Neuromuscular disorders are diseases that affect the muscles and the motor neurons, which facilitate the central nervous system’s communication with the muscles. These diseases cause a decline in neuromuscular communication and the degradation of muscle tissue overtime. These disorders are not curable, but there are a wide array of treatments and therapies that are geared toward improving the quality of life for affected patients, but testing their effectiveness is difficult to do with the current technology, as methods of monitoring gait, or walking pattern, leave out several crucial variables. Stevens professor Damiano Zanotto and professor Jaqueline Montes of Columbia University Irving Medical Center (CUMIC) are working to develop AI techniques to analyze in-shoe sensor data to better analyze the progress of children living with neuromuscular disorders. The National Institute of Health (NIH) has awarded a $2 million multi-principal investigator R01 award to support this study, and several of the nation’s top medical care providers are also involved.
This study focuses on two disorders: Duchenne muscular dystrophy, a genetic disorder that leads to progressive muscle weakness in young boys, and spinal muscular dystrophy (SMA), which causes a loss of motor neurons, and in turn, the deterioration of muscle tissue. SMA is the leading cause of infant mortality in the United States, as it affects the muscles responsible for speaking, swallowing, and breathing. These diseases affect roughly one million children worldwide. These, as well as other neuromuscular disorders, significantly affect movement and gait. In the worst cases, these diseases can be fatal.
Monitoring gait is important when assessing an individual’s response to treatment. Muscle strength and proprioception, which is your body’s ability to tell where your limbs are in relation to one another, are linked to good posture and a consistent and balanced gait. Problems with voluntary movement translate to disruptive gait, which in turn can damage joints or increase the risk of falling. Existing techniques for monitoring physical activity include step counters and distance trackers, neither of which give an accurate view of “‘real-world performance nor the improvement in or progression of a disease,’” says Professor Zanotto. They cannot assess gait quality, “‘such as stride length, time, and velocity, nor kinetic metrics that evaluate a person’s dynamic stability.’”
The AI-enhanced in-shoe sensors previously developed by Professor Zanotto capture not only gait quality but also “unique kinetic data — such as the force of a foot striking the ground,” which other technologies can’t do. These were developed with the support of the National Science Foundation, Muscular Dystrophy Association and New Jersey Department of Health. In the previous study, the sensors were tested on 600 individuals, only a small number of whom were healthy individuals. This meant that there was insufficient control data to which data of affected individuals could be compared.
The NIH study is the first time this technology will be used to monitor a patient’s gait for one full month. This study will be conducted at clinical sites at CUIMC, Boston Children’s Hospital (which is affiliated with Harvard Medical School) and Stanford Medical School. Professor Montes explains that the team plans to “enroll up to 100 individuals—33 with SMA, 33 with DMD and 40 healthy controls—across the 3 clinical sites.” With ambulatory patients as young as 5 years old, and a balance of child, adolescent, and adult participants, the team intends to employ fairness and accuracy in the acquisition of data and its analysis.
This study plans to collect data from two one-month periods one year apart. With both everyday monitoring and benchmark clinical assessments, the group hopes to isolate the portions of the motion data that give the most information about the condition of a patient in relation to the progression of their disease. Dr. Zanotto notes the effect the study will have on drug development in the future. Having more sensitive outcome parameters will allow future clinical trials of drugs to use smaller sample sizes and run more quickly. This means that results can be acquired sooner and treatments can be given to patients in a quicker and more cost-efficient way. In addition, the team hopes to use the collected data to train machine-learning algorithms that can predict the progress of a patient’s disease months into the future.
