Demystifying CNS Symptom Tracking with Advanced AI

By Lei Guan

In clinical trials of potential new therapies for Central Nervous System (CNS) disorders, it can be challenging for site-based clinicians to reliably collect symptomatology data. Patient reported outcomes can often be incomplete and symptoms don’t always present in face-to-face clinical visits. This can lead to studies that take longer to produce conclusive efficacy data while also requiring a lot of hands-on support from site teams.

Common symptoms of CNS disorders include:

  • Headache
  • Tremors and/or seizures
  • Changes in facial or vocal affect
  • Slurred speech
  • Tingling or numbness in the extremities
  • Weakness 
  • Loss of sight or changes in sight
  • Memory loss
  • Changes to cognitive ability
  • Lack of coordination

These kinds of symptoms typically fluctuate in severity over time. They can also be impacted by factors outside the therapy being studied. For instance, the patient’s level of physical activity, diet and use of other medications – which can be prescribed for comorbidities or can be simple over-the-counter treatments – can all cause measurable changes to these symptoms. All of this complicates both the patients’ ability to accurately record outcome data and the site clinician’s ability to reliably tie symptom changes to the drug under trial.

At AiCure, we’ve developed a way for our proprietary artificial intelligence (AI) to recognize and identify changes in patient symptoms of CNS disorders with a high degree of accuracy and reliability. 

AI and Digital Biomarkers

Digital biomarkers are patient data, either physical or behavioral, collected and measured via technology. They are collected with digital devices, ranging from medical devices to wearables and familiar smart devices like phones and tablets. Digital biomarkers are ideal for measuring data points that are difficult for humans to collect without aid, such as CNS symptom data.

They work by employing the previously mentioned devices to record patient activity. This can mean recording things like dosing activity (recording the patient taking each dose), or it can mean asking the patient to perform a specific series of actions. With AiCure’s solution, we utilize the Patient Connect app which is downloaded on a smart device, usually the patient’s own device. Using their smartphone or tablet, patients record video and/or audio of themselves going through a series of tasks. During recording, AI algorithms are running within the smartphone app to support and guide patients going through these tasks, ensuring high quality data is captured. The AI then analyzes the recordings to identify the presence of symptoms, even those so subtle that they would be easy for a human observer to miss. Planning for these recordings to happen in conjunction with trial medication dosing allows for more frequent data collection versus relying on patient reported outcomes complemented by data collected during clinic visits. Using this technology allows study teams to detect and identify immediately even minute changes in patients’ behavior and symptomatology. Some of these changes can be difficult to see during clinical visits, and sometimes they aren’t present at all during face-to-face interactions. In this way, digital biomarkers can provide added insight into what happens with patients between visits. 

Wearable devices are similarly being deployed more and more in trials in order to capture data from patients around-the-clock. Access to data that doesn’t rely on patient reporting or in-visit interviews is playing an increasingly important role in the measurement of drug efficacy, particularly with disease states such as CNS. Innovative approaches that apply any and all relevant touchpoints for gathering data along with AI to analyze that data will continue to grow in importance.


Solutions exist to help sponsors and sites overcome data collection challenges related to CNS patients. Because these sets of symptoms are prone to natural fluctuation, technological tools to help quickly and reliably capture symptom changes are an asset to study teams. The use of technology powered by AI can remove many obstacles to CNS data capture, allowing studies to run more quickly and to produce better overall data. For more information on Digital Biomarkers and AiCure’s AI-driven approach to clinical research, visit