From my basic understanding of stroke there are 2 classes of...

From my basic understanding of stroke there are 2 classes of stroke; ischemic and hemorhagic. both types starve the brain of life supporting blood flow. There are also mini strokes known as Transient Ischemic Attack (TIA)

Ischemic strokes can be classed within two subtypes; Embolic and Thrombotic.
Embolic stroke is caused by clot formation somewhere else in the body (in elderly population, it usually forms in the heart from atrial fibrillation) and has released from its origin, the clot then travels in the body eventually flowing until its size obstructs an artery which feeds the brain or part of the brain. Thrombotic stroke on the other hand the clot forms within the brain eventually obstructing flow of blood causing Ischemia to brain tissue.

Hemorrhagic strokes can also be classed into two subtypes Intracerebral and Subarachnoid.
Intracerebral means a blood vessel that has ruptured within the brain
Subarachnoid basically means a blood vessel on the outside of the brain has ruptured

From my understanding CT and MRI scans can see these changes they can also see the secondary effects when tissue death has occurred, obviously MRI with more specific resolution. Both have pros and cons and limitations to use but both are bulky, very expensive and not mobile. Step in EMV's unit, as I understand EMV's device it is not aiming to replace these technologies, at this point in the game, but it is to fill a void in diagnosis treatment and convenience in use. If these units can effectively identify stroke types in the field eg paramedics they can send them to appropriate facilities. If these unit can identify stroke types providing a diagnosis therefor treatment and be afforded by smaller healthcare facilities in regional areas then saves transporting these critical patients to facilities which can scan. Early treatment show better outcomes therefor less money spent on rehab and improved quality of life. Additionally, some pts in critical care units are too unstable to transport to MRI and CT therefore filling another gap providing a safe and convenient scan to identify concerns with a patients neurological status which are at risk of bleeding re-bleeding following stroke.

Ultimately one can assume they will be assessing the accuracy of the scans compared to current scans from CT and if a correlation of relevance following algorithmic tweaks can be identified and accepted by medical staff then there would be something to work with for further development.