Article
Computer Science, Interdisciplinary Applications
Zhan Gao, Qiuhao Zong, Yiqi Wang, Yan Yan, Yuqing Wang, Ning Zhu, Jin Zhang, Yunfu Wang, Liang Zhao
Summary: Liver vessel segmentation from computed tomography is a challenging task due to small vessel size and imbalanced distribution of vessels and liver tissues. This study proposes a sophisticated model and elaborated dataset, utilizing a newly conceived Laplacian salience filter and pyramid deep learning architecture. Experimental results show significant improvement over existing methods, achieving higher Dice scores on available datasets.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Chemistry, Analytical
Ahmed Sharafeldeen, Mohamed Elsharkawy, Norah Saleh Alghamdi, Ahmed Soliman, Ayman El-Baz
Summary: A new segmentation technique using a probabilistic model and modified algorithm was proposed to accurately delineate lung structures in CT images, showing promising results in COVID-19 patients' data.
Article
Instruments & Instrumentation
Wenjun Tan, Luyu Zhou, Xiaoshuo Li, Xiaoyu Yang, Yufei Chen, Jinzhu Yang
Summary: This paper reviews 12 different pulmonary vascular segmentation algorithms for lung CT and CTA images, and objectively evaluates their performances. Most of the algorithms show admirable performance in pulmonary vascular extraction and segmentation, with the top three algorithms having dice coefficients around 0.80. Integrating methods that consider spatial information, fuse multi-scale feature map, or have excellent post-processing to deep neural network training and optimization process are significant for improving the accuracy of pulmonary vascular segmentation.
JOURNAL OF X-RAY SCIENCE AND TECHNOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
David Olayemi Alebiosu, Anuja Dharmaratne, Chern Hong Lim
Summary: Tuberculosis is a bacterial infection that affects the lungs. This study proposes a novel approach to segment tuberculosis-affected areas in CT images using a new model called DAvoU-Net. The proposed model out-performs existing methods and achieves higher accuracy in tuberculosis image segmentation.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Chemistry, Physical
K. A. A. Aziz, M. I. Saripan, F. F. A. Saad, R. S. A. R. Abdullah, N. Waeleh
Summary: This study investigated the performance of computed tomography lung classification using image processing and Markov Random Field. A multilevel thresholding and Markov Random Field approach was proposed to improve the segmentation process. The results revealed that Markov Random Field using Metropolis algorithm gave the best performance for CT image lung classification. The output from this study is important for lung cancer analysis research and computer aided diagnosis development.
RADIATION PHYSICS AND CHEMISTRY
(2022)
Article
Clinical Neurology
Shalini A. Amukotuwa, Angel Wu, Kevin Zhou, Inna Page, Peter Brotchie, Roland Bammer
Summary: The study found that Tmax maps derived from CT perfusion can accurately and rapidly identify patients with DMVOs, demonstrating high sensitivity and specificity for this condition.
Article
Computer Science, Interdisciplinary Applications
Haowen Pang, Shouliang Qi, Yanan Wu, Meihuan Wang, Chen Li, Yu Sun, Wei Qian, Guoyan Tang, Jiaxuan Xu, Zhenyu Liang, Rongchang Chen
Summary: In this study, two synthesizers were developed to achieve mutual synthesis between non-contrast CT (NCCT) and contrast-enhanced CT (CECT) using generative adversarial networks. The results demonstrated the effectiveness of the synthesizers in high-quality synthesis of NCCT and CECT images, with the training process being crucial to their performance.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Ilkay Yildiz Potter, Diana Yeritsyan, Sarah Mahar, Jim Wu, Ara Nazarian, Aidin Vaziri, Ashkan Vaziri
Summary: The purpose of this study was to develop an automated method using CT imaging and machine learning for bone tumor segmentation and classification to assist clinicians in determining the need for biopsy. A dataset of 84 femur CT scans with confirmed bone lesions was used to train a deep learning model for tumor segmentation and classification. The results showed similar classification performance to existing deep learning models and demonstrated the potential of the proposed approach in aiding clinical decision-making for biopsy.
JOURNAL OF DIGITAL IMAGING
(2023)
Article
Computer Science, Information Systems
Hyung Min Kim, Taehoon Ko, In Young Choi, Jun-Pyo Myong
Summary: This study developed an algorithm that combines lung segmentation and deep learning models to diagnose patients with asbestosis in segmented CT images with high accuracy. The algorithm outperformed radiologists in diagnosing asbestosis.
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS
(2022)
Article
Engineering, Electrical & Electronic
Kai Hu, Hui Tan, Yuan Zhang, Wei Huang, Xieping Gao
Summary: Recently, accurate segmentation of COVID-19 infection from CT scans has become crucial for its diagnosis and treatment. To address the challenges posed by various infection characteristics, a novel multiscale wavelet guidance network (MWG-Net) is proposed, which integrates wavelet domain information into the CNN encoder and decoder. Experimental results demonstrate the superiority of MWG-Net in COVID-19 lung infection segmentation, outperforming state-of-the-art methods on different datasets.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Computer Science, Artificial Intelligence
Jiannan Liu, Bo Dong, Shuai Wang, Hui Cui, Deng-Ping Fan, Jiquan Ma, Geng Chen
Summary: This study proposes a novel two-stage cross-domain transfer learning framework for accurately segmenting COVID-19 lung infections from CT images. Through effective infection segmentation and a novel transfer learning strategy, the framework addresses issues such as low boundary contrast and large infection variations, showing superior segmentation accuracy in experiments.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Health Care Sciences & Services
Negar Farzaneh, Craig A. Williamson, Jonathan Gryak, Kayvan Najarian
Summary: Machine learning framework proposed in this study offers an explainable approach for predicting the long-term functional outcomes in traumatic brain injury patients. Through the incorporation of statistical inference and human expert validation layers, the predicted risk scores are made more trustworthy and reliable.
NPJ DIGITAL MEDICINE
(2021)
Article
Computer Science, Information Systems
Ruikun Li, Yi-Jie Huang, Huai Chen, Xiaoqing Liu, Yizhou Yu, Dahong Qian, Lisheng Wang
Summary: In this paper, a method of segmenting hepatic vessels by utilizing the connectivity prior is proposed. By integrating a graph neural network (GNN) into a convolutional neural network (CNN), the graphical connectivity information of hepatic vessels is modeled for segmentation. Experimental results demonstrate that the proposed method outperforms related works in terms of accuracy and connectivity.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Medicine, General & Internal
Chandra Sekhara Rao Annavarapu, Samson Anosh Babu Parisapogu, Nikhil Varma Keetha, Praveen Kumar Donta, Gurindapalli Rajita
Summary: This article proposes an end-to-end deep learning approach for lung nodule segmentation, which incorporates a Bi-FPN between an encoder and a decoder architecture. The proposed model outperforms existing deep learning models and achieves a Dice Similarity Coefficient of over 80% on both LUNA-16 and QIN Lung CT datasets.
Article
Radiology, Nuclear Medicine & Medical Imaging
Q. Li, X. Li, X-Y Li, X-Q He, Z-G Chu, T-Y Luo
Summary: The study investigated the value of DESCT in evaluating the histological subtypes of SILADC. Significant differences were found between two groups with different prognosis. DESCT showed potential clinical application prospects in assessing the histological subtypes of SILADC.
CLINICAL RADIOLOGY
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Espen Asak Ruud, Knut Stavem, Jonn Terje Geitung, Arne Borthne, Vidar Soyseth, Haseem Ashraf
Summary: This study analyzed predictors of pneumothorax and chest drainage after CT-guided lung biopsy, finding significant factors such as emphysema, needle time, and insertion location. The study revealed a pneumothorax rate of 39% and a chest drainage rate of 10%, with lateral body position and insertion through interlobar fissure being important predictors.
EUROPEAN RADIOLOGY
(2021)
Article
Genetics & Heredity
Laurike Harlaar, Pierluigi Ciet, Gijs van Tulder, Alice Pittaro, Harmke A. van Kooten, Nadine A. M. E. van der Beek, Esther Brusse, Piotr A. Wielopolski, Marleen de Bruijne, Ans T. van der Ploeg, Harm A. W. M. Tiddens, Pieter A. van Doorn
Summary: This study aimed to identify early signs of diaphragmatic weakness in Pompe patients using chest MRI. Results showed that even in early-stage Pompe disease, the motion of the diaphragm is reduced and the shape is more curved during inspiration. MRI can be used to detect early signs of diaphragmatic weakness in Pompe patients, which might help to select patients for early intervention.
ORPHANET JOURNAL OF RARE DISEASES
(2021)
Article
Clinical Neurology
Thom S. Lysen, Pinar Yilmaz, Florian Dubost, M. Arfan Ikram, Marleen de Bruijne, Meike W. Vernooij, Annemarie Luik
Summary: Higher sleep efficiency measured by actigraphy was found to be associated with increased perivascular space load in the centrum semiovale of the brain, contrary to the hypothesis. No other sleep characteristics were found to be associated with perivascular space load in other brain regions in this middle-aged and elderly population.
JOURNAL OF SLEEP RESEARCH
(2022)
Review
Radiology, Nuclear Medicine & Medical Imaging
Ivan Dudurych, Susan Muiser, Niall McVeigh, Huib A. M. Kerstjens, Maarten van den Berge, Marleen de Bruijne, Rozemarijn Vliegenthart
Summary: Research on computed tomography (CT) bronchial parameter measurements shows conflicting results, and there is heterogeneity in the methodology and population of the studies. Significant differences exist between populations for parameters such as wall area percentage, but there is overlap in their ranges.
EUROPEAN RADIOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Shuai Chen, Zahra Sedghi Gamechi, Florian Dubost, Gijs van Tulder, Marleen de Bruijne
Summary: Posterior-CRF is a segmentation method that incorporates CNN-learned features in a CRF for medical image segmentation, outperforming existing methods in terms of performance metrics across various tasks.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Clinical Neurology
Laurike Harlaar, Pierluigi Ciet, Gijs van Tulder, Esther Brusse, Remco G. M. Timmermans, Wim G. M. Janssen, Marleen de Bruijne, Ans T. van der Ploeg, Harm A. W. M. Tiddens, Pieter A. van Doorn, Nadine A. M. E. van der Beek
Summary: The aim of this exploratory study was to evaluate diaphragmatic function across various neuromuscular diseases using spirometry-controlled MRI. The results showed that the diaphragmatic function was impaired in patients with myopathies and motor neuron diseases, and significantly abnormal in Pompe patients. The study suggests that spirometry-controlled MRI can be used to investigate respiratory dysfunction in neuromuscular diseases.
NEUROMUSCULAR DISORDERS
(2022)
Article
Physics, Multidisciplinary
Jon Sporring, Sune Darkner
Summary: This paper examines the representational power of local overlapping histograms for discrete binary signals. It presents an algorithm, with linear complexity in signal size and factorial complexity in window size, for generating a set of signals that share a sequence of densely overlapping histograms. The paper also provides values for the sizes of the number of unique signals for a given set of histograms, as well as bounds on the number of metameric classes.
Article
Computer Science, Artificial Intelligence
Alain Lalande, Zhihao Chen, Thibaut Pommier, Thomas Decourselle, Abdul Qayyum, Michel Salomon, Dominique Ginhac, Youssef Skandarani, Arnaud Boucher, Khawla Brahim, Marleen de Bruijne, Robin Camarasa, Teresa M. Correia, Xue Feng, Kibrom B. Girum, Anja Hennemuth, Markus Huellebrand, Raabid Hussain, Matthias Ivantsits, Jun Ma, Craig Meyer, Rishabh Sharma, Jixi Shi, Nikolaos V. Tsekos, Marta Varela, Xiyue Wang, Sen Yang, Hannu Zhang, Yichi Zhang, Yuncheng Zhou, Xiahai Zhuang, Raphael Couturier, Fabrice Meriaudeau
Summary: This paper presents the results of the EMIDEC challenge, which aims to automatically assess the state of the heart after myocardial infarction (MI). The challenge focuses on distinguishing between non-infarct and pathological exams and automatically calculating the extent of myocardial infarction using deep learning methods. The results show that automatic classification of exams is achievable, and the segmentation of the myocardium is possible, although the segmentation of the diseased area needs improvement.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Clinical Neurology
Tavia E. Evans, Maria J. Knol, Petra Schwingenschuh, Katharina Wittfeld, Saima Hilal, M. Arfan Ikram, Florian Dubost, Kimberlin M. H. van Wijnen, Petra Katschnig, Pinar Yilmaz, Marleen de Bruijne, Mohamad Habes, Christopher Chen, Soenke Langer, Henry Volzke, M. Kamran Ikram, Hans J. Grabe, Reinhold Schmidt, Hieab H. H. Adams, Meike W. Vernooij
Summary: This study identified determinants of perivascular spaces (PVS) burden by pooling data from multiple cohort studies and using a uniform rating method. The results showed that PVS count increases with age, men have more PVS in the mesencephalon but less in the hippocampus, and higher blood pressure is associated with increased PVS in all regions. Furthermore, other factors such as high-density lipoprotein cholesterol levels, glucose levels, APOE genotypes, and presence of lacunes are also associated with PVS burden.
Article
Psychiatry
Nick Y. Larsen, Ninna Vihrs, Jesper Moller, Jon Sporring, Xueke Tan, Xixia Li, Gang Ji, Grazyna Rajkowska, Fei Sun, Jens R. Nyengaard
Summary: This study used imaging techniques to investigate pyramidal cells in the brain region BA46 of patients with schizophrenia and major depressive disorder. The findings suggest that these patients have lower neuron number and density, which may affect prefrontal connections, but do not affect the spatial organization of the cells.
TRANSLATIONAL PSYCHIATRY
(2022)
Article
Computer Science, Artificial Intelligence
Silas Nyboe Orting, Hans Jacob Teglbjaerg Stephensen, Jon Sporring
Summary: Mathematical morphology is an essential tool for post-processing, but there is a lack of satisfactory definitions for morphology on probabilistic representations of categorical images. This is because categories are inherently unordered. In this work, we propose two approaches to address this issue and demonstrate their effectiveness.
JOURNAL OF MATHEMATICAL IMAGING AND VISION
(2023)
Article
Multidisciplinary Sciences
Carlos Benitez Villanueva, Hans J. T. Stephensen, Rajmund Mokso, Abdellatif Benraiss, Jon Sporring, Steven A. Goldman
Summary: This study investigates the relationship between astroglial cells and medium spiny neurons (MSN) synapses in Huntington's disease (HD). The results show that HD astrocytes have impaired connection and function with synaptic sites compared to normal astrocytes, leading to striatal hyperexcitability and the development of HD.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2023)
Article
Computer Science, Interdisciplinary Applications
Jon Sporring, Sune Darkner
Summary: This article explores local overlapping histograms of functions between discrete domains and codomains, introducing a simple algebra for local histograms. By separating overlapping domains into non-overlapping ones, the article demonstrates how to estimate the size of possible histogram sets based on the local histogram domains and enumerates the number of functions that share a specific set of local histograms. Furthermore, a decoding algorithm is presented to calculate the set of functions that share a given set of overlapping histograms.
FRONTIERS IN COMPUTER SCIENCE
(2022)
Article
Imaging Science & Photographic Technology
Soumick Chatterjee, Kartik Prabhu, Mahantesh Pattadkal, Gerda Bortsova, Chompunuch Sarasaen, Florian Dubost, Hendrik Mattern, Marleen de Bruijne, Oliver Speck, Andreas Nuernberger
Summary: The pathology of blood vessels in the brain can cause serious neurodegenerative diseases. This paper proposes a deep learning architecture to automatically segment small vessels in MRI images, improving diagnosis accuracy.
JOURNAL OF IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Laurike Harlaar, Pierluigi Ciet, Gijs van Tulder, Harmke A. van Kooten, Nadine A. M. E. van der Beek, Esther Brusse, Marleen de Bruijne, Harm A. W. M. Tiddens, Ans T. van der Ploeg, Pieter A. van Doorn
Summary: MRI can evaluate the progression of Pompe disease and the effectiveness of treatment by assessing changes in diaphragmatic curvature. Once severe diaphragmatic weakness occurs, improvement in diaphragm muscle function seems unlikely.
EUROPEAN RADIOLOGY
(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)