BACKGROUND: High variability of intracranial arterial blood flow velocities by Transcranial color-coded sonography (TCCS) has been found in clinical practice. This study aimed to improve diagnostic accuracy by analyzing influencing factors of middle cerebral artery (MCA) blood flow velocity detected by TCCS. METHODS: In total, 328 MCA vessels were classified as normal (27.1 %) or having mild (30.2 %), moderate (23.2 %), and severe (19.5 %) stenosis based on computed tomography angiography (CTA). Based on morphology, MCAs were classified as type I (98) or type II (230). Differences in peak systolic velocity (PSV) detected by TCCS and TCD were analyzed at different degrees of MCA stenosis (MCAS), correction angles, and morphology. RESULTS: The mean rank of MCAS of TCCS was higher than that of TCD and CTA (P <
0.05), with no significant difference between TCD and CTA. The PSV measured by TCCS was significantly higher than that of TCD, but when the correction angle of TCCS was ≤30°, the two PSVs were comparable (P >
0.05). TCCS and TCD measured significantly different PSVs when the TCCS correction angles were >
30° in type I and at all correction angles in type II (P <
0.05). The optimal cut-off values for MCAS diagnosis using TCCS increased with increasing correction angles. CONCLUSIONS: The differences in PSVs measured using TCCS and TCD were related to the correction angles and morphology of the MCA. Using optimal cut-off values based on the correction angles allows for more accurate MCAS diagnosis.