Article
Oncology
Yunfei Zhang, Ruofan Sheng, Chun Yang, Yongming Dai, Mengsu Zeng
Summary: This study assessed the effectiveness of tri-exponential Intra-Voxel Incoherent Motion (tri-IVIM) MRI in identifying microvascular invasion (MVI) in hepatocellular carcinoma (HCC) prior to surgery. The results showed that tri-IVIM outperformed bi-IVIM in identifying MVI-positive HCC. A predictive nomogram integrating tri-IVIM metrics and clinical risk factors achieved the highest diagnostic accuracy.
JOURNAL OF HEPATOCELLULAR CARCINOMA
(2023)
Article
Oncology
Qi Liu, Jinggang Zhang, Man Jiang, Yue Zhang, Tongbing Chen, Jilei Zhang, Bei Li, Jie Chen, Wei Xing
Summary: In assessing the histopathological features of PDAC, the D and f values derived from the IVIM model showed higher sensitivity and diagnostic performance compared to the conventional DWI model. A significant negative correlation was found between D values and fibrosis, while a significant positive correlation was observed between f values and fibrosis. IVIM-DWI may serve as an imaging biomarker for predicting the fibrosis grade of PDAC.
FRONTIERS IN ONCOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
Jini Raju, Ushadevi C. Amma, Ansamma John
Summary: IVIM imaging is a magnetic resonance imaging technique that allows quantitative evaluation of diffusion and perfusion without the use of contrast agents. However, determining the optimal number and range of b values is critical to minimize errors and improve image quality. Identifying the most suitable b values is essential for accurate quantitative estimation of IVIM parameters.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yohan Son, Jinsick Park, Jeong Min Lee, Robert Grimm, In Young Kim
Summary: This study found a significant negative correlation between hepatic steatosis and DWI parameters, with ADC, D, and D-app being lower in the steatosis group and K-app being higher compared to the non-steatosis group. Perfusion-related parameters did not show statistical significance. It suggests that hepatic steatosis can affect DWI parameters and should be considered a possible confounding factor in DWI-based assessment of liver fibrosis.
ACADEMIC RADIOLOGY
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jian Zhao, Meifeng Wang, Xiaohui Ding, Yonggui Fu, Cheng Peng, Huanhuan Kang, Huiping Guo, Xu Bai, Qingbo Huang, Shaopeng Zhou, Xiaojing Zhang, Kan Liu, Lin Li, Huiyi Ye, Xu Zhang, Xin Ma, Haiyi Wang
Summary: This study evaluated the consistency of venous tumor thrombus (VTT) in renal cell carcinoma (RCC) using IVIM-DWI derived parameters and found that these parameters have the potential to predict the VTT consistency of RCC.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Gregory Simchick, Diego Hernando
Summary: The purpose of this study was to obtain precise tri-exponential intravoxel incoherent motion (IVIM) quantification in the liver using 2D (b-value and first-order motion moment [M-1]) IVIM-DWI acquisitions and region of interest (ROI)-based fitting techniques. The results showed high repeatability and reproducibility in the estimations of the diffusion coefficient, perfusion fractions, and blood velocity SDs in the right liver lobe using the 2D (b-M-1) acquisition in conjunction with BVD fitting. The study concluded that the 2D (b-M-1) IVIM-DWI data acquisition in conjunction with BVD fitting enables highly precise tri-exponential IVIM quantification in the right liver lobe.
MAGNETIC RESONANCE IN MEDICINE
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Keita Fujimoto, Yoshifumi Noda, Nobuyuki Kawai, Kimihiro Kajita, Yuta Akamine, Hiroshi Kawada, Fuminori Hyodo, Masayuki Matsuo
Summary: This study aimed to compare the diagnostic values of different diffusion-weighted imaging models in differentiating hepatic hemangiomas and liver metastases. The results showed that ADC and DDC values can be considered as quantitative imaging biomarkers for distinguishing these two types of liver lesions.
EUROPEAN JOURNAL OF RADIOLOGY
(2021)
Article
Oncology
Junjiao Hu, Xin Yu, Peidi Yin, Bin Du, Xiangran Cai
Summary: This study evaluated the predictive value of IVIM diffusion-weighted imaging (DWI) in assessing the immune responses induced by immunogenic chemotherapy. The results showed that IVIM-DWI can effectively evaluate the treatment response and the D value may serve as a sensitive imaging marker for identifying the antitumor immune response initiated by immunogenic chemotherapy.
FRONTIERS IN ONCOLOGY
(2022)
Review
Oncology
Fei Wang, Chun Yue Yan, Cai Hong Wang, Yan Yang, Dong Zhang
Summary: In this study, the predictive power of diffusion kurtosis imaging (DKI) and intravoxel incoherent motion (IVIM)-diffusion-weighted imaging (DWI) parameters for pathological grades and microvascular invasion (MVI) in hepatocellular carcinoma (HCC) was investigated and compared. The results showed that certain parameters had high sensitivity and specificity in the preoperative prediction of HCC pathological grades and MVI. Additionally, MD and D values had superior diagnostic efficacy in predicting HCC grades, while D value outperformed other parameters in predicting MVI+ in HCC.
FRONTIERS IN ONCOLOGY
(2022)
Article
Medicine, General & Internal
Chang Guo, Kai Zheng, Qiang Ye, Zixiao Lu, Zhuoyao Xie, Xin Li, Yinghua Zhao
Summary: There were independent correlations between the D parameter derived from IVIM DWI and the K-trans, K-ep, and V-e parameters derived from DCE-MRI. The factors that affected their correlations mainly included BME, gender, and the scope of lesions.
FRONTIERS IN MEDICINE
(2022)
Article
Multidisciplinary Sciences
Jelena Djokic Kovac, Marko Dakovic, Aleksandra Jankovic, Milica Mitrovic, Vladimir Dugalic, Daniel Galun, Aleksandra uric-Stefanovic, Dragan Masulovic
Summary: The study demonstrated the potential value of IVIM-related parameters in differentiating intrahepatic mass-forming cholangiocarcinoma and hypovascular liver metastases, with a higher diagnostic performance of the D parameter compared to ADC.
Article
Oncology
Yuhui Qin, Chen Chen, Haotian Chen, Fabao Gao
Summary: Parameters derived from IVIM-DWI, such as ADC, D, D*, and f, can serve as important factors in predicting survival in patients with nasopharyngeal carcinoma, helping to select high-risk patients and anticipate long-term outcomes.
FRONTIERS IN ONCOLOGY
(2022)
Article
Oncology
Xue Wang, Jiao Song, Shengfa Zhou, Yi Lu, Wenxiao Lin, Tong San Koh, Zujun Hou, Zhihan Yan
Summary: The study found that there was no significant difference between IVIM parameters obtained from the segmented method with a b-value cutoff of 200 s/mm(2) and the simultaneous fitting method. Tissue diffusivity (D) and perfusion fraction (f) were significantly lower in cervix cancer compared to normal tissue. Therefore, both the segmented method and simultaneous fitting method can be used to differentiate between cervix cancer and normal tissue.
Article
Radiology, Nuclear Medicine & Medical Imaging
Ruba Alkadi, Osama Abdullah, Naoufel Werghi
Summary: This study evaluates the potential of using different b-values of DWI in classifying prostate cancer. By combining machine learning techniques with parametric maps generated using various models, a potentially powerful system for prostate cancer diagnosis was proposed.
JOURNAL OF DIGITAL IMAGING
(2022)
Article
Oncology
Hong-Wei Li, Gao-Wu Yan, Jin Yang, Li-Hua Zhuo, Anup Bhetuwal, Yong-Jun Long, Xu Feng, Hong-Chao Yao, Xing-Xiong Zou, Ruo-Han Feng, Han-Feng Yang, Yong Du
Summary: The aim of this study was to compare the diagnostic performance of diffusion kurtosis imaging (DKI), intravoxel incoherent motion (IVIM) and diffusion-weighted imaging (DWI) in detecting and grading hepatocellular carcinoma (HCC). The results showed that mean diffusivity, mean diffusional kurtosis, true diffusion coefficient, and apparent diffusion coefficient were correlated with HCC grade, and DKI and IVIM parameters performed best in predicting highly differentiated HCC.
Article
Biochemistry & Molecular Biology
Maxime Taquet, Stephen M. Smith, Anna K. Prohl, Jurriaan M. Peters, Simon K. Warfield, Benoit Scherrer, Paul J. Harrison
Summary: Psychiatry is shifting from accepting distinct diagnoses to a representation of psychiatric illness that crosses diagnostic boundaries. A study shows that the brain vulnerability network may form a common neurobiological root for different psychiatric disorders, funneling genetic risks through shared neurobiological mechanisms.
MOLECULAR PSYCHIATRY
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Tess E. Wallace, Jonathan R. Polimeni, Jason P. Stockmann, W. Scott Hoge, Tobias Kober, Simon K. Warfield, Onur Afacan
Summary: This study developed a method for slice-wise dynamic distortion correction for EPI using FID navigators and demonstrated its efficacy relative to an established data-driven technique. Results showed that the FID-navigated distortion correction accurately corrected image geometry in the presence of induced magnetic field perturbations and yielded significant temporal SNR gains in functional MRI scans.
MAGNETIC RESONANCE IN MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Jaume Coll-Font, Onur Afacan, Jeanne S. Chow, Richard S. Lee, Simon K. Warfield, Sila Kurugol
Summary: Early identification of kidney function deterioration in newborns with congenital kidney disease is crucial for determining surgical intervention. However, motion artifacts in DCE-MR imaging can introduce errors in tracer kinetic model estimation. A novel LTI model is proposed to improve registration stability and accuracy, showing promising results in aligning volumes and fitting the tracer kinetic model.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Clinical Neurology
F. Machado-Rivas, O. Afacan, S. Khan, B. Marami, C. K. Rollins, C. Ortinau, C. Velasco-Annis, S. K. Warfield, A. Gholipour, C. Jaimes
Summary: A new algorithm enabled real-time tracking of developmental changes in the cerebellar peduncles and revealed significant alterations in various metrics as gestational age increased.
AMERICAN JOURNAL OF NEURORADIOLOGY
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jaume Coll-Font, Onur Afacan, Scott Hoge, Harsha Garg, Kumar Shashi, Bahram Marami, Ali Gholipour, Jeanne Chow, Simon Warfield, Sila Kurugol
Summary: DW-MRI of the kidneys can provide information about renal tissue microstructure, but physiological motion and main B-0 field inhomogeneities can lead to large geometric distortions, reducing image quality and quantitative imaging marker accuracy.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Tess E. Wallace, Onur Afacan, Camilo Jaimes, Joanne Rispoli, Kristina Pelkola, Monet Dugan, Tobias Kober, Simon K. Warfield
Summary: The integrated motion metrics from FIDnavs in structural MRI are a valuable predictor of diagnostic image quality, leading to significant time and cost savings when applied to pediatric MRI examinations.
MAGNETIC RESONANCE IN MEDICINE
(2021)
Article
Computer Science, Interdisciplinary Applications
Arvind Balachandrasekaran, Alexander L. Cohen, Onur Afacan, Simon K. Warfield, Ali Gholipour
Summary: Functional MRI (fMRI) is widely used for studying the functional organization of the brain, but the signal can be contaminated by subject motion artifacts. This study proposes a new approach to recover the censored fMRI signal using structured low rank matrix completion, reducing the adverse effects of motion in fMRI analysis.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Computer Science, Interdisciplinary Applications
Yao Sui, Onur Afacan, Camilo Jaimes, Ali Gholipour, Simon K. Warfield
Summary: High spatial resolution is important for the interpretation and analysis of Magnetic Resonance Imaging (MRI), but direct acquisition is time-consuming and costly. Super-resolution reconstruction (SRR) allows for a balance between high spatial resolution, high signal-to-noise ratio (SNR), and reduced scan times. Deep learning has improved SRR, but requires large-scale training datasets of high-resolution images. We developed a dataset-free learning method using generative neural networks for tailored SRR. With three short duration scans, we achieved high-quality brain MRI with enhanced spatial resolution and SNR. Our approach outperformed state-of-the-art methods while reducing costs.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Fedel Machado-Rivas, Jasmine Gandhi, Jungwhan John Choi, Clemente Velasco-Annis, Onur Afacan, Simon K. Warfield, Ali Gholipour, Camilo Jaimes
Summary: This study used a spatiotemporal MRI atlas to evaluate the volumetric growth of intracranial structures in healthy fetuses, taking into account gestational age, sex, and laterality. The results showed that the volumetric growth of the fetal brain exhibited complex trajectories dependent on structure, gestational age, sex, and laterality.
Article
Radiology, Nuclear Medicine & Medical Imaging
Tess E. Wallace, Tobias Kober, Jason P. Stockmann, Jonathan R. Polimeni, Simon K. Warfield, Onur Afacan
Summary: In this study, a method for real-time field control using rapid FID navigator measurements was implemented, and its effectiveness in mitigating dynamic field perturbations and improving image quality was evaluated. The results showed that real-time shimming with FID navigator successfully compensated for field offsets and reduced field fluctuations, leading to an improvement in image quality.
MAGNETIC RESONANCE IN MEDICINE
(2022)
Article
Neurosciences
Benjamin F. Zwick, George C. Bourantas, Saima Safdar, Grand R. Joldes, Damon E. Hyde, Simon K. Warfield, Adam Wittek, Karol Miller
Summary: Invasive intracranial electroencephalography (iEEG) or electrocorticography (ECoG) can be used for epilepsy surgery planning, but accurate solution of the forward problem requires accurate representation of patient's brain geometry and tissue conductivity after electrode implantation. A biomechanics-based image warping procedure using preoperative MRI and postoperative CT allows non-rigid registration of preoperative image data to postoperative configuration, enabling accurate solution of the iEEG forward problem on deformed brain geometry.
Article
Multidisciplinary Sciences
Arka N. Mallela, Hansen Deng, Ali Gholipour, Simon K. Warfield, Ezequiel Goldschmidt
Summary: The human cerebrum has a specific arrangement of lobes, primary gyri, and connectivity that underlies cognition. The development of this arrangement is not well understood, and current models do not explain the global configuration of the cerebral lobes. The insula, a part of the cerebrum buried in the Sylvian fissure, has unique morphology and location. This study quantitatively examines the unique development of the insula and identifies differences in migration patterns that may explain these unique characteristics.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Tess E. Wallace, Davide Piccini, Tobias Kober, Simon K. Warfield, Onur Afacan
Summary: This study aims to develop a self-navigated motion compensation strategy for 3D radial MRI, which can compensate for continuous head motion by measuring rigid body motion parameters from the central k-space acquisition point. The results show that self-encoded FID navigators can accurately obtain rigid body motion parameters and substantially improve image quality through retrospective correction. This approach is suitable for subjects with frequent head motion.
MAGNETIC RESONANCE IN MEDICINE
(2023)
Article
Computer Science, Artificial Intelligence
Serge Didenko Vasylechko, Simon K. Warfield, Sila Kurugol, Onur Afacan
Summary: We present a generative model for synthesizing large-scale 3D datasets for quantitative mapping of myelin water fraction (MWF). Our model combines MR physics signal decay model with a probabilistic multi-component T2 model, and can be used to train CNN models for MWF estimation. Experimental results show that our synthetically trained CNN achieves superior accuracy compared to competing methods.
MEDICAL IMAGE ANALYSIS
(2024)
Proceedings Paper
Computer Science, Artificial Intelligence
Noam Korngut, Elad Rotman, Onur Afacan, Sila Kurugol, Yael Zaffrani-Reznikov, Shira Nemirovsky-Rotman, Simon Warfield, Moti Freiman
Summary: The SUPER-IVIM-DC method introduced in this study enables IVIM analysis of DWI data with a limited number of b-values, reducing the long acquisition times associated with this analysis. This method provides clinically feasible biomarkers for non-invasive fetal lung maturity assessment.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT II
(2022)
Article
Computer Science, Artificial Intelligence
Hong Liu, Dong Wei, Donghuan Lu, Xiaoying Tang, Liansheng Wang, Yefeng Zheng
Summary: This study proposes a framework based on hybrid 2D-3D convolutional neural networks for obtaining continuous 3D retinal layer surfaces from OCT volumes. The framework works well with both full and sparse annotations and utilizes alignment displacement vectors and layer segmentation to align the B-scans and segment the layers. Experimental results show that the framework outperforms state-of-the-art 2D deep learning methods in terms of layer segmentation accuracy and cross-B-scan 3D continuity.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Simon Oxenford, Ana Sofia Rios, Barbara Hollunder, Clemens Neudorfer, Alexandre Boutet, Gavin J. B. Elias, Jurgen Germann, Aaron Loh, Wissam Deeb, Bryan Salvato, Leonardo Almeida, Kelly D. Foote, Robert Amaral, Paul B. Rosenberg, David F. Tang-Wai, David A. Wolk, Anna D. Burke, Marwan N. Sabbagh, Stephen Salloway, M. Mallar Chakravarty, Gwenn S. Smith, Constantine G. Lyketsos, Michael S. Okun, William S., Zoltan Mari, Francisco A. Ponce, Andres Lozano, Wolf-Julian Neumann, Bassam Al-Fatly, Andreas Horn
Summary: Spatial normalization is a method to map subject brain images to an average template brain, allowing comparison of brain imaging results. We introduce a novel tool called WarpDrive, which enables manual refinements of image alignment after automated registration. The tool improves accuracy of data representation and aids in understanding patient outcomes.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Ricards Marcinkevics, Patricia Reis Wolfertstetter, Ugne Klimiene, Kieran Chin-Cheong, Alyssia Paschke, Julia Zerres, Markus Denzinger, David Niederberger, Sven Wellmann, Ece Ozkan, Christian Knorr, Julia E. Vogt
Summary: This study presents interpretable machine learning models for predicting the diagnosis, management, and severity of suspected appendicitis using ultrasound images. The proposed models utilize concept bottleneck models (CBM) that facilitate interpretation and intervention by clinicians, without compromising performance or requiring time-consuming image annotation.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Jian-Qing Zheng, Ziyang Wang, Baoru Huang, Ngee Han Lim, Bartlomiej W. Papiez
Summary: This article introduces a new method for medical image registration, which utilizes a separable motion backbone and a residual aligner module to better handle the discontinuous motion of multiple neighboring objects. The proposed method achieves excellent registration results on abdominal CT scans and lung CT scans.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Xiangqiong Wu, Guanghua Tan, Hongxia Luo, Zhilun Chen, Bin Pu, Shengli Li, Kenli Li
Summary: This study develops a user-friendly framework for the automated diagnosis of thyroid nodules in ultrasound videos, simulating the diagnostic workflow of radiologists. By interpreting image characteristics and modeling temporal contextual information, the efficiency and generalizability of the diagnosis can be improved.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Riddhish Bhalodia, Shireen Elhabian, Jadie Adams, Wenzheng Tao, Ladislav Kavan, Ross Whitaker
Summary: This paper introduces DeepSSM, a deep learning-based framework for image-to-shape modeling. By learning the functional mapping from images to low-dimensional shape descriptors, DeepSSM can directly infer statistical representation of anatomy from 3D images. Compared to traditional methods, DeepSSM eliminates the need for heavy manual preprocessing and segmentation, and significantly improves computational time.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Florentin Liebmann, Marco von Atzigen, Dominik Stutz, Julian Wolf, Lukas Zingg, Daniel Suter, Nicola A. Cavalcanti, Laura Leoty, Hooman Esfandiari, Jess G. Snedeker, Martin R. Oswald, Marc Pollefeys, Mazda Farshad, Philipp Furnstahl
Summary: This study presents a marker-less approach for automatic registration and real-time navigation of lumbar spinal fusion surgery using a deep neural network, avoiding radiation exposure and surgical errors. The method was validated on an ex-vivo surgery and a public dataset.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Piyush Tiwary, Kinjawl Bhattacharyya, A. P. Prathosh
Summary: Domain shift refers to the change of distributional characteristics between training and testing datasets, leading to performance drop. For medical image tasks, domain shift can be caused by changes in imaging modalities, devices, and staining mechanisms. Existing approaches based on generative models suffer from training difficulties and lack of diversity. In this paper, the authors propose the use of energy-based models (EBMs) for unpaired image-to-image translation in medical images. The proposed method, called Cycle Consistent Twin EBMs (CCT-EBM), employs a pair of EBMs in the latent space of an Auto-Encoder to ensure translation symmetry and coupling between domains.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Yutong Xie, Jianpeng Zhang, Lingqiao Liu, Hu Wang, Yiwen Ye, Johan Verjans, Yong Xia
Summary: This paper proposes a hybrid pre-training paradigm that combines self-supervised learning and supervised learning to improve the representation quality for medical image segmentation tasks. It introduces a reference task in self-supervised learning and optimizes the model using a gradient matching method. The experimental results demonstrate the effectiveness of this approach on multiple medical image segmentation benchmarks.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Youyi Song, Jing Zou, Kup-Sze Choi, Baiying Lei, Jing Qin
Summary: Cell classification is crucial for intelligent cervical cancer screening, but the variation in cells' appearance and shape poses challenges. A new learning algorithm, worse-case boosting, is proposed to improve classification accuracy for under-represented data. Experimental results demonstrate the effectiveness of this algorithm in two publicly available datasets, achieving a 4% improvement in accuracy.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Sangjoon Park, Eun Sun Lee, Kyung Sook Shin, Jeong Eun Lee, Jong Chul Ye
Summary: The increasing demand for AI systems to monitor human errors and abnormalities in healthcare presents challenges. This study presents a model called Medical X-VL, which is tailored for the medical domain and outperformed current state-of-the-art models in two medical image datasets. The model enables various zero-shot tasks for monitoring AI in the medical domain.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Anna Klimovskaia Susmelj, Berkan Lafci, Firat Ozdemir, Neda Davoudi, Xose Luis Dean-Ben, Fernando Perez-Cruz, Daniel Razansky
Summary: Optoacoustic imaging is a technique that uses optical excitation and ultrasound detection for biological tissue imaging. The quality of the images depends on the extent of tomographic coverage provided by the ultrasound detector arrays. However, full coverage is not always possible due to experimental constraints. The proposed signal domain adaptation network aims to reduce limited-view artifacts in the images.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Srijay Deshpande, Muhammad Dawood, Fayyaz Minhas, Nasir Rajpoot
Summary: In this work, a novel framework called SynCLay is proposed for automated synthesis of histology images based on user-defined cellular layouts. The framework can generate realistic and high-quality histology images with different cellular arrangements, which is helpful for studying the role of cells in the tumor microenvironment. The framework integrates a nuclear segmentation and classification model to refine nuclear structures and generate nuclear masks. Evaluation using quantitative metrics and feedback from pathologists shows that the synthetic images generated by SynCLay have high realism scores and can accurately differentiate between benign and malignant tumors.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Ahmed H. Shahin, An Zhao, Alexander C. Whitehead, Daniel C. Alexander, Joseph Jacob, David Barber
Summary: Survival analysis is a valuable tool in healthcare for predicting the time to specific events. This paper introduces CenTime, a novel approach that directly estimates the time to event. The method performs well with censored data and can be easily integrated with deep learning models. Compared to standard methods, CenTime offers superior performance in predicting event time while maintaining comparable ranking performance.
MEDICAL IMAGE ANALYSIS
(2024)
Article
Computer Science, Artificial Intelligence
Bingyuan Liu, Jose Dolz, Adrian Galdran, Riadh Kobbi, Ismail Ben Ayed
Summary: Most segmentation losses, such as CE and Dice, are variants of the Cross-Entropy or Dice losses. This work provides a theoretical analysis that shows a deeper connection between CE and Dice than previously thought. From a constrained-optimization perspective, both CE and Dice decompose into similar ground-truth matching terms and region-size penalty terms. The analysis uncovers hidden region-size biases: Dice has an intrinsic bias towards extremely imbalanced solutions, while CE implicitly encourages the ground-truth region proportions. Based on this analysis, a principled and simple solution is proposed to explicitly control the region-size bias.
MEDICAL IMAGE ANALYSIS
(2024)