Microwave imaging (MWI) method could provide a great opportunity for early stroke diagnosis in the future and thus reduce the health consequences caused by stroke. Based on different dielectric properties of healthy and stroke tissue MWI systems can help to differentiate the type of stroke in prehospital care. The main purpose of this contribution was to validate the newly designed multilevel 24-port MWI system on synthetic numerical data. Inside the 3D human head phantom, the different stroke phantom types (HEM - haemorrhagic or ISCH - ischemic) with different diameters were placed. Using the reconstruction algorithm based on Born Approximation and TSVD the stroke phantoms could be followed and distinguished. The numerical analysis of MWI system showed promising results where positions, diameters and types of stroke phantoms were successfully reconstructed. The system proved some limitations as a disability to detect objects with a size lower than half of the used wavelength which can be eliminated by increasing the operating frequency range in the future.