Atrial fibrillation (AF) is a condition where the heartbeat becomes fast and irregular. AF can be a complication of acute illness. Even short episodes of AF, double the risk of developing blood clots or a stroke.
We wish to understand which patients have a high risk of developing it. Using modern compute technology (“artificial intelligence”), computer scientists can develop tools that update in real time when receiving patient data to inform healthcare professionals about the risk for AF. This would allow doctors to take measures to prevent AF or to request further investigations. In this project, we wish to collect data to develop such a tool, also called “digital twin”. Testing of this digital twin in patients will be part of a follow-up study. We wish to collect this data using an approved and licensed wireless monitoring system.
Patients over 50 years old admitted to hospital for acute illness or high-risk operations can take part. Patients will have heart rate, breathing rate, blood pressure and oxygen levels measured as per the normal hospital practice. Patients will also wear a wireless patch to measure heart rate and rhythm for up to 7 days. The patch sends data to a tablet and records the electrical signals of the heart continuously.
We will enrol 25 patients in the second part of the study, in which alerts for any abnormal readings of blood pressure, oxygen saturation in the blood, or heart rate will be sent to the ward nurse on a mobile phone. In questionnaires or interviews, we will ask these patients and the nursing staff for their opinions on the extra monitoring and how useful the alerts are in improving patient care.
Atrial fibrillation (AF) is a common heart rhythm disorder and a major cause of stroke, especially in older adults. AF increases the risk of stroke by causing blood clots that can travel to the brain. However, AF can be difficult to detect because it may occur without symptoms or in irregular, brief episodes. In many cases of unexplained strokes, the cause could be undetected AF that occurs intermittently. Standard tests like ECGs and Holter monitors often miss these brief episodes, especially if the monitoring period is limited.
Early detection of AF is important because blood thinners can prevent future strokes, but these medications are only prescribed if AF is found. One of the goals of the TAILOR study is to generate a cohort of patients with an ischemic stroke who will be monitored using wearable cardiac devices to improve the detection of AF by using a longer monitoring period and advanced AI technology.
In addition to heart monitoring, the TAILOR study leverages advanced brain imaging techniques, like MRI, to better understand how stroke affects the brain and recovery. By comparing MRI scans taken at the time of stroke and 90 days later, we can track changes in the brain, such as damage to the brain tissue and how the brain adapts. This will help us to better understand what factors influence stroke outcomes, including disability levels at three months and changes in cognitive and behavioural abilities.
The TAILOR study will be conducted at two centres: Hospital del Mar in Barcelona, Spain, and Radboud University Medical Center in Nijmegen, The Netherlands. As part of the larger TARGET project, the TAILOR study aims to enhance AF detection and stroke outcome prediction using "virtual twin" technology. These models will help doctors personalise treatment plans, improving outcomes and reducing healthcare costs.
Stroke care includes a rehabilitation program which is recognised as a cornerstone of the recovery period. The intensity of therapy is related to functional recovery although there is high variability in the amount of time and techniques applied in therapy sessions. There is a need to better describe stroke rehabilitation protocols to develop a better understanding of current practices and to increase the internal validity and generalisability of clinical trial results. Typically published studies do not differentiate between stroke aetiology, at most they may define ischaemic or haemorrhagic events. Identifying differences in aetiology and stroke recovery can help the scientific community to develop predictive models of recovery optimising available resources.
The PEARL study, as part of the TARGET project, is a prospective observational cohort study of patients with acute stroke admitted to Stroke Unit for 2 years. The main goal is the data recruitment of the time-course of AF-stroke patients compared with non-AF in terms of functional recovery, following high or medium rehabilitation intensity programs and, testing a new screening tool developed for prognostic management and function. The functional outcomes will be compared before and after applying the developed selection AI-virtual twin tool, helping to personalise rehabilitation decisions.
Mapping the time course of recovery of sitting and standing balance and walking after stroke
This study aims to investigate how sitting and standing balance, as well as gait initiation, recover during the first six months after a stroke. By mapping these recovery patterns, the research seeks to provide insights that can improve rehabilitation strategies for stroke survivors.
Participants will be recruited shortly after admission to Inkendaal Rehabilitation Hospital and will be followed up longitudinally, with assessments at baseline and at fixed time points throughout the first six months. The study includes in-lab measurements using EMG sensors, accelerometry, and force plates to capture detailed data during various tasks. Additionally, wearable Activ8 sensors are used to monitor daily activities such as sitting, standing, and walking over several days, providing insights into real-world recovery. Together, these tools provide a comprehensive understanding of recovery both inside and outside the clinical gait lab setting.
During in-lab assessments, participants perform tasks using an interactive serious game that involves reaching for virtual targets displayed on a screen. While performing the game, participants' (biomechanical) data are recorded to assess parameters such as postural sway, weight-bearing, reaction time, balance adjustments, muscle activity etc. The game is designed to simulate real-life scenarios by including both predictable tasks, where the target location is known, and unpredictable tasks, requiring rapid adjustments in movement and posture. A dual-task paradigm further integrates cognitive demands, such as decision-making under time pressure.
By combining these data, the study aims to identify key dynamics of the recovery process and support the development of personalised rehabilitation strategies. These findings will contribute to optimising care for stroke survivors and improving their quality of life.