Unique mathematical models.
There is currently no biomarker that can indicate Parkinson’s disease severity. Assessments are based on visual interpretation of motor symptoms, mapped against a standardized scale. Rating the severity of Parkinson’s disease requires a highly skilled neurologist, specialized in movement disorders. Therefore, rating scales are seldom successfully used outside of specialized clinics.
ANLIVA™ will produce clinical biomarkers from sensor data: Reproducible, reliable indicators of Parkinson-related symptoms. The platform is the result of several years of evidence-based product development in cooperation with clinicians and key opinion leaders. The idea to support the daily challenge of tracking Parkinson’s disease symptom progression was born from the needs of the clinic.
Our core algorithms have evolved from more than a decade of research collaborations and show promising preliminary results.
How does ANLIVA™ work?
ANLIVA™ transforms sensor- and image data into objective, clinical biomarkers to assist quantification and tracking of Parkinson’s symptoms. Our proprietary algorithms combine sensor data with advanced mathematical models of physical reality.
At launch, we will capture and interpret eye motor control and involuntary movements (dyskinesias) to track disease progression and treatment efficiency. More symptom inputs will be announced, with eye blink frequency next on our roadmap.
Symptom measurements are captured using off-the-shelf hardware such as smartphones and eye-tracking cameras. All data is processed using our proprietary algorithms on servers owned and fully controlled by Stardots.
Results are delivered through a cloud-connected web client and mobile app. This allows both physicians and Parkinson’s patients to track symptoms and monitor treatment progress.
Based on the needs of the clinic. Developed from university research.
Our mathematical models and algorithms have been developed under Professor Alexander Medvedev for over a decade. They are the result of several joint research projects between Uppsala University and Uppsala University Hospital.
Parkinson’s disease symptoms can manifest themselves in many ways and make automatic measurements challenging. ANLIVA relies on physical and biological knowledge to the largest extent possible. This has some benefits to machine- or deep-learning approaches. For one, we do not need large data sets tied to an explicit understanding of the markers we are looking to quantify.
ANLIVA’s symptom quantification uses mathematical models that correspond to physical reality to create a so-called “digital twin” of the patient. We fit patient data into mathematical models to describe each patient and how therapy can relieve symptoms. As a result, clinicians can closely follow symptom development and optimize patient treatment, even without physical patient contact.