Detection and localization of cell swelling in stroke / brain injury, partly via transfocal impedence-matching, partly through microwave imaging in conjunction with Medfield Diagnostics AB. The latter technique is aiming to improve diagnosis of acute brain injury / stroke, not only in hospitals in the neuro-intensive care unit, but also during emergency ambulance transportation, including to enable early pre-hospital decision support for thrombolytic therapy in acute stroke.
Project Leader: Mikael Elam
In this project we are developing a microwave based technique for diagnosis of stroke patients. A stroke can be caused either by a blood clot or a bleeding and the treatment is entirely different for the different cases. The single most important factor for a successful outcome for the patient is an early start of the treatment and efficient diagnostic equipment is therefore vital. Today the diagnostics are made with x-ray computed tomography or magnetic resonance imaging at the hospital. With a microwave based technique the equipment can be made much more compact and even fitted into an ambulance, allowing for earlier diagnostics and consequently earlier start of treatment than today is possible.
In Sweden roughly 30 000 people suffer a stroke each year and about 15-30% of them become permanently disabled. In the United States the corresponding number is approximately 700 000 persons per year. Approximately 85% of all strokes are ischemic (blood clot induced) and 15% are hemorrhagic (bleeding) strokes. Stroke is placed third among reasons for acute death, and first among reasons for neurological dysfunction. Stroke is also resulting in most days of hospital nursing and therefore the most costly disease within the western world health care. The total social cost resulting from stroke in Sweden is estimated to EUR 1.2 billion per year. Among stroke survivors, 20% have serious remaining dysfunctions. A much larger proportion has less conspicuous dysfunctions which still seriously affect quality of life for the patient and relatives.
Early thrombolytic treatment has been proven very successful for ischemic stroke patients. If given to hemorrhagic stroke patients the treatment in itself could instead be lethal. The effect of the treatment is better the sooner it is given and it cannot be given later than 3 hours since start of the symptoms, as the tissue damage at this time result in a high risk for bleeding. Today computer tomography (CT) and sometimes magnetic resonance imaging (MRI) are used to diagnose the ischemic from the hemorrhagic strokes. The examinations in these facilities are however time consuming and together with transportation and other procedures in the chain of care the result is that only 2–3 % of the ischemic stroke patients receives the thrombolytic therapy.
Microwave technology is a strong and novel candidate to fulfill the needed gap of a diagnostic method in the care of acute stroke patients. It has high sensitivity to differentiate between blood and brain tissue due to the high difference in permittivity. It is also a technology with components that can be made compact and thus be implemented earlier in the health care chain, in emergency rooms and in ambulances. The same microwave technology has also attracted extensive interest in the research community as a novel method for breast cancer detection as well as several other applications that are under investigation.
In microwave based stroke detection a number of transmitting and receiving antennas are mounted in a helmet and worn during the examination. Microwave measurements are made using every possible combination of antennas as transmitter and receiver. In order to make the diagnosis advanced signal analysis is required and in this project we investigate two different approaches. The first is a statistical classification algorithm which makes the diagnosis based on comparison with a large bank of data from healthy volunteers and patients with a known diagnosis. The second is an imaging algorithm where the dielectric properties of the scull are reconstructed based on the measured data.
The company Medfield Diagnostics in close collaboration with Chalmers and the Sahlgrenska University Hospital have performed an early clinical study of about 20 patients. Using the classification algorithm for the signal analysis promising results have been obtained in distinguishing between healthy persons, hemorrhagic stroke patients and ischemic stroke patients. Further clinical studies are however being planned and the development of the measurement system and signal analysis algorithms are ongoing. Further development is still needed before the technique can be introduced in clinical practice.
Some future challenges in the development are:
• Ensure that the system is robust and insensitive to disturbances from outside and patient movement.
• Improve the performance of the classification algorithm
• Further develop and improve the imaging algorithm
• Verify the microwave technique in further clinical studies
