View Presentation Video

Chenqing Wu¹, Haoting Wu¹, Cheng Zhou¹, Jingwen Chen¹, Jiaqi Wen¹, Xiaocao Liu¹, Jianmei Qin¹, Zhengye Cao¹, Tao Guo¹, Xiaojun Guan¹, Xiaojun Xu¹, Baorong Zhang², and Minming Zhang^1
1Department of Radiology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China, ²Department of Neurology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China

Keywords: Parkinson's Disease, Neurodegeneration

View Presentation Video

Keywords: Breast, Treatment, dynamic contrast-enhanced magnetic resonance imaging, pathological complete response, radiomics

Based on pre-treatment and early treatment dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and clinical characteristics, we established a pathological complete response (pCR) prediction model using a deep learning radiomic (DLR) method that achieved good performance in the training and validation cohorts. The model can help clinicians evaluate whether the patient can reach pCR after neoadjuvant chemotherapy (NAC) and can provide an effective diagnostic reference for accurate medical treatment of patients receiving NAC.

View Presentation Video

Yannik Streibel¹, Jessica Hunger^1,2,3, Chenchen Pan^4,5, Verena Turco², Manuel Fischer¹, Volker Sturm¹, Kianush Karimian-Jazi^1,4, Dennis A. Agardy^2,3,5, Giacomo Annio^6,7, Rami Mustapha⁸, Christopher B. Rodell⁹, Wolfgang Wick^4,5, Ralph Sinkus^6,7, Sabine Heiland¹, Frank Winkler^4,5, Michael Platten^2,10, Martin Bendszus¹, Michael O. Breckwoldt^1,2, and Katharina Schregel^1,4
1Department of Neuroradiology, Heidelberg University Hospital, Heidelberg, Germany, ²Clinical Cooperation Unit Neuroimmunology and Brain Tumor Immunology, German Cancer Consortium (DTK) within the German Cancer Research Center (DKFZ), Heidelberg, Germany, ³Faculty of Biosciences, Heidelberg University, Heidelberg, Germany, ⁴Clinical Cooperation Unit Neurooncology, German Cancer Consortium (DTK) within the German Cancer Research Center (DKFZ), Heidelberg, Germany, ⁵Department of Neurology, Heidelberg University Hospital, Heidelberg, Germany, ⁶INSERM UMRS1148 - Laboratory for Vascular Translational Science, University Paris, Paris, France, ⁷School of Biomedical Engineering and Imaging Sciences, King’s College London, London, United Kingdom, ⁸Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom, ⁹School of Biomedical Engineering, Science and Health Systems, Drexel University, Philadelphia, PA, United States, ¹⁰Department of Neurology, University Medical Center Mannheim, Heidelberg University, Mannheim, Germany

Keywords: Elastography, Neuro

MRI and MRE were used to monitor the effects of immunotherapy on tumor volume, FA and biomechanics of murine orthotopic glioma. Treated tumors were significantly smaller, softer and had lower FA than controls. This difference was most pronounced when comparing tumor stiffness of both groups. Controls revealed heterogeneous tumor stiffness. We hypothesize that this is caused by viable tumor cells alternated with necrotic areas and presumably immune-suppressive iba1-positive cells. In contrast, biomechanical properties of treated animals correlated better with presence of macrophages/microglia and likely reflected anti-tumorigenic inflammation. Thus, MRE could prove useful for monitoring glioma response to therapy.

View Presentation Video

Shengcai Wei¹, Jing Cheng¹, and Yi Wang^1
1Department of Radiology, Peking University People's Hospital, Beijing, China

Keywords: Radiomics, Machine Learning/Artificial Intelligence, Histopathological growth patterns; colorectal liver metastases; transformation

Our study used an MRI-based radiomics model to predict the transformation of the histopathological growth pattern (HGP) of colorectal liver metastases (CRLMs) before and after chemotherapy. After collecting data, drawing regions of interest and analyzing radiomics, we enrolled 152 patients and 299 liver metastases (99 pure desmoplastic (pdHGP) and 174 non-pdHGP). The pdHGP in the non-chemotherapy group and the post-chemotherapy group accounted for 28.3% and 42.5%, respectively (P=0.019). The fused MRI-based radiomics model demonstrated good predictive performance, and it could predict pdHGP before chemotherapy (25.3%), which was significantly different (p=0.034) compared with postoperative pdHGP after chemotherapy (39.1%).

View Presentation Video

Upasana Roy¹, Malin Pedersen¹, Carol Box¹, Jessica K.R. Boult¹, Antonio Rullan¹, Michael Schmohl², Sebastian Carotta³, Kevin J. Harrington¹, and Simon P. Robinson^1
1Radiotherapy and Imaging, The Institute of Cancer Research, Sutton, United Kingdom, ²Pharma GmbH & Co. KG, Boehringer Ingelheim, Biberach, Germany, ³RCV GmBH & Co KG, Boehringer Ingelheim, Vienna, Austria

Keywords: Cancer, Drug Development, Preclinical, Diffusion Imaging, Relaxometry

Whilst cancer immunotherapies have shown marked and durable tumour responses in some patients, the majority derive no benefit.  Strategies are being exploited to enhance tumour responses to immuno-oncology agents, including pharmacological activation of the STING pathway to create a more inflamed microenvironment.  Multi-parametric MRI revealed a dose-dependent increase in murine tumour ADC in response to a STING agonist that was associated with histologically confirmed increase in tumour cell death.  An acute, transient increase in tumour R₂*, consistent with vascular occlusion, was also found.  ADC is a sensitive early imaging biomarker of tumour response to STING agonism.   

View Presentation Video

Anum S. Kazerouni¹, Lanell M. Peterson¹, Isaac Jenkins², Alena Novakova¹, Hannah M. Linden¹, David Hockenbery¹, David Mankoff³, Jennifer Specht¹, and Savannah C. Partridge^1
1University of Washington, Seattle, WA, United States, ²Fred Hutch Cancer Center, Seattle, WA, United States, ³University of Pennsylvania, Philadelphia, PA, United States

Keywords: Breast, Cancer

We investigated the combination of DCE-MRI and ¹⁸F-FDG PET for assessing metabolism/perfusion mismatch and predicting pathological response and recurrence-free survival (RFS) in women undergoing neoadjuvant chemotherapy (NAC) for breast cancer. Thirty-five patients with localized breast cancer were imaged with both modalities at 3 timepoints over the course of NAC. Greater mid-treatment decreases in imaging measures were predictive of pathological response and associated with improved long-term outcome. Furthermore, mid-treatment decreases in metabolism/perfusion ratios were predictive of RFS. These results indicate a complementary relationship between DCE-MRI and ¹⁸F-FDG PET metrics and potential value of metabolism/perfusion mismatch as a marker of response.

View Presentation Video

Wei Huang¹, Andy Kaempf¹, Brooke Beckett¹, Alina Tudorica¹, Jeong Youn Lim¹, and Christopher Ryan^1
1Oregon Health & Science University, Portland, OR, United States

Keywords: Cancer, Treatment, DCE-MRI, DW-MRI, Therapy Response, Ktrans, ADC

Thirty patients with lower extremity soft tissue sarcoma were treated with preoperative chemoradiotherapy and underwent MRI exams including DCE- and DW-MRI pre-therapy, after the first cycle of therapy regimen, and after completion of therapy but before surgery.  Ten patients had optimal and the other 20 had sub-optimal pathologic responses based on necrosis percentage of the surgical tumor specimens.  Quantitative DCE-MRI markers and apparent diffusion coefficient (ADC) were superior to radiographic tumor size measurement in early prediction and evaluation of therapy response.  Multi-parametric approach of combining DCE- and DW-MRI markers further improved early prediction of optimal vs. sub-optimal response.

View Presentation Video

Emilia Palmér¹, Jesper Brovall², Oscar Jalnefjord^1,2, Karin Petruson³, Fredrik Nordström ^1,2, Anna Karlsson^1,2, Maria Ljungberg^1,2, and Maja Sohlin^1,2
1Department of Medical Radiation Sciences, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, ²Department of Medical Physics and Biomedical Engineering, Sahlgrenska University Hospital, Gothenburg, Sweden, ³Department of Oncology and Radiotherapy, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

Keywords: Multi-Contrast, Radiotherapy, Diffusion/Other Diffusion Imaging Techniques

Tumor oxygenation is a biomarker proposed as a predictor of radiation therapy (RT) response.  Here, the feasibility of Oxygen-Enhanced MRI, intravoxel incoherent motion, and diffusion kurtosis imaging for monitoring of oxygenation changes in head and neck cancers was evaluated. Seven patients were examined pre- and mid-RT. No relative change in population mean longitudinal relaxation rate was observed following RT. A general increase was noticed in population mean diffusion and capillary perfusion fraction, and a decrease in population mean kurtosis. The implementation of these techniques was clinically feasible, and relative changes in almost every derived biomarker could be observed following RT.

View Presentation Video

Keywords: Machine Learning/Artificial Intelligence, Modelling

Prior characterization of treatment-effect and tumor recurrence using deep learning approaches have not optimized for spatial classification within a single lesion, which could improve surgical planning and treatment. 10mm patches of pre-surgical anatomical and physiological images surrounding the locations of histopathologically-confirmed tissue samples were used to train our models. Including physiological images, pretraining on unlabeled data in an autoencoding task, and training with an alternative cross-validation approach that enabled many networks to be ensembled, we achieved an ensembled test AUROC of 0.814 and generated spatial maps of tumor probability and model uncertainty. Performance decreased when removing any of these components.

View Presentation Video

Anjali Datta¹, Gustavo Chau Loo Kung², Kristin Quah³, Daniel Barbosa⁴, Chelsea Li⁵, David Purger⁵, Allan Wang⁵, Yosefi Chodakiewitz¹, Pejman Ghanouni¹, Vivek Buch⁵, and Jennifer McNab^1
1Radiology, Stanford University, Stanford, CA, United States, ²Bioengineering, Stanford University, Stanford, CA, United States, ³Electrical Engineering, Stanford University, Stanford, CA, United States, ⁴Neurosurgery, University of Pennsylvania, Philadelphia, PA, United States, ⁵Neurosurgery, Stanford University, Stanford, CA, United States

Keywords: Data Analysis, Tractography & Fibre Modelling, Neurosurgical Targeting

MRI-guided focused-ultrasound (MRgFUS) is an FDA-approved treatment for essential tremor. Unfortunately, the target to be ablated cannot be directly visualized on standard imaging, and its location varies interindividually. Since suboptimal lesion location can lead to side effects that impair quality of life, a method that locates personalized ablative targets that correlate with better quality-of-life outcomes (and not just with tremor reduction) could provide significant benefit. We present a clinically-feasible patient-specific probabilistic-tractography-based method for personalized targeting of MRgFUS treatment and show that similarity between the target it generates and the ablated lesion predicts QoL outcome in a dataset of 36 patients.

View Presentation Video

xiaoyu jiang¹, Jingping Xie², John Gore², and Junzhong Xu^3
1Radiology, Vanderbilt University Medical Center, nashville, TN, United States, ²Vanderbilt University Medical Center, nashville, TN, United States, ³Vanderbilt University Medical Center, Brentwood, TN, United States

Keywords: Quantitative Imaging, Cancer, microstructure; diffusion; apoptosis

Reliable and sensitive methods for assessing the response of breast cancer to treatment are critical for timely adjustments of therapies for individual patients, and development of novel therapies.  Different from the conventional tumor-volume-based criteria, we hypothesize that more specific microstructural information on the cellular level in tumors, such as changes in cell size, may provide more accurate characterization of therapeutic response. Using both in vitro cell experiments and in vivo xenograft experiments, we demonstrated that temporal changes in MR-derived cell sizes in triple-negative MDA-MB-231 tumors treated with either drug vehicle or paclitaxel provide a new means to assess treatment response.        

View Presentation Video

Chu-Yu Lee¹, Michael S Petronek², Varun Monga³, Benjamin J Miller⁴, Mohammed M Milhem³, Vincent A Magnotta¹, and Bryan G Allen^2
1Department of Radiology, The University of Iowa, Iowa City, IA, United States, ²Department of Radiation Oncology, Free Radical and Radiation Biology Program, The University of Iowa, Iowa City, IA, United States, ³Department of Internal Medicine, The University of Iowa, Iowa City, IA, United States, ⁴Department of Orthopedics and Rehabilitation, The University of Iowa, Iowa City, IA, United States

Keywords: Muscle, Tumor, Soft-tissue sarcomas; Treatment assessment; Relaxometry

Soft-tissue sarcomas is commonly treated by neoadjuvant radiation therapy followed by surgical resection. Current assessment of neoadjuvant therapy relies on pathological examinations after surgery. Nonetheless, noninvasive imaging assessment can be performed before and during the treatment, offering the opportunity for predicting and early assessing treatment response. This pilot study applied T₂* mapping to evaluate neoadjuvant therapy in seven patients with soft-tissue sarcomas before, during the treatment, and before surgery. The results showed strong or moderate correlations between T₂* measurements and percent necrosis from pathological examinations, suggesting the potential for using T₂* mapping to predict and early assess treatment response.     

View Presentation Video

David A Hormuth II^1,2, Maguy Farhat³, Julianna Bronk³, Holly Langshaw³, Thomas E Yankeelov^1,2,4,5,6,7, and Caroline Chung^3
1Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin, Austin, TX, United States, ²Livestrong Cancer Institutes, The University of Texas at Austin, Austin, TX, United States, ³Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁴Biomedical Engineering, The University of Texas at Austin, Austin, TX, United States, ⁵Oncology, The University of Texas at Austin, Austin, TX, United States, ⁶Diagnostic Medicine, The University of Texas at Austin, Austin, TX, United States, ⁷Imaging Physics, The University of Texas MD Anderson Cancer Center, Houston, TX, United States

Keywords: Tumors, Radiotherapy

Technical advances in imaging and radiotherapy in the last decade have motivated the development of patient-specific treatment plans accounting for the underlying tumor biology that ultimately informs treatment efficacy. A fundamental challenge is how to leverage imaging data to optimize patient response to radiotherapy. Towards this end, we have developed an approach to forecast tumor response prior to the conclusion of therapy to enable a spatially-resolved map of response to chemoradiation. Our forecasting approach can accurately identify statistically significant changes in cell density that could potentially inform treatment adaptations.

View Presentation Video

Li Jun Qian¹, Xu Hua Gong¹, Ming Xuan Feng², Hai Nan Ren¹, Yan Zhou¹, Jian Rong Xu¹, Qiang Xia², and Yang Song^3
1Radiology, Renji Hospital, Shanghai Jiao Tong University, School of Medicine, Shanghai, China, ²Liver Surgery, Renji Hospital, Shanghai Jiao Tong University, School of Medicine, Shanghai, China, ³MR Scientific Marketing, Siemens Healthineers Ltd., Shanghai, China

Keywords: Cancer, Liver, Hepatoblastoma

This study evaluates the diagnostic efficacy of contrast-enhanced MR imaging for preoperative POSTTEXT staging and prediction of vascular involvement in pediatric hepatoblastoma after neoadjuvant chemotherapy. The findings revealed that MR preoperative POSTTEXT staging has a good level of agreement with the reference standard. MR provides high predictive performance for identifying portal vein involvement and moderate predictive performance for identifying hepatic vein/inferior vena cava involvement.

View Presentation Video

Sheng Qing Lin¹, Sebastian Fonseca¹, Durga Udayakumar^1,2, Alberto Diaz de Leon¹, Orhan Oz¹, Gurbakhash Kaur³, Aimaz Afrough³, Larry D. Anderson Jr.³, and Ananth J. Madhuranthakam^1,2
1Radiology, UT Southwestern Medical Center, Dallas, TX, United States, ²Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, ³Internal Medicine, UT Southwestern Medical Center, Dallas, TX, United States

Keywords: Cancer, Quantitative Imaging, Disease Biomarkers

WBMRI-DETECT has been previously shown to have improved lesion conspicuity and shorter scan times compared to WBMRI-STIR and WBMRI-DWIBS in multiple myeloma. In this work, we demonstrate that quantitative biomarkers measured through WBMRI, fat fraction (FF) and apparent diffusion coefficient (ADC), have weak correlation suggesting both biomarkers to provide complementary information. Additionally, both ADC and FF showed negative correlation with the semiquantitative SUV_max from FDG-PET, demonstrating potential use of quantitative MRI biomarkers for the assessment of therapy response.