Springbok Analytics Publishes Predictive Disease Progression Model for Muscular Dystrophy

Cision Canada

3 days ago

AI-based digital twin modeling predicts patient-specific decline, addressing key trial design and biomarker challenges in FSHD
CHARLOTTESVILLE, Va, July 16, 2025 /CNW/ - Springbok Analytics, a leader in AI-driven muscle analytics, today announced the publication in Scientific Reports of a new disease progression model for facioscapulohumeral muscular dystrophy (FSHD). Springbok's model employs a machine learning–based approach that predicts patient-specific muscle decline, functional outcomes, and fat infiltration using data derived from whole-body MRI.
Click here to read the full paper.
Facioscapulohumeral muscular dystrophy (FSHD) is a progressive, genetic neuromuscular disorder affecting approximately 1 in 7,500 individuals. It is driven by the aberrant expression of the DUX4 protein, which leads to skeletal muscle wasting and functional decline. Although conventional clinical endpoints, such as the 6-minute walk test (6MWT), Timed Up and Go (TUG), and reachable workspace (RWS), are commonly used to monitor disease progression, these measures often lack sensitivity in short-duration trials due to high variability and patient heterogeneity. The recent inability of RWS to differentiate outcomes in a Phase III FSHD trial further underscored these limitations.
"Traditional functional metrics are simply not the best way to measure disease given the complexity and heterogeneity of FSHD," said Silvia Blemker, Ph.D., Chief Scientific Officer and Co-Founder at Springbok Analytics. "What we've shown in this study is that muscle MRI, combined with advanced machine learning, can capture that complexity and make it actionable, supporting more sensitive, predictive, and patient-specific trial designs."
Springbok's advanced modeling combines baseline imaging biomarkers with clinical and functional data to predict how an individual's disease will progress. The result is a patient-specific simulation, or "digital twin," capable of forecasting future declines in muscle structure and performance. This approach outperforms current methods that rely on pooled muscle group metrics or single-slice measurements, which can dilute signal and miss early changes.
"FSHD doesn't follow a predictable pattern from person to person, and treating every patient as if they use the same muscles in the same way prevents therapeutic breakthroughs," said Scott Magargee, CEO and Co-Founder at Springbok Analytics. "Our team has built a model that learns from each person's muscle composition, clinical markers, and functional data to simulate how their disease may progress, establishing a foundation for digital twins and synthetic control arms that is exciting for not only the FSHD community but many other populations that could benefit from precision healthcare."
Study Highlights:
Personalized Prediction: The model accurately forecasted annual changes in fat fraction (a marker of muscle tissue replacement) and lean muscle volume using baseline MRI data.
Functional Correlation: Model predictions aligned with changes in TUG time, a standard test of lower-body mobility, linking imaging biomarkers to meaningful, patient-relevant outcomes.
Enhanced Sensitivity: The study revealed detailed spatial patterns of muscle degradation not visible in grouped or averaged analyses.
"The model doesn't just predict what might happen," added Blemker. "It tells us why, where, and how fast it's happening for each individual. That's a powerful shift in how we approach trial design, and ultimately, how we understand muscle disease biology."
Springbok's AI-powered platform, which received FDA 510(k) Clearance last year, rapidly analyzes dozens of individual muscles from a single MRI scan, quantifying volume, asymmetry, and fat infiltration. The technology is already integrated into multiple neuromuscular trials, including a first-in-human study of Epicrispr Biotechnologies' investigational therapy, EPI-321.
As interest in MRI biomarkers grows across rare and neuromuscular disease trials, this publication establishes a framework for incorporating high-resolution, AI-powered models into trial design, patient stratification, and biomarker development.
To explore how Springbok's platform can support your trial or development program, contact: [email protected].
About Springbok Analytics:
Springbok Analytics is a leading muscle health analytics company dedicated to advancing health and performance outcomes through innovative, AI-driven solutions that deliver a clearer, more comprehensive view of muscle health.
Built on more than 15 years of research and scientific validation, Springbok's FDA-cleared technology transforms MRI data into personalized, 3D visualizations of muscle health. These detailed analyses provide precise metrics, including individual muscle volume and composition, fat infiltration, asymmetries, scar tissue, edema, and tendon morphology. By offering a more accurate and complete understanding of musculoskeletal health, Springbok enhances diagnostic accuracy, treatment monitoring, research capabilities, and performance optimization.