Prediction of placenta accreta spectrum using texture analysis on coronal and sagittal T2-weighted imaging

Letter to the editor: Radiomics features of patients with placenta accreta spectrum: A quantification of heterogeneity caused by intraplacental T2‐hypointense bands

“…Was the ITHB correlated with the selected texture features other than shape in their study? In our previous study, ITHB significantly correlated with texture features obtained from T2WI 5 . The correlation between ITHB and texture features was not surprising, since the presence of ITHB results in a heterogeneous signal intensity in the placenta on T2WI.…”

“…In our previous study, ITHB significantly correlated with texture features obtained from T2WI. 5 The correlation between ITHB and texture features was not surprising, since the presence of ITHB results in a heterogeneous signal intensity in the placenta on T2WI. In general, a model composed of variables that are correlated with each other leads to a reduction of the generalization performance.…”

Letter to “Development and validation of a magnetic resonance imaging‐based nomogram for predicting invasive forms of placental accreta spectrum disorders”

“…We performed quantitative texture analysis on the axial, coronal, and sagittal planes of T2WI, and showed better diagnostic performance in quantitative texture analysis with coronal and sagittal planes than in visual MRI signs for differentiating PAS from the normal group. 4 Although our texture analysis cannot quantify the bulging or loss of double line signs reported by Liet al, texture analysis showed superiority at evaluating intraplacental heterogeneity compared to visual MRI signs. The design of the method to quantify the bulging or loss of double line signs would be a future task.…”

“…Our previous study scored four visual MRI signs (outward bulging of the placenta, intraplacental abnormal vascularity, intraplacental T2 dark band, and placental heterogeneity) on the three planes. 4 Indeed, four visual MRI signs on axial, coronal, and sagittal planes showed different diagnostic ability in differentiating PAS from the normal group (Table 1).…”

Application of Texture and Volume Model Analysis to Dedicated Axillary High-resolution 3D T2-weighted MR Imaging: A Novel Method for Diagnosing Lymph Node Metastasis in Patients with Clinically Node-negative Breast Cancer

“…For the texture features of the LN with the largest least axis length, we performed feature selection by the least absolute shrinkage and selection operator (LASSO) algorithm. 21,[24][25][26] We constructed the texture model using the selected texture features of the LN with the largest least axis length by Random Forest (Fig. 2).…”