning phase.

Most of the work was learning to find things. In an animal the size of a mouse, the liver’s blood vessels and the gallbladder sit close together, and recognising them reliably is the skill the whole method rests on. That includes the arrangement radiologists call the Mickey Mouse sign, the shape formed in cross-section by the portal vein with the hepatic artery and bile duct beside it. A drill ran through all three days: remove the probe, reposition it, find the landmarks again. Doing that repeatedly is what makes a measurement taken this week comparable to one taken next month, which is the whole basis of following a single animal over time.

The settings that govern image quality were worked through one at a time, including gain, depth, where the focus sits and how hard the probe presses, each changing what the tissue looks like on screen. Both freehand and scaffold-supported probe positioning were practised, to compare flexibility against stability.

Days one and two ended in dissection. Having imaged a specimen, the team then opened it and compared what they had seen on screen with the anatomy in front of them. Felli notes that this step is rarely part of standard ultrasound training, and it is what turns a two-dimensional image into a three-dimensional understanding of where things are.

Doppler, which measures blood flow, could only be covered in theory, since a cadaver has no circulation. The angles and positioning for the portal vein and hepatic artery were located on the specimens instead, ready for live work later.

Back in Bern, the scanning sequence is now being used during animal procedures, starting from a fixed reference point at the base of the sternum and working through the liver lobes to the main vessels and gallbladder. The same ultrasound platform was already available there. The technique is intended for the group’s SNSF-funded project on liver fibrosis, where the aim is to stop dosing each animal on a fixed schedule and instead measure when it has reached the disease stage the study requires.

Two things still need work. The team knows from the training that anaesthesia and breathing both interfere with imaging in a living animal, but managing that in practice will take several sessions to settle. And with two operators now scanning, keeping their measurements consistent with each other requires continued practice.