Leonardino A. Digma¹, Christine H. Feng¹, Christopher C. Conlin², Ana E. Rodriguez-Soto², Kanha Batra³, Aaron Simon¹, Roshan Karunamuni¹, Joshua Kuperman², Rebecca Rakow-Penner², Michael E. Hahn², Anders M. Dale², and Tyler M. Seibert^1,4
1Department of Radiation Medicine and Applied Sciences, UCSD School of Medicine, La Jolla, CA, United States, ²Department of Radiology, UCSD School of Medicine, La Jolla, CA, United States, ³Department of Electrical and Computer Engineering, UC San Diego, La Jolla, CA, United States, ⁴Department of Bioengineering, UC San Diego, La Jolla, CA, United States

Bone is among the most common sites of cancer metastases. Diffusion weighted imaging (DWI) can be used to detect these metastases. However, when acquired with echo-planar imaging, DWI suffers distortions due to static magnetic field inhomogeneities. In this study, we first used the reverse polarity gradient (RPG) technique to measure spatial distortions of bone metastases on DWI. Next, we demonstrated that RPG can be used to correct these distortions and produce diffusion images that more accurately reflect the underlying anatomy. Taken together, findings support the use of distortion correction techniques to improve localization of bone metastases on DWI.

Antonio Candito¹, Matthew D Blackledge¹, Fabio Zugni², Richard Holbrey¹, Sebastian Schäfer³, Matthew R Orton¹, Ana Ribeiro⁴, Matthias Baumhauer³, Nina Tunariu¹, and Dow-Mu Koh^1
1The Institute of Cancer Research, London, United Kingdom, ²IEO, European Institute of Oncology IRCCS, Milan, Italy, ³Mint Medical, Heidelberg, Germany, ⁴The Royal Marsden NHS Foundation Trust, London, United Kingdom

We develop a model of treatment response in patients with advanced prostate cancer (APC), based on measurements of median Apparent Diffusion Coefficient (ADC) and tumour Diffusion Volume (tDV) derived using Whole-Body Diffusion Weighted Imaging (WBDWI) and blood biomarkers. The model showed that change in ADC is positively correlated with response in this patient population.  Furthermore, our model predicted response with an accuracy of 87%, and sensitivity/specificity of 75/93% using a probability cut-off of 0.5.  

Antonio Candito¹, Matthew D Blackledge¹, Fabio Zugni², Richard Holbrey¹, Sebastian Schäfer³, Matthew R Orton¹, Ana Ribeiro⁴, Matthias Baumhauer³, Nina Tunariu¹, and Dow-Mu Koh^1
1The Institute of Cancer Research, London, United Kingdom, ²IEO, European Institute of Oncology IRCCS, Milan, Italy, ³Mint Medical, Heidelberg, Germany, ⁴The Royal Marsden NHS Foundation Trust, London, United Kingdom

We employed a deep transfer-learning model to classify whether images from whole-body diffusion-weighted MRI (WBDWI) contain metastatic bone lesions. Our results demonstrate sensitivity/specificity of 0.87/0.89 on 8 test patients, who were not included in the model training. Such a model may accelerate radiological assessment of disease extent from WBDWI, which currently can be cumbersome to interpret due to the large quantity of data (approximately 200-250 images per patient).

Christopher C Conlin¹, Christine H Feng², Leonardino A Digma², Ana E Rodriguez-Soto¹, Joshua M Kuperman¹, Dominic Holland³, Rebecca Rakow-Penner¹, Tyler M Seibert^1,2,4, Michael E Hahn¹, and Anders M Dale^1,3,5
1Department of Radiology, UC San Diego School of Medicine, La Jolla, CA, United States, ²Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine, La Jolla, CA, United States, ³Department of Neurosciences, UC San Diego School of Medicine, La Jolla, CA, United States, ⁴Department of Bioengineering, UC San Diego Jacobs School of Engineering, La Jolla, CA, United States, ⁵Halıcıoğlu Data Science Institute, UC San Diego, La Jolla, CA, United States

Whole-body DWI is increasingly used to assess bone involvement in prostate cancer. Multicompartmental diffusion modeling can outperform conventional DWI techniques for evaluating tumors, but has yet to be applied to whole-body imaging. In this study, we determined an optimal multicompartmental model for describing whole-body diffusion and applied it to examine metastatic bone lesions in vivo. We found that a 4-compartment model best characterized whole-body diffusion. Compartmental signal-contributions revealed by this model show improved bone-lesion conspicuity and may help to assess microstructural changes that accompany prostate-cancer bone involvement.

Suhong Qin¹, Ailian Liu¹, SHUANG MENG¹, Lihua Chen¹, Qinhe Zhang¹, Qingwei Song¹, and Yunsong Liu^1
1The First Affiliated Hospital of Dalian Medical University, Dalian, China

It remains a challenge to diagnose the oligometastatic prostate cancer (PCa) due to the ambiguous definition of oligometastatic PCa. Previous studies had shown that diffusion kurtosis imaging (DKI) is a non-gaussian diffusion weighted imaging (DWI) method that yields more accurate results on the microstructural complexity of prostate cancer tissue structure. This study indicated that performances of DKI cannot differentiate between oligometastatic and widely metastatic PCa, however it has the potential to assess tumor load and aggressiveness in metastatic PCa.

Cyrano Chatziantoniou^1,2, Reineke Schoot², Roelof van Ewijk², Simone ter Horst², Rick van Rijn³, Hans Merks², Alexander Leemans¹, and Alberto de Luca^1
1Image Science Institute, UMC Utrecht, Utrecht, Netherlands, ²Princess Máxima Center for Pediatric Oncology, Utrecht, Netherlands, ³Department of Radiology, Academic Medical Centre Amsterdam, Amsterdam, Netherlands

Rhabdomyosarcoma is a rare form of cancer that is particularly prevalent in children. There is a pressing need for new imaging biomarkers for monitoring treatment response, such diffusion. No standards currently exist to annotate rhabdomyosarcoma on MRI and measure the diffusion.

This work analyses segmentation strategies in recent literature on saromas and compares them on a toy example. It shows that there is a large inconsistency in the application of segmentation strategies and that the different methods for segmenting the tumor can yield high variation in measured diffusion. 

Ramesh Paudyal¹, Linda Chen², Jung Hun Oh¹, Kaveh Zakeri², Vaios Hatzoglou³, Chiaojung Jillian Tsai², Nancy Lee², and Amita Shukla-Dave^1,3
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ³Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States

The study aims to assess quantitative imaging (QI) metrics from pre-treatment (TX) non-gaussian intravoxel incoherent DW- and fast exchange regime (FXR) DCE-MRI for predicting locoregional failure (LRF) in nasopharyngeal carcinoma (NPC) patients. Cumulative incidence analysis (CIA) was performed on the two subgroups dichotomized with Youden’s index. Competing-risks regression based on Fine and Gray’s (FG) proportional sub hazards model was used to estimate survival subdistribution hazard ratios (SHRs). The pre-TX ADC, D, f, and ti cutoff values from CIA analysis and K cutoff value from the competing risk regression analysis indicated these QI’s could predict the LRF in NPC patients.

Lukas Lundholm¹, Mikael Montelius¹, Oscar Jalnefjord¹, Eva Forssell-Aronsson¹, and Maria Ljungberg^1
1Department of Radiation Physics, Institute of Clinical Sciences, Gothenburg, Sweden

VERDICT dMRI allows for estimation of microstructural parameters in tumours which could facilitate planning and assessment of treatment. Due to the complexity of the model there is a risk for overfitting to data and there is hence a need to determine the reliability of the estimated parameters. A mouse model of human SI-NETs (n=5) was measured twice using the same dMRI protocol and the VERDICT model was fitted to data. Results showed an overall good repeatability of the tumour mean parameter values estimated by VERDICT. However, some local clusters of voxels showed larger differences between repeated scans.

Archana Vadiraj Malagi¹, Devasenathipathy Kandasamy², Kedar Khare³, Deepam Pushpam⁴, Rakesh Kumar⁵, Sameer Bakhshi⁴, and Amit Mehndiratta^1,6
1Centre for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi, India, ²Department of Radiodiagnosis, All India Institute of Medical Sciences Delhi, New Delhi, India, ³Department of Physics, Indian Institute of Technology Delhi, New Delhi, India, ⁴Department of Medical Oncology, Dr. B.R. Ambedkar Institute-Rotary Cancer Hospital (IRCH), All India Institute of Medical Sciences Delhi, New Delhi, India, ⁵Department of Nuclear Medicine, All India Institute of Medical Sciences Delhi, New Delhi, India, ⁶Department of Biomedical Engineering, All India Institute of Medical Sciences Delhi, New Delhi, India

PET/CT plays an important role in diagnosis and assessment of treatment response in lymphoma. The goal of this study was to evaluate the role of IVIM-DKI parameters in comparison to PET parameters in lymphoma. PET images were registered onto IVIM-DKI at b=0s/mm² images for tumor ROI using 3D-multimodal affine registration. Qualitatively, IVIM parameters with state-of-the-art Total-Variation produced better quality parameter maps. Tumor appeared hyperintense in SUV and K maps and hypointense in diffusion and perfusion parameters. No correlation was observed between IVIM-DKI with PET parameters. IVIM with Total-Variation showed substantial reproducibility as compared to conventional IVIM, DKI and SUV parameters.

Qi Xie¹, Wenjuan He¹, Zhilin Tan¹, Yajie Wang¹, Jinbin Wu¹, Xiyan Shao², Yiming Yang³, Jing Zhang⁴, Kangwei Wang⁵, Guiqin Wang⁶, Qifeng Pan¹, and Yunzhu Wu^7
1Medical Imaging Department, Nansha Hospital, Guangzhou First People's Hospital, School of Medicine, South China University of Technology, Guangzhou, China, ²Ultrasound Imaging Department, Longgang District People’s Hospital, Shenzhen, China, ³Department of Radiology, The Second Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangdong Provincial Hospital of Traditional Chinese Medicine, Guangzhou, China, ⁴Department of Pathology, Cancer Center, Sun Yat-sen University, Guangzhou, China, ⁵Department of Pathology, Nansha Hospital, Guangzhou First People's Hospital, School of Medicine, South China University of Technology, Guangzhou, China, ⁶Medical Record Department, Nansha Hospital, Guangzhou First People's Hospital, School of Medicine, South China University of Technology, Guangzhou, China, ⁷MR Scientific Marketing, SIEMENS Healthcare Ltd., Guangzhou, China

This study has demonstrated that ADC value of DWI with single-exponential model and diffusion coefficient D of IVIM are valuable for discriminating 5-FU-responsive and 5-FU-resistant colon cancer in vivo. IVIM is expected to be a simple and practical way to quantitatively monitor tumor resistance in vivo, and ADC &D value may be an imaging biomarker.

Xiaojiao Pei¹, Tao Jiang¹, Zhenyu Pan¹, Yufei Lian¹, Yueluan jiang², and qinglei Shi ^3
1Radiology, Beijing Chaoyang Hospital Affiliated to Capital Medical University, Beijing, China, ²MR Scientific Marketing, Diagnosis Imaging, Siemens Healthineers China, Beijing, China, ³Scientific Marketing, Diagnosis Imaging, Siemens Healthineers China, Beijing, China

Intravoxel incoherent motion diffusion-weighted MR imaging (IVIM-DWI) can simultaneously acquire diffusion parameters and perfusion parameters without intravenous administration, compared with conventional diffusion imaging. This study investigated the application of IVIM in the newly diagnostic multiple myeloma (NDMM) patients.

Amaresha Shridhar Konar¹, Akash Deelip Shah², Ramesh Paudyal¹, Jung Hun Oh¹, Eve LoCastro¹, David Aramburu Nuñez¹, Nathaniel Swinburne², Robert J. Young², Andrei I. Holodny², Kathryn Beal³, Vaios Hatzoglou², and Amita Shukla-Dave^1,2
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, ³Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, United States

In the clinical settings it is essential to accurately assess, whether or not a brain metastases has been successfully treated or whether it requires additional treatment, especially in high dose radiation therapy, such as  stereotactic radiosurgery (SRS). The present prospective study aims to determine the ability of Diffusion Weighted (DW)- and Dynamic Contrast Enhanced (DCE)-MRI to predict the long-term response of brain metastases within 72 hours of SRS. The preliminary results are promising as it will inform the treating physicians at an early time point about which patients will benefit from SRS (or not).

Andre Monteiro Paschoal^1,2, Raquel Andrade Moreno^3,4, Antonio Carlos dos Santos⁵, and Renata Ferranti Leoni^6
1LIM44, Instituto e Departamento de Radiologia, Faculdade de Medicina, Universidade de Sao Paulo, Sao Paulo, Brazil, ²InBrain Lab, University of Sao Paulo, Ribeirao Preto, Brazil, ³Instituto do Cancer do Estado de Sao Paulo, Sao Paulo, Brazil, ⁴Memorial Sloan-Kettering Cancer Center, New York, NY, United States, ⁵Departamento de Clinica Medica, Faculdade de Medicina de Ribeirao Preto, Universidade de Sao Paulo, Ribeirao Preto, Brazil, ⁶InBrain Lab - University of Sao Paulo, Ribeirao Preto, Brazil

The assessment perfusion in glioma is an important information in the tumor characterization. Traditional MRI methods employed to analyze glioma are based on the application of contrast agents to enhance the T1 signal and/or to measure perfusion within the tumor cells. The analysis of tumor microvasculature however is hampered by the size of contrast agent molecules. In that sense, SWI can be employed, although it does not provide quantitative information. IVIM maps showed to be an interesting alternative for such assessment, even showing potential of an early biomarker to assess early changes in perfusion, specially in low grade gliomas.

ruicheng ba¹, Hongxi Zhang², Zhongwei Gu³, Yuhao Liao¹, Xingwang Yong¹, Zhiyong Zhao¹, Yi Zhang¹, and Dan Wu^1
1Key Laboratory for Biomedical Engineering of Ministry of Education, Department of Biomedical Engineering, College of Biomedical Engineering & Instrument Science, Zhejiang University, Hang zhou, Zhejiang, China, ²Children's Hospital, Zhejiang University School of Medicne，Department of Radiology, Hang zhou, Zhejiang, China, ³Children's Hospital, Zhejiang University School of Medicne, Department of Pathology, Hangzhou, Zhejiang, China

Diffusion-time dependent diffusion MRI (dMRI) has been proposed to characterize tumor microstructure. This study aimed to evaluate diagnostic value of time-dependent dMRI based microstructural mapping to grade and categorize pediatric brain tumor at 3T. Oscillating and pulsed gradient dMRI was performed in 35 pediatric patients. Cell diameter (d), intracellular fraction (f_in) and extracellular diffusivity (D_ex) were fitted based on the IMPULSED model. Higher cellularity (f_in/d) and f_in and lower D_ex were found in high-grade glioma than low-grade ones. The cellularity further increased in medulobastoma compared with glioma.  Cellularity achieved the highest area-under-the-curve than the other dMRI metrics in grading glioma.

Richa Singh Chauhan¹, Nihar Kathrani², Jitender Saini¹, Maya D Bhat¹, Karthik Kulanthaivelu¹, Vani Santosh³, Nishanth S⁴, and Subhas Konar^4
1Neuroimaging and Interventional Radiology, NIMHANS, BENGALURU, India, ²Interventional Radiology, Paras Hospital, Gurgaon, India, ³Neuropathology, NIMHANS, BENGALURU, India, ⁴Neurosurgery, NIMHANS, BENGALURU, India

H3K27M mutant diffuse midline gliomas are a newly classified entity in the 2016 World health organization classification of CNS tumors. They are high grade tumors, with the mere presence of this mutation confers them a WHO grade IV designation, irrespective of their histologic morphology. Patients harboring these tumors have dismal prognosis and shorter overall survival. Furthermore, since these are deep-seated lesions involving the eloquent brain areas, biopsy can be challenging with substantial risk of morbidity. Our work proposes non-invasive multiparametric MRI-based imaging attributes to detect the H3K27M mutation pre-operatively. Results demonstrate a considerable accuracy on 123 patients.

Gao Eryuan¹, Gao Ankang¹, Zhang Huiting², Wang Shaoyu², Yan Xu², Bai Jie¹, and Cheng Jingliang^1
1Dept. of MRI, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, Zhengzhou, China, ²MR Scientific Marketing, Siemens Healthcare, Shanghai, China, Shanghai, China

This study aimed to investigate the efficiency of four advanced diffusion models, including diffusion tensor imaging (DTI), diffusion kurtosis imaging (DKI), neurite orientation dispersion and density imaging (NODDI) and mean apparent propagator (MAP) in grading of glioma. Through histogram analysis of parameters, we found that axial diffusivity (AD)_maximum, mean diffusivity (MD)_maximum and radial diffusivity (RD)_maximum from DTI, and Q-space inverse variance (QIV)_maximum and QIV_range from MAP had significant differences and good diagnostic efficiency in all comparisons among different grading of glioma.

Gao Ankang¹, Gao Eryuan¹, Zhang Huiting², Wang Shaoyu², Yan Xu², Bai Jie¹, and Cheng Jingliang^1
1Dept. of MRI, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, ²MR Scientific Marketing, Siemens Healthcare, Shanghai, China, Shanghai, China

The Isocitrate dehydrogenase (IDH) genotyping and epigenetic 1p/19q codeletion as two key molecular markers are included in the glioma WHO 2016 classification. Gaussian or non-Gaussian diffusion models were recently proposed to provide additional microstructure information. In present work, we applied four diffusion models in glioma grading and genotyping, including DTI, DKI, MAP-MRI and NODDI models, which could be acquired within a single scan.

Rifeng Jiang¹, Kaiji Deng¹, Yixin Guo², and Zhongshuai Zhang^3
1Fujian Medical University Union Hospital, Fuzhou, China, ²Fujian Medical University, Fuzhou, China, ³MR Scientific Marketing, Siemens Healthcare, Shanghai, China

This study found that NODDI seems to be a more potent approach in evaluating the early CST infiltration by HGG, and can evaluate the CST destruction with a similar performance to MD by providing additional information about neurite density for HGG-induced CST injury.

Jan Brabec¹, Filip Szczepankiewicz², Jaromír Šrámek³, Elisabet Englund⁴, Johan Bengzon⁵, Linda Knutsson^1,6, Carl-Fredrik Westin^7,8, Pia C Sundgren^2,9, and Markus Nilsson^2
1Medical Radiation Physics, Lund University, Lund, Sweden, ²Diagnostic Radiology, Lund University, Lund, Sweden, ³Institute of Histology and Embryology, First Faculty of Medicine, Charles University, Prague, Czech Republic, ⁴Division of Oncology and Pathology, Department of Clinical Sciences, Lund University, Lund, Sweden, ⁵Division of Neurosurgery, Department of Clinical Sciences, Lund University, Lund, Sweden, ⁶Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ⁷Harvard Medical School, Boston, MA, United States, ⁸Radiology, Brigham and Women’s Hospital, Boston, MA, United States, ⁹Lund University Bioimaging Center, Lund University, Lund, Sweden

Mean diffusivity (MD) in tumors is often too readily interpreted as cellularity. Here, we investigated which microstructural features explain the MD in meningioma tumors. We performed high-resolution MRI and histological imaging on excised mengioma tumor samples and coregistered MD to histology. We found that cellularity alone was a poor predictor of MD, and that the prediction improved markedly by considering also the microcyst density. A qualitative analysis also indicated that collagen fibers, stroma and psammoma bodies affect the MD.

Osamu Togao¹, Akio Hiwatashi², Toru Chikui³, Kazufumi Kikuchi², Yukiko Kami³, Kenji Tokumori⁴, and Kousei Ishigami^2
1Molecular Imaging & Diagnosis, Kyushu University, Fukuoka, Japan, ²Clinical Radiology, Kyushu University, Fukuoka, Japan, ³Oral and Maxillofacial Radiology, Kyushu University, Fukuoka, Japan, ⁴Clinical Radiology, Teikyo University, Omuta, Japan

The aim of the present study was to determine whether the gamma distribution (GD) model is useful in the differentiation of glioblastomas (GBs) and primary CNS lymphomas (PCNSLs). The GD model well described the histological features of PCNSLs and GBs, and its use enabled the significant differentiation of these tumors. The κ, f2, and f3 values were significantly smaller and the f1 values were significantly larger in the PCNSLs than in the GBs. The GD model-derived parameters correlated well with the IVIM-derived parameters. The GD model may therefore contribute to the characterization of brain tumor from the histological viewpoint.