Jul 6 2026
In this interview, News Medical-Life Sciences speaks with Professor Emma Hodson-Tole, Professor of Neuromuscular Biomechanics at Manchester Metropolitan University, about the urgent need for improved diagnostic tools for motor neuron disease (MND), and how her team's research is combining ultrasound imaging, computer vision and artificial intelligence to create a non-invasive approach that could reduce the need for painful needle examinations while supporting earlier diagnosis and improved disease monitoring.
Image credit: Sheeyla/Shutterstock.com
From your perspective, what are the biggest scientific and clinical knowledge gaps that still limit our understanding of motor neuron disease today?
Motor neuron disease is an umbrella term describing a group of progressive neurodegenerative diseases that affect the communication between nerves and muscles throughout the body. As these nerve cells deteriorate, muscles become progressively weaker and smaller, making everyday activities such as walking, lifting objects, speaking, swallowing and breathing increasingly difficult. The most common form is amyotrophic lateral sclerosis, which often progresses rapidly. Although many people think of MND as a very rare disease, the lifetime risk is around one in 300 people.
One of the biggest challenges is that we still do not understand the cause of MND. While some cases have a strong genetic component, the majority appear to develop without any obvious family history. This suggests that multiple genetic, environmental and lifestyle factors interact in ways that we do not yet fully understand. Without knowing these mechanisms, we cannot predict who will develop the disease or intervene before symptoms appear.
The disease is also highly heterogeneous. Every person's experience is different, suggesting that different biological mechanisms may underlie similar clinical symptoms. Until we understand these differences, we cannot reliably stratify patients into biologically meaningful groups. This has major implications for clinical trials because variations between patients can mask whether potential treatments are truly effective.
Could you explain the current diagnostic pathway and where you believe there is the greatest need for more patient-friendly approaches?
There is currently no simple biological test for MND, such as a blood test. Diagnosis therefore relies on a combination of clinical assessment and investigations designed to exclude other neurological conditions.
Early symptoms can be very subtle. People may notice they trip more often, struggle with fine hand movements, or experience slurred speech at the end of the day. Because MND most commonly affects people over the age of 50, these symptoms are often initially attributed to normal aging, delaying diagnosis.
The diagnostic process usually includes taking a detailed medical history, assessing muscle strength, reflexes, coordination, and speech, as well as blood tests and MRI scans. A key part of the pathway is electromyography, where needles are inserted into muscles throughout the body to record electrical activity. Patients consistently tell us this is the most unpleasant part of the diagnostic process.
Unfortunately, many people have to undergo repeated electromyography examinations because there is insufficient evidence to make a definitive diagnosis on the first assessment. This prolongs uncertainty for patients and delays access to clinical trials, which currently represent the only opportunity for many people to receive experimental treatments that may slow disease progression.
There is therefore a clear need for tools that reduce reliance on repeated needle examinations while helping clinicians diagnose the disease earlier and monitor progression in a way that is comfortable for patients.
What inspired you and your team to develop a non-invasive approach for monitoring motor neuron disease, and how has your background in neuromuscular biomechanics influenced this work?
My research has always centered on understanding how muscle structure influences movement and function. As part of this work, we were already using ultrasound imaging to study muscle behavior during movement. Ultrasound is widely available, inexpensive compared with MRI, does not expose patients to radiation, and is completely pain-free, making it an attractive clinical tool.
At the same time, we were developing computer software to automate measurements from ultrasound images. Around then, changes to the diagnostic criteria for MND meant that involuntary muscle twitches, known as fasciculations, became increasingly important during diagnosis.
We knew from published research that these fasciculations could often be detected using ultrasound and, for some muscles, ultrasound could even outperform electromyography. However, clinicians had no automated software capable of analyzing ultrasound videos and detecting these twitches objectively. Instead, videos had to be reviewed manually, a time-consuming process that can introduce variability across clinicians and hospitals.
We recognized an opportunity to combine our expertise in ultrasound image analysis with this unmet clinical need, creating automated tools that could make ultrasound a practical and consistent method for supporting MND diagnosis.
Could you explain how your technology works and which biomechanical features you are analyzing to assess motor neuron disease?
Our current work primarily focuses on detecting involuntary muscle twitches in ultrasound videos. We use computer vision techniques to analyze muscle movement throughout the image sequences, and then combine these measurements with machine learning approaches that classify whether the observed muscle characteristics are consistent with motor neuron disease.
Importantly, the ultrasound videos contain much more information than muscle twitches alone. They also capture changes in muscle size, shape and architecture, all of which influence muscle function and strength.
Dr. Emma Hodson-Tole | Research Profiles
Video credit: MMU ENGAGE/Youtube.com
How does this non-invasive approach compare with traditional needle electromyography in terms of diagnostic performance and monitoring disease progression?
Our early studies have shown that the ultrasound image analysis techniques compare very well with needle electromyography across different muscles throughout the body. Typically, we have achieved accuracies above 85% when comparing the two techniques.
We are still in the early stages of combining image analysis with machine learning for disease classification, but our initial findings are very encouraging and indicate high classification accuracy.
It is important to recognize, however, that these studies have so far involved people who have already received a diagnosis of motor neuron disease, compared with healthy volunteers. At this stage, the disease is relatively advanced, and the clinical signs are much more obvious.
Our next priority is to study people while they are still going through the diagnostic process. This will allow us to determine how sensitive the technology is during the earliest stages of disease and whether it can support earlier diagnosis. It will also enable us to test the tools against other neurological conditions that can mimic MND, such as multiple sclerosis or a trapped nerve, to ensure they accurately distinguish between disorders.
Alongside this work, we have recently completed longitudinal data collection from people living with MND, who have kindly returned for repeated assessments approximately every three months. We are now beginning to investigate how well our methods can monitor disease progression over time, which is another area where objective, non-invasive measurements could make a significant contribution.
From both the patient and clinician perspective, how could this technology change the clinical workflow for diagnosing and monitoring motor neuron disease?
Needle electromyography provides valuable clinical information that ultrasound cannot completely replace, so our ambition is not to eliminate needle testing altogether.
Instead, we want ultrasound to become an additional clinical tool that clinicians can use quickly and easily during routine appointments. The information gathered from ultrasound could help determine whether a needle examination is actually necessary or whether sufficient evidence already exists to support a diagnosis.
If successful, this approach would reduce the number of painful needle examinations many patients currently undergo. Equally important, if ultrasound enables earlier diagnosis, it could shorten the period during which patients live with uncertainty about the cause of their symptoms.
Earlier diagnosis also has practical implications, as it allows patients to enter clinical trials sooner and gain earlier access to experimental therapies that may slow disease progression.
What are the next milestones for this technology, and what types of collaborations will be most important for translating it into routine clinical practice?
Our immediate priority is to begin collecting data from people before they receive a diagnosis. That will allow us to evaluate how well the tools perform in the earliest stages of disease, when they have the greatest potential for clinical impact.
At the same time, we are analyzing the longitudinal datasets we have already collected to understand how effectively the technology can monitor disease progression and whether additional analytical approaches could improve those measurements.
From a collaboration perspective, we are particularly interested in working with both industrial and clinical partners. We would welcome partnerships with organizations experienced in biomedical imaging technologies and companies that supply tools to NHS clinical settings. Their expertise would be invaluable in helping develop these research tools into products suitable for routine healthcare use.
We are equally keen to strengthen collaborations with neurophysiologists and clinicians involved in diagnosing motor neuron disease and related neurological conditions. Their input is essential to ensure the technology integrates naturally into existing clinical workflows and delivers the features clinicians need in everyday practice.
Developing these partnerships is also critical when applying for future funding, as demonstrating a clear pathway towards clinical translation is increasingly important for securing research investment.
How are advanced data analysis and AI contributing to your work, and why is explainability so important?
Advanced data analysis and AI are playing a key role in our current work. We use image analysis techniques to identify ultrasound features we already know are clinically relevant, and machine learning algorithms then use those features to classify whether muscles are affected by motor neuron disease.
An important aspect of our research is ensuring these AI systems remain explainable. It is not enough for an algorithm simply to provide an answer. We also need to understand which aspects of the ultrasound data contributed to that decision so clinicians can have confidence in the results and interpret them appropriately.
Ultimately, our objective is not to replace clinicians with artificial intelligence. Instead, we want to develop tools that provide objective, evidence-based information to support clinical decision-making, improve patient experience, and help determine whether additional investigations are required or whether sufficient evidence already exists to make a diagnosis.
Save this interview for later. Download the full free PDF and explore the latest advances in non-invasive motor neuron disease research.
You recently attended the Bionow BioAI Summit. What did you find most valuable about the meeting, and how do events like this help advance motor neuron disease research?
The most valuable thing about attending the Bionow BioAI Summit was the opportunity to connect with potential collaborators and gain feedback. It also provided a platform from which we could raise awareness of MND, and the need for improved diagnosis and effective treatments. If even one group leaves the meeting recognizing that their expertise could support challenges faced by the MND community, then that is a brilliant result.
The problems associated with motor neuron disease are incredibly complex, and no single research group can solve them alone. Many pieces of the puzzle need to come together before we can truly transform patient outcomes.
Our work addresses one part of that challenge by developing better diagnostic tools. The more researchers, clinicians, engineers and industry partners who become involved in tackling different aspects of the disease, the greater the likelihood that we will develop solutions capable of improving lives. Events such as the Bionow BioAI Summit play an important role in bringing those communities together and accelerating progress towards making motor neuron disease a treatable condition, or ultimately a disease of the past.
Where can readers find more information?
About the Researcher
Professor Emma Hodson-Tole is Professor of Neuromuscular Biomechanics at Manchester Metropolitan University and an internationally recognised researcher in muscle function, biomechanics and neuromuscular imaging. She leads multidisciplinary research that combines biomechanics, medical ultrasound, image analysis and artificial intelligence to improve understanding of muscle function in health and disease, with a particular focus on motor neuron disease.
Professor Hodson-Tole completed her postgraduate research in biomechanics and neuromuscular function. Throughout her academic career she has developed expertise in musculoskeletal imaging, muscle architecture, movement analysis and computational approaches for analysing complex biological data. Her work has been supported by major UK research funders and involves close collaboration with clinicians, engineers, computer scientists and patient groups.
Her current research aims to develop objective, non-invasive ultrasound technologies capable of supporting earlier diagnosis and improved monitoring of motor neuron disease. By integrating advanced image analysis and explainable artificial intelligence with established clinical imaging techniques, Professor Hodson-Tole's work seeks to reduce reliance on invasive diagnostic procedures while improving patient experience and enabling more sensitive assessment of disease progression. Her research reflects a strong commitment to translating engineering and biomechanical innovation into practical clinical tools that can improve outcomes for people living with neurological disease.
New microneedle patch offers enzyme-free continuous glucose monitoring