Roshni Senthilkumar^1,2, Sai Man Cheung², Kwok-Shing Chan³, and Jiabao He^2
1Radiology Physics, University Hospital Coventry and Warwickshire, Coventry, United Kingdom, ²Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Fosterhill, Aberdeen, Scotland, ³Donders Institute for Brain, Cognition and Behaviour, AJ Nijmegen, Germany

MR relaxation properties of T1 and T2 are well known to alter in cancer, reflecting diseased tissue micro-environment. Multiple compartment models have been developed to better approximate tissue classes within an imaging voxel, allowing more accurate estimation of disease load for more precise treatment planning and monitoring. However, multiple compartment models is more sensitive to noise and suffers from potential overfitting, due to from significantly increased number of variables for fitting exponential functions and consequently more complex cost function. We therefore conducted numerical simulation to establish, the first in the literature, applicability condition of multiple compartment model in breast cancer.

Michael Berks¹, Damien J McHugh², Nuria Porta³, Ross A Little¹, Susan Cheung¹, Gordon C Jayson^4,5, Geoff J M Parker^6,7, and James P B O'Connor^1,8,9
1Quantitative Biomedical Imaging Laboratory, Division of Cancer Sciences, University of Manchester, Manchester, United Kingdom, ²Medical Physics, The Christie Hospital NHS Trust, Manchester, United Kingdom, ³Clinical Trials and Statistics Unit, The Institute of Cancer Research, London, United Kingdom, ⁴Division of Cancer Sciences, University of Manchester, Manchester, United Kingdom, ⁵Department of Medical Oncology, The Christie Hospital NHS Trust, Manchester, United Kingdom, ⁶Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, ⁷Bioxydyn Ltd, Manchester, United Kingdom, ⁸Department of Radiology, The Christie Hospital NHS Trust, Manchester, United Kingdom, ⁹Division of Radiotherapy and Imaging, The Institute of Cancer Research, London, United Kingdom

DCE-MRI biomarkers such as change in median K^trans have a proven role in drug development in phase I/II trials. There is current interest in using approaches such as radiomics to extract additional information relating to spatial heterogeneity from images and one emerging application is to apply these analyses to clinical trial data where imaging is used to monitor pharmacodynamic change in the tumour microenvironment. Here, we explore the properties of radiomics features extracted from maps of K^trans and aim to identify features that are repeatable at baseline, show consistent treatment effect and provide additional, independent information to the median K^trans.

Labib Shahid¹, Juan Pablo Gonzalez-Pereira¹, Cody Johnson¹, Yanheng Li², David Rowinski², and Alejandro Roldán-Alzate^1
1University of Wisconsin-Madison, Madison, WI, United States, ²Convergent Science, Inc., Madison, WI, United States

Multi-channel urodynamics is an invasive diagnostic method used to assess bladder biomechanics. MRI-based CFD has shown high potential as a clinical tool primarily focused on cardiovascular flows. In this pilot study, we demonstrate MRI-based CFD as a tool to study urodynamics. Real-time MRI on one subject provided bladder wall surfaces at multiple time points during voiding. We developed a surface mapping algorithm that processes the bladder geometries before inputting them into a CFD simulation. Coupling MRI with CFD successfully visualized and quantified urine flow dynamics. This provides a non-invasive tool to investigate urodynamics in common urological conditions such as BPH/LUTS.

Brandon Whitcher¹, Marjola Thanaj¹, Madeleine Cule², Nicolas Basty¹, Elena P. Sorokin², Jimmy D. Bell¹, and E. Louise Thomas^1
1Research Centre for Optimal Health, University of Westminster, London, United Kingdom, ²Calico Life Sciences LLC, South San Francisco, CA, United States

We investigated changes in body composition in 3,088 free-living participants of the UK Biobank imaging study. Statistical models utilised organ and tissue segmentations from the neck-to-knee Dixon acquisitions and multiecho sequences of the liver/pancreas.  A significant decrease in grip strength was observed, and small, but statistically significant, decreases in all skeletal muscle measurements. Significant increases in visceral adipose tissue and intermuscular fat in the thighs were also found in the absence of changes in BMI, waist circumference and ectopic-fat deposition. We have shown that even after a relatively short period of time significant changes in body composition are observable.

Steffen Albert¹, Barbara D. Wichtmann², Wenzhao Zhao³, Jürgen Hesser³, Ulrike I. Attenberger², Lothar R. Schad¹, and Frank G. Zöllner^1
1Computer Assisted Clinical Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, ²Department of Diagnostic and Interventional Radiology, University Hospital Bonn, Bonn, Germany, ³Data Analysis and Modeling in Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany

We evaluated the influence of normalization (setting mean and standard deviation, histogram matching and percentiles) on the segmentation of rectal cancer on multimodal images when operating on multicenter data as part of a Radiomics pipeline. We used two different networks for segmentation. When training and evaluating on all data or data from a single center, normalization did not play a significant role. In contrast, when training on one center and evaluating on all others, it did play a major role. Best results are obtained by normalization using percentiles. Fixing the mean and standard deviation did not work well.

Yoshihiko Fukukura¹, Fumitaka Ejima¹, Takuro Ayukawa¹, Kiyohisa Kamimura¹, Masanori Nakajo¹, Hiroaki Nagano¹, Koji Takumi¹, Marcel Dominik Nickel², Hiroshi Imai³, and Takashi Yoshiura^1
1Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan, ²MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany, ³Siemens Healthcare K.K., Tokyo, Japan

This study focused on the feasibility of dynamic contrast-enhanced MRI (DCE-MRI) with motion-resolved compressed sensing T1-weighted volumetric interpolated breath-hold examination (CS-VIBE) for prediction of treatment response to chemotherapy and patient outcome in pancreatic ductal adenocarcinoma (PDAC). Our results showed tumor ve was significantly higher in the response group than in the non-response group and was an independent predictor of progression-free survival in patients with PDAC treated with chemotherapy. These results suggest that DCE-MRI obtained with CS-VIBE may be useful for predicting treatment response to chemotherapy and patient outcome in PDAC.