An analysis of the angle dependence in strain (rate) imaging of the left ventricle

In this paper, we analyze the angle dependence of longitudinal strain estimates in the left ventricle (LV). We also assess the effect of moving the sample volume (SV) according to the left ventricular motion. The analysis was performed by calculating the true strain and estimated ("Doppler-based") strain in a thick-walled deforming ellipsoid that incorporates chamber deformation and torsion. LV diameter, wall thickness and base-to-apex length were measured in 15 healthy individuals, averaged to obtain a representative cardiac cycle and used as input to the model. Firstly, we simulated myocardial deformations as imaged from the apical view with the transducer at various lateral positions (0 mm corresponded to a position 10 mm straight above the apex). For the apical, mid and basal segments the true systolic strains were: -16.4, -16.7 and -17.0 %, respectively. The optimal transducer positions, giving negligible measurement errors, were (moving/fixed SV): 7/8 mm, 15/17 mm and 21/25 mm. When the transducer was positioned at 0 mm (i.e. straight above the apex), the estimated strains were (moving/fixed SV): -15.2/-14.7, -14.5/-14.0 and -14.3/-13.7%. The ranges of lateral transducer locations with estimated strains (moving SV) within 80% of the true strain were -5-17 mm, -4-32 mm and 3-45 mm. Secondly, we calculated estimated strain in the basal segment when varying the beam-to-wall angle from 0 to 30 degrees. The estimated peak strain was within 80% of true peak strain for beam-to-wall angles less than 16 degrees. In conclusion, the conventional clinical imaging window straight above the apex gives measurement errors for the mid and basal segments in the range 13-16% (moving SV), suggesting that the transducer position should be optimized with these segments in mind. A moving SV reduces the distance between the optimal transducer positions of the three segments. Since the actual transducer position is the same for all three segments, this in effect reduces the overall measurement error. A beam-to-wall misalignment of more than 15 degrees should be avoided.