Integrating Sensor Fusion for Improved Insights and Study Outcomes
by Anna Keil
Senior Software Engineer, Clinical ink
When patients present for treatment at a clinic, physicians are trained to combine the results of objective assessments with their general observations to form a more complete picture of the patient’s health status. In much the same way, by fusing data from different real-world assessments using technology to continually collect and analyze data, we can gain a more complete — and indeed a more accurate — picture of the patient for clinical research.
Defining Continuous Data Collection
Continuous data collection involves tracking various aspects of a patient’s everyday life using wearables and mobile devices equipped with sensors. There are two types of continuous data: active data and passive data.
- Active data involves performing specific tasks within a defined period to provide known data patterns. For example, in Parkinson’s disease studies, patients may be asked to perform instrumented motor tasks to measure steps, gait, and tremors.
- Passive data collection occurs naturally in the context of patients’ normal routines, eliminating the need for active participation and providing a comprehensive view of a patient’s behavior and health in real-world settings.
Enhancing Traditional Clinical Measures with Continuous Data
Integrating continuous data with traditional clinical measures enables a more holistic view of patients and their conditions. It provides numerous benefits to studies such as greater objectivity, sensitivity, and more opportunities to detect patient or disease changes.
Additionally, it enhances patient convenience, improves compliance rates (Clinical ink has observed up to a 98.3% compliance rate), and provides real-time situational awareness in clinical studies. By fusing data from wearables, sensors, and other sources, researchers can develop digital endpoints and biomarkers, gaining a deeper understanding of therapy efficacy and its impact on a patient’s quality of life, and enabling more sensitive patient monitoring.
Examples of sensor fusion with continuous data streams include:
- Voice data combined with patient diaries for respiratory studies
- Passive data from smartwatches integrated with active movement tasks for movement disorder research
- Mobility data from GPS or Geographics Information Systems (GISes) in conjunction with other health-related information to assess a patient’s quality of life across various disease states
Sensor fusion is a key driver of innovation in clinical research. By leveraging advanced technologies alongside sensor fusion, such as disease-focused feature engineering and artificial intelligence, researchers can accurately classify disease status and predict outcomes with better accuracy.
Ensuring Success in Continuous Data Capture and Analysis
Capturing and analyzing continuous data requires expertise in designing data capture processes and selecting the appropriate platforms for ingestion, storage, processing, monitoring, export, and analysis. Without the right skills and tools, managing the volume and complexity of continuous data can become overwhelming.
Partnering with experienced teams, such as the Clinical ink Advanced Technology Team, ensures the successful capture, processing, and analysis of complex eSource data and empowers researchers to make informed decisions and meet study goals effectively.
Read more about the Clinical ink Advanced Technology Team and its goals in this DPHARM interview.
