A stroke patient with a blood clot in the brain needs treatment within four hours in order not to decease or suffer severe disability. Therefore, the Chalmers research team are now developing an instrument for faster stroke diagnostics. The system is based on microwave technology and aims to enable distinction between a bleeding stroke and a stroke caused by a blood clot.
If thrombolysis treatment is quickly started on a stroke patient, the patient’s life can be saved and improved. The problem is that the cause of the stroke needs to be an absolute certainty in order to start such treatment. 15 % of all strokes are actually caused by bleedings, in which case administration of blood thinners would be life threatening.
The diagnostics method used today is computerized tomography (CT). The trombolytic treatment has to be given to the patient within four and a half hours, otherwise the brain tissue will already be dead and the treatment will have no effect. However, if the diagnosis can be confirmed already in the ambulance, perhaps even making it possible to give thrombolysis treatment on the way to hospital, this could dramatically change the lives of millions of stroke patients.
Stroke Finder consists of three parts – a helmet-like antenna system which is placed on the patient’s head, a microwave unit plus a computer for equipment control, data collection and signal processing. Research is carried out for all three parts within the research group.
Clinical studies with the aim to verify this technique are taking place at Sahlgrenska University Hospital in Gothenburg.
Every year, about 15 million people worldwide suffer a stroke, according to the World Health Organization. In developed countries, stroke is the third leading cause of death, exceeded only by coronary heart disease and cancer. Furthermore, stroke is the leading cause of serious long-term disability.
In Sweden, around 30,000 people suffer a stroke each year, of which a third become permanently disabled.
The critical difference between blood clots and bleeding
Approximately 85% of all strokes are ischemic (blood clot induced) and 15% are haemorrhagic (bleeding) strokes. Stroke is the third cause behind acute death and the first cause for neurological dysfunction in Swedish health care.
Thrombolysis (blood clot resolving medication) has been shown to improve the outcome from ischaemic stroke, if given within three hours from the occurrence of the stroke. This requires the ability to quickly distinguish between ischaemic and haemorrhagic stroke. For successful thrombolytic treatment, new cerebral ischemic events need to be identified within 15 to 30 minutes. It is partly due to these limited time windows that less than 5% of patients who could benefit from this treatment actually receive it. The total cost of treatments and rehabilitation for stroke patients is estimated to SEK 12 billion per year in Sweden, a cost which could be greatly reduced if more patients received the right treatment from the start.
Diagnostics methods today
Standard techniques for stroke diagnostics and monitoring are EEG, CT, MRI, and fMRI. The common factors of these modalities are the relatively involved procedures and the associated time involved, not only in the procedure itself, but also in getting access to the relevant diagnostics tool.
Ultrasound is the technique which comes closest to the same level of technology and accessibility as that of microwaves. However, in contrast to microwaves, ultrasound has problems passing through scull bones and so would not be an option for this kind of diagnostics. Similar problems are observed for near Infrared imaging and spectroscopy, impedance measurements and impedance tomography.
The microwave techniques previously mentioned are novel and innovative methods which are not commercially available at present.
In this stroke detection experiment, the multi-static scattering matrix is measured at a large number of frequencies. This measurement strategy is similar to what is commonly used in microwave tomography experiments. The antenna helmet is built using patch antennas of the shape of an isosceles triangle. The antenna is excited by a probe-feed via a coaxial transmission line positioned at the centerline of the patch. This experimental system uses 10 patch antennas that have been mounted in a helmet.
The multi-static matrix is measured using each antenna as a transmitter as well as a receiver. A network analyzer is used as the transmitting and receiving unit. A particular problem with the helmet is that it is rigid, while the shape and size of different patients vary. For the antenna to fit individual patients the solution has been to mount flexible plastic bags fillable with water inside the helmet. The purpose of the water bags is to fill up the space between the antennas and the skull but also to provide impedance matching between the antenna and the skull.
For signal analysis purposes, the project team has developed a statistical classifier algorithm. Measurements have been made both on patients with stroke and on healthy volunteers. The comparison of measurement data between the patient and the healthy volunteer group has then been used to create an algorithm, which can be used to diagnose patients.
Plans for the future involve performing further measurements on more patients in order to obtain additional clinical data. Tests on a large number of patients are required in order to verify the accuracy of this technology and also to assess limitations.
Another future challenge is to miniaturize the Stroke Finder system in order to fit it into an ambulance. If it takes more than four hours for a stroke patient to get to hospital and get a diagnosis, their brain tissue will already be dead and the treatment will have no effect. However, if the diagnosis can be confirmed already in the ambulance, perhaps even making it possible to give thrombolysis treatment on the way to hospital, this could dramatically change the lives of millions of stroke patients.
Measurements have been made using the system on a large number of healthy volunteers and stroke patients. So far, clinical measurement results are encouraging and show great potential.
However, many more patients must be included in the clinical verification before the system can be used as the only diagnostic modality in the ambulance, making it accurate enough to base life and death decisions on it.