University of Bath

BLADDERSENS: Detecting bladder volume and pressure from sacral nerve signals

An EPSRC-funded research project detecting bladder volume and pressure to aid the management of the urinary bladder following spinal cord trauma.

Managing the urinary bladder after trauma to the spinal cord is a top priority for clinicians and patients. In the past, kidney damage due to high bladder pressures and/or infection was a common cause of death following such an injury. Infections still raise mortality and morbidity, made worse by the increasing risk of antibiotic resistance.

GS Brindley developed a neuroprosthesis for controlling the bladder after spinal cord injury (SCI). This provided a cheaper alternative of achieving complete voluntary micturition (CVM) than intermittent sterile catheterisation. The Brindley method employs sacral anterior root stimulation (SARS) but it is not a popular method. Instead, a new neuroprosthesis that doesn't require rhizotomy is needed. But, a key component of such a new neuroprosthesis hasn't been developed or tested yet.

Our project is developing the remaining systems to enable the specification and design of a future complete closed loop bladder prosthesis.

Improving on the Brindley system

The two functions of the bladder are storage and voiding. Several improvements on the Brindley stimulator have been proposed to manage storage and voiding without rhizotomy. We can manage incontinence by using a technique called neuromodulation. This suppresses the voiding reflex by using low intensity electrical stimulation of the large afferents in the pudendal nerve or sacral posterior roots.

As in the Brindley system, stimulating the sacral nerve roots (extradural or intrathecal) produces bladder contractions. However, there will also be sphincter contractions that prevent micturition if this is done without a rhizotomy. Using a high-frequency current (about 2kHz) to block the pudendal nerves bilaterally, prevents the unwanted contractions. This blockade is reversible and is only needed during voiding. This high-frequency blocking is already used in devices for blocking pain signals after amputation.

Yet, despite these developments, there are still two missing requirements needed for a complete, practical neuroprosthetic. There is no practical method for detecting the onset of bladder contractions and triggering neuromodulation in a chronic implant. There is also no method to 'inform' the patient of how full their bladder is and when it should be emptied.

One solution is to insert small absolute pressure sensors into the bladder through the urethra. But there is a drawback to this in chronic impairment of the sensors due to encrustation. Also, the extreme change in shape as the bladder fills and empties makes it difficult to an intramural device.

But surgically-implanted electrodes are necessary anyway and the nerves that innervate the bladder adapt with the bladder shape. So an innovative approach is to use the bladder afferant neural signals themselves. Afferent fibres from the bladder wall reach the spinal cord via the posterior sacral roots and are part of the mixed extradural root.

Previously, we investigated using tripolar microchannels to detect bladder activity at the intrathecal roots in rats. The method requires that the dura around the spinal cord is open (durotomy). This presents several risks such as:

  • damage to the delicate and intrathecal roots at that level
  • post-operative spinal fluid leak
  • long-term damage to the roots themselves

This is a significant deterrent to uptake of the new device.

Developing the velocity selective recording (VSR) method

The extradural roots (most often S3) that are within the sacrum 'outside the dura' offer an alternative site for the electrodes. They are routinely implanted here by some surgeons in the Brindley procedure. This involves removing the bony roof of the sacrum (a laminectomy), which is routine practice.Our project investigates if we can estimate bladder pressure and volume using electrodes implanted at this extradural site, both acutely and chronically. We will use sheep for these experiments as the nerves are more like those in humans.

The conduction velocity of the sensory (myelinated) afferent fibres from mechanoreceptors of the genital region in sheep is ~41.4 m/s (+/- 14.7 m/s). This is similar to bladder afferents coding for pressure and volume information in humans, 38 and 41 m/s respectively.

We propose using a method we are developing called velocity selective recording (VSR) to discriminate between these signals. This method may make this possible without using nerve interfaces that are difficult and risky to implant.

The principle of VSR is based on the fact that action potentials (APs) propagate at specific velocities. And that these velocities are related to the nerve fibre function (for myelinated nerves).

Recording the velocity spectrum of neural traffic will provide a signature relating the recorded signals to their function. It will also discriminate between afferent and efferent traffic on a mixed nerve, such as extradural bladder nerves. The method requires multi-electrode cuffs (MECs). We have used such devices with as many as 11 electrodes along the length of a nerve cuff.

We have established VSR as a powerful technique to analyse electrically evoked electroneurogram (ENG). We have further developed the method to measure the AP rate in a number of bands of propogation velocity. This is important as it allows us to measure the intensity of neural activity (e.g. bladder pressure) as well as identify the presence and origin of the signal.