Risk of recurrence after chemoradiotherapy identified by multimodal MRI and 18F-FDG-PET/CT in locally advanced cervical cancer.

BACKGROUND AND PURPOSE: MRI, applying dynamic contrast-enhanced (DCE) and diffusion-weighted (DW) sequences, and 18F-fluorodeoxyglucose (18F-FDG) PET/CT provide information about tumor aggressiveness that is unexploited in treatment of locally advanced cervical cancer (LACC). We investigated the potential of a multimodal combination of imaging parameters for classifying patients according to their risk of recurrence. MATERIALS AND METHODS: Eighty-two LACC patients with diagnostic MRI and FDG-PET/CT, treated with chemoradiotherapy, were collected. Thirty-eight patients with MRI only were included for validation of MRI results. Endpoints were survival (disease-free, cancer-specific, overall) and tumor control (local, locoregional, distant). Ktrans, reflecting vascular function, apparent diffusion coefficient (ADC), reflecting cellularity, and standardized uptake value (SUV), reflecting glucose uptake, were extracted from DCE-MR, DW-MR and FDG-PET images, respectively. By applying an oxygen consumption and supply-based method, ADC and Ktrans parametric maps were voxel-wise combined into hypoxia images that were used to determine hypoxic fraction (HF). RESULTS: HF showed a stronger association with outcome than the single modality parameters. This association was confirmed in the validation cohort. Low HF identified low-risk patients with 95% precision. Based on the 50th SUV-percentile (SUV50), patients with high HF were divided into an intermediate- and high-risk group with high and low SUV50, respectively. This defined a multimodality biomarker, HF/SUV50. HF/SUV50 increased the precision of detecting high-risk patients from 41% (HF alone) to 57% and showed prognostic significance in multivariable analysis for all endpoints. CONCLUSION: Multimodal combination of MR- and FDG-PET/CT-images improves classification of LACC patients compared to single modality images and clinical factors.

Autodelineation of cervical cancers using multiparametric magnetic resonance imaging and machine learning.

BACKGROUND: Tumour delineation is a challenging, time-consuming and complex part of radiotherapy planning. In this study, an automatic method for delineating locally advanced cervical cancers was developed using a machine learning approach. MATERIALS AND METHODS: A method for tumour segmentation based on image voxel classification using Fisher?s Linear Discriminant Analysis (LDA) was developed. This was applied to magnetic resonance (MR) images of 78 patients with locally advanced cervical cancer. The segmentation was based on multiparametric MRI consisting of T2- weighted (T2w), T1-weighted (T1w) and dynamic contrast-enhanced (DCE) sequences, and included intensity and spatial information from the images. The model was trained and assessed using delineations made by two radiologists. RESULTS: Segmentation based on T2w or T1w images resulted in mean sensitivity and specificity of 94% and 52%, respectively. Including DCE-MR images improved the segmentation model?s performance significantly, giving mean sensitivity and specificity of 85?93%. Comparisons with radiologists? tumour delineations gave Dice similarity coefficients of up to 0.44. CONCLUSION: Voxel classification using a machine learning approach is a flexible and fully automatic method for tumour delineation. Combining all relevant MR image series resulted in high sensitivity and specificity. Moreover, the presented method can be extended to include additional imaging modalities.