Pretreatment prediction of the response to chemotherapy for spine metastases of breast cancer using histogram signatures from diffusion weighted imaging with multiple b-values: a retrospective study
Pretreatment diffusion weighted imaging (DWI) is effective to predict the response to chemotherapy for spine metastases of breast cancer, but current methods of analyzing DWI data with multiple b-values require a predefined mathematical model for curve fitting. This study proposes and validates a method of combining histogram signatures and deep learning to predict treatment response for spine metastases of breast cancer without the need of fitting MRI signals at multiple b-values into a predefined diffusion model. 37 patients with breast cancer with 191 spine metastatic lesions were included in this study. All patients underwent MRI scanning with DWI before chemotherapy and were divided into an investigating group ( n = 19) and a test group ( n = 18). Two radiologists created region of interest (ROI) by manually delineating lesion borders on pretreatment DWI images. DWI signals inside ROI at 10 b-values were converted into histograms and concatenated into a histogram signature. A convolutional neural network (CNN) was constructed to classify histogram signatures into a progressive disease (PD) group or a non-PD group. The predictive performance of eight fitted diffusion parameters, ADC, DDC, α, D*, D true , frac, D app and K app , were compared with histogram signature by receiver operating characteristic (ROC) curves. The area under ROC curve (AUC) of PD prediction for spine metastases of breast cancer was 0.882 (95%CI: 0.803–0.938) by the model of histogram signatures. Among the eight fitted diffusion parameters, D app had the best performance for PD prediction with an AUC of 0.685 (95%CI: 0.585–0.774). The AUC of histogram signature was significantly (Z = 3.312, P = 0.0009) higher than that of D app value. A deep‑learning model based on pretreatment histogram signatures derived from multi‑b‑value DWI data shows potential for predicting response to chemotherapy in breast‑cancer‑derived spine metastases.