The role of input imaging combination and ADC threshold on segmentation of acute ischemic stroke lesion using U-Net

BackgroundTo evaluate the effect of the weighting of input imaging combo and ADC threshold on the performance of the U-Net and to find an optimized input imaging combo and ADC threshold in segmenting acute ischemic stroke (AIS) lesion.MethodsThis study retrospectively enrolled a total of 212 patients having AIS. Four combos, including ADC-ADC-ADC (AAA), DWI-ADC-ADC (DAA), DWI-DWI-ADC (DDA), and DWI-DWI-DWI (DDD), were used as input images, respectively. Three ADC thresholds including 0.6, 0.8 and 1.8 x 10(-3) mm(2)/s were applied. Dice similarity coefficient (DSC) was used to evaluate the segmentation performance of U-Nets. Nonparametric Kruskal-Wallis test with Tukey-Kramer post-hoc tests were used for comparison. A p < .05 was considered statistically significant.ResultsThe DSC significantly varied among different combos of images and different ADC thresholds. Hybrid U-Nets outperformed uniform U-Nets at ADC thresholds of 0.6 x 10(-3) mm(2)/s and 0.8 x 10(-3) mm(2)/s (p < .001). The U-Net with imaging combo of DDD had segmentation performance similar to hybrid U-Nets at an ADC threshold of 1.8 x 10(-3) mm(2)/s (p = .062 to 1). The U-Net using the imaging combo of DAA at the ADC threshold of 0.6 x 10(-3) mm(2)/s achieved the highest DSC in the segmentation of AIS lesion.ConclusionsThe segmentation performance of U-Net for AIS varies among the input imaging combos and ADC thresholds. The U-Net is optimized by choosing the imaging combo of DAA at an ADC threshold of 0.6 x 10(-3) mm(2)/s in segmentating AIS lesion with highest DSC.

Automated summary

This study aimed to evaluate the impact of input imaging combo and ADC threshold on the performance of U-Net in segmenting AIS lesions and to find the optimal combo and threshold. Results showed significant variations in segmentation performance among different combos and ADC thresholds. The best combo was DAA with an ADC threshold of 0.6 x 10(-3) mm(2)/s, achieving the highest Dice similarity coefficient (DSC) in AIS lesion segmentation.