Style transfer generative adversarial networks to harmonize multisite MRI to a single reference image to avoid overcorrection.

Recent work within neuroimaging consortia have aimed to identify reproducible, and often subtle, brain signatures of psychiatric or neurological conditions. To allow for high-powered brain imaging analyses, it is often necessary to pool MR images that were acquired with different protocols across multiple scanners. Current retrospective harmonization techniques have shown promise in removing site-related image variation. However, most statistical approaches may over-correct for technical, scanning-related, variation as they cannot distinguish between confounded image-acquisition based variability and site-related population variability. Such statistical methods often require that datasets contain subjects or patient groups with similar clinical or demographic information to isolate the acquisition-based variability. To overcome this limitation, we consider site-related magnetic resonance (MR) imaging harmonization as a style transfer problem rather than a domain transfer problem. Using a fully unsupervised deep-learning framework based on a generative adversarial network (GAN), we show that MR images can be harmonized by inserting the style information encoded from a single reference image, without knowing their site/scanner labels a priori. We trained our model using data from five large-scale multisite datasets with varied demographics. Results demonstrated that our style-encoding model can harmonize MR images, and match intensity profiles, without relying on traveling subjects. This model also avoids the need to control for clinical, diagnostic, or demographic information. We highlight the effectiveness of our method for clinical research by comparing extracted cortical and subcortical features, brain-age estimates, and case-control effect sizes before and after the harmonization. We showed that our harmonization removed the site-related variances, while preserving the anatomical information and clinical meaningful patterns. We further demonstrated that with a diverse training set, our method successfully harmonized MR images collected from unseen scanners and protocols, suggesting a promising tool for ongoing collaborative studies. Source code is released in USC-IGC/style_transfer_harmonization (github.com).

Evaluation of Copanlisib in Combination with Eribulin in Triple-negative Breast Cancer Patient-derived Xenograft Models.

The PI3K pathway regulates essential cellular functions and promotes chemotherapy resistance. Activation of PI3K pathway signaling is commonly observed in triple-negative breast cancer (TNBC). However previous studies that combined PI3K pathway inhibitors with taxane regimens have yielded inconsistent results. We therefore set out to examine whether the combination of copanlisib, a clinical grade pan-PI3K inhibitor, and eribulin, an antimitotic chemotherapy approved for taxane-resistant metastatic breast cancer, improves the antitumor effect in TNBC. A panel of eight TNBC patient-derived xenograft (PDX) models was tested for tumor growth response to copanlisib and eribulin, alone or in combination. Treatment-induced signaling changes were examined by reverse phase protein array, immunohistochemistry (IHC) and 18F-fluorodeoxyglucose PET (18F-FDG PET). Compared with each drug alone, the combination of eribulin and copanlisib led to enhanced tumor growth inhibition, which was observed in both eribulin-sensitive and -resistant TNBC PDX models, regardless of PI3K pathway alterations or PTEN status. Copanlisib reduced PI3K signaling and enhanced eribulin-induced mitotic arrest. The combination enhanced induction of apoptosis compared with each drug alone. Interestingly, eribulin upregulated PI3K pathway signaling in PDX tumors, as demonstrated by increased tracer uptake by 18F-FDG PET scan and AKT phosphorylation by IHC. These changes were inhibited by the addition of copanlisib. These data support further clinical development for the combination of copanlisib and eribulin and led to a phase I/II trial of copanlisib and eribulin in patients with metastatic TNBC. SIGNIFICANCE: In this research, we demonstrated that the pan-PI3K inhibitor copanlisib enhanced the cytotoxicity of eribulin in a panel of TNBC PDX models. The improved tumor growth inhibition was irrespective of PI3K pathway alteration and was corroborated by the enhanced mitotic arrest and apoptotic induction observed in PDX tumors after combination therapy compared with each drug alone. These data provide the preclinical rationale for the clinical testing in TNBC.

Style Transfer Using Generative Adversarial Networks for Multi-Site MRI Harmonization.

Large data initiatives and high-powered brain imaging analyses require the pooling of MR images acquired across multiple scanners, often using different protocols. Prospective cross-site harmonization often involves the use of a phantom or traveling subjects. However, as more datasets are becoming publicly available, there is a growing need for retrospective harmonization, pooling data from sites not originally coordinated together. Several retrospective harmonization techniques have shown promise in removing cross-site image variation. However, most unsupervised methods cannot distinguish between image-acquisition based variability and cross-site population variability, so they require that datasets contain subjects or patient groups with similar clinical or demographic information. To overcome this limitation, we consider cross-site MRI image harmonization as a style transfer problem rather than a domain transfer problem. Using a fully unsupervised deep-learning framework based on a generative adversarial network (GAN), we show that MR images can be harmonized by inserting the style information encoded from a reference image directly, without knowing their site/scanner labels a priori. We trained our model using data from five large-scale multi-site datasets with varied demographics. Results demonstrated that our style-encoding model can harmonize MR images, and match intensity profiles, successfully, without relying on traveling subjects. This model also avoids the need to control for clinical, diagnostic, or demographic information. Moreover, we further demonstrated that if we included diverse enough images into the training set, our method successfully harmonized MR images collected from unseen scanners and protocols, suggesting a promising novel tool for ongoing collaborative studies.

REGION SPECIFIC AUTOMATIC QUALITY ASSURANCE FOR MRI-DERIVED CORTICAL SEGMENTATIONS.

Quality control (QC) is a vital step for all scientific data analyses and is critically important in the biomedical sciences. Image segmentation is a common task in medical image analysis, and automated tools to segment many regions from human brain MRIs are now well established. However, these methods do not always give anatomically correct labels. Traditional methods for QC tend to reject statistical outliers, which may not necessarily be inaccurate. Here, we make use of a large database of over 12,000 brain images that contain 68 parcellations of the human cortex, each of which was assessed for anatomical accuracy by a human rater. We trained three machine learning models to determine if a region was anatomically accurate (as 'pass', or 'fail') and tested the performance on an independent dataset. We found good performance for the majority of labeled regions. This work will facilitate more anatomically accurate large-scale multi-site research.