Dystonia UK was pleased to award seed funding in 2016 to Professor Ian Loram and his team at Manchester Metropolitan University for their research project. In this article, Prof Ian Loram explains the learnings and next steps for turning the proof of concept into a robust clinical tool.

Dystonia is a neurological movement disorder caused by incorrect messages from the brain to the muscles. Symptoms for cervical dystonia (CD) include uncontrollable muscle contractions in the neck which cause awkward postures, discomfort, and pain. Following diagnosis, Botulinum NeuroToxin (BoNT) treatment is the accepted standard of care for patients with CD. When the dystonic muscles are injected accurately the relief and improvement in quality of life can be very great1.

The neck though is a complex structure containing many layers of muscles and other delicate structures including arteries, veins, and nerves (Figure 1 Neck muscles). To identify dystonic muscles and inject effectively requires detailed understanding of this complex three-dimensional structure. Use of ultrasound imaging (US) to guide injections has been proven to improve treatment outcomes5,6. However, expertise to interpret ultrasound images of the neck and use US to guide injections takes a long time to acquire7.

Therefore, people with dystonia often experience a shortage of medical professionals with the specialist training to understand and treat dystonia2. This shortage means people are waiting too long to receive a diagnosis, there are too few clinical staff injecting BoNT, and it takes too long to create new trained injectors. Commonly, there are long waiting times for treatment and up to 40% of people with CD discontinue long-term BoNT treatment following perceived lack of efficacy3,4.

Figure 1. Neck muscles: deep to surface layers from left to right.

Studying US and muscle anatomy in relation to dystonia, Prof Ian Loram and Dr Ryan Cunningham at Manchester Metropolitan University (MMU) in collaboration with neurologists at Salford Royal NHS Foundation Trust led by Dr Chris Kobylecki conducted a seed project in 2016 supported by Dystonia UK. Their aim was to develop a method for visualization of neck muscles boundaries using ultrasound to accelerate training of clinicians and increase accuracy of injections. 35 volunteers with cervical dystonia and 25 age matched volunteers without dystonia participated in this study. Participants agreed to ultrasound and MRI recording of their neck muscles, motion analysis of their posture and movement and a clinical assessment. Using this data, the researchers annotated meticulously thousands of ultrasound images and trained an artificial intelligence system to recognize automatically the muscles in an ultrasound image (Figure 2).    

Figure 2.  Neural Network analysis of original ultrasound image (left) to generate automatically the muscle boundaries and labels that are added (right).

This method demonstrates automated identification and display neck muscles in real time in the ultrasound image. The value of this approach is its potential to train clinicians to interpret the muscle structures within the ultrasound images so they can become proficient in injecting. This “proof of concept” was published in a prestigious IEEE Journal8 (Figure 3). 

Figure 3. Tool for real-time analysis and visualization of cervical muscles for cervical dystonia

Link to Publication and Media (Video) demonstrating proof of concept

Starting from this work, the research group funded by a pharmaceutical company decided to translate the proof of concept to a robust tool meeting clinical standard. This is the aim of the current project, “A clinical tool for real-time visualization and analysis of neck muscles for cervical dystonia”. Generally, this tool will enable an expanded pool of people who understand cervical dystonia and are trained to treat it. The main benefit will be to facilitate the training of injectors to meet the needs of local clinics. The tool should also help to guide injections, resulting in more precise treatment with potentially fewer side effects and improved outcomes. This should result in an increased number of local clinics so people do not have to travel so far and should lead to reduced waiting times to receive BoNT injections. This current project is focused on cervical dystonia, however, we expect that in the future people with other forms of dystonia will also benefit.

Your help is needed! Join our study and support the development of this tool for people with cervical dystonia…

Manchester Metropolitan University logo

  1. Albanese, A. et al. Practical guidance for CD management involving treatment of botulinum toxin: a consensus statement. J. Neurol. 262, 2201–2213 (2015).
  2. Benson, M. et al. Development of a patient journey map for people living with cervical dystonia. Orphanet J. Rare Dis. 17, 1–9 (2022).
  3. Gill, C. E. et al. Continuation of long-term care for cervical dystonia at an academic movement disorders clinic. Toxins (Basel). 5, 776–783 (2013).
  4. Jinnah, H. A., Comella, C. L., Perlmutter, J., Lungu, C. & Hallett, M. Longitudinal studies of botulinum toxin in cervical dystonia: Why do patients discontinue therapy? Toxicon 147, 89–95 (2018).
  5. Tyślerowicz, M., Dulski, J., Gawryluk, J. & Sławek, J. Does Ultrasound Guidance Improve the Effectiveness of Neurotoxin Injections in Patients with Cervical Dystonia? (A Prospective, Partially-Blinded, Clinical Study). Toxins (Basel). 14, (2022).
  6. Fietzek, U. M. et al. The role of ultrasound for the personalized botulinum toxin treatment of cervical dystonia. Toxins (Basel). 13, (2021).
  7. Lagnau, P. et al. Ergonomic Recommendations in Ultrasound-Guided Botulinum Neurotoxin Chemodenervation for Spasticity: An International Expert Group Opinion. Toxins (Basel). 13, (2021).
  8. Loram, I. et al. Objective Analysis of Neck Muscle Boundaries for Cervical Dystonia Using Ultrasound Imaging and Deep Learning. IEEE J. Biomed. Heal. Informatics 24, 1016–1027 (2020).


Published: 9th August 2023