Article
Radiology, Nuclear Medicine & Medical Imaging
Dominik Etter, Geoff Warnock, Frederic Koszarski, Tilo Niemann, Nidaa Mikail, Susan Bengs, Ronny. R. R. Buechel, Philipp Kaufmann, Catherine Gebhard, Alexia Rossi
Summary: This study aimed to investigate the impact of tube voltage and iterative reconstruction on the mean attenuation of pericoronary adipose tissue (PCAT(MA)) derived from coronary computed tomography angiography (CCTA). It was found that both tube voltage and reconstruction type significantly affected PCAT(MA), and it is recommended to use the same tube voltage and reconstruction type in multicenter and longitudinal studies.
EUROPEAN RADIOLOGY
(2023)
Review
Cardiac & Cardiovascular Systems
Haipeng Liu, Aleksandra Wingert, Jian'an Wang, Jucheng Zhang, Xinhong Wang, Jianzhong Sun, Fei Chen, Syed Ghufran Khalid, Jun Jiang, Dingchang Zheng
Summary: The review focuses on recent studies of CT-based coronary plaque extraction, categorizing them into 2D and 3D methods. The analysis of data, methods, and evaluation in each category highlights the need for methodological innovations to improve accuracy in clinical applications. Advanced techniques such as de-blooming algorithms, standardized datasets, and machine learning could enhance the efficiency and accuracy of coronary plaque extraction in future studies.
FRONTIERS IN CARDIOVASCULAR MEDICINE
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Arwed Elias Michael, Denise Schoenbeck, Jendrik Becker-Assmann, Julius Henning Niehoff, Thomas Flohr, Bernhard Schmidt, Christoph Panknin, Matthias Baer-Beck, Tilman Hickethier, David Maintz, Alexander Christian Bunck, Jan Borggrefe, Marcus Wiemer, Volker Rudolph, Jan Robert Kroeger
Summary: This study investigates the effect of using optimized kernels on CT imaging of cardiac stents. The results show that in PCCT UHR mode, using optimized kernels significantly improves the assessment of the in-stent lumen of small cardiac stents. It is recommended to use optimized kernels for imaging of cardiac stents.
EUROPEAN JOURNAL OF RADIOLOGY
(2023)
Article
Rheumatology
George A. Karpouzas, Sarah R. Ormseth, Elizabeth Hernandez, Matthew J. Budoff
Summary: In this study, statin therapy was associated with lower long-term cardiovascular risk in rheumatoid arthritis patients, especially those with higher inflammation levels. Additionally, statin therapy modified the impact of inflammation on new coronary plaque formation and predicted regression and calcification of prevalent noncalcified lesions.
Article
Radiology, Nuclear Medicine & Medical Imaging
Hye Joo Park, Seo-Youn Choi, Ji Eun Lee, Sanghyeok Lim, Min Hee Lee, Boem Ha Yi, Jang Gyu Cha, Ji Hye Min, Bora Lee, Yunsub Jung
Summary: This study compared the image quality and radiation dose of a deep learning image reconstruction algorithm (DLIR) with iterative reconstruction (IR) and filtered back projection (FBP) at different tube voltages and tube currents. The results showed that DLIR significantly reduced noise and artifacts and improved overall image quality compared to FBP and hybrid IR. Despite the reduced image sharpness, low-dose CT with DLIR seemed to have a greater potential for dose optimization.
EUROPEAN RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Salim A. Si-Mohamed, Joel Greffier, Jade Miailhes, Sara Boccalini, Pierre-Antoine Rodesch, Aurelie Vuillod, Niels van der Werf, Djamel Dabli, Damien Racine, David Rotzinger, Fabio Becce, Yoad Yagil, Philippe Coulon, Alain Vlassenbroek, Loic Boussel, Jean-Paul Beregi, Philippe Douek
Summary: The study evaluated the image quality of spectral photon-counting CT (SPCCT) compared to dual-layer CT (DLCT) with different reconstruction algorithms. Results showed that SPCCT had lower noise magnitude and higher detectability for nodules compared to DLCT. SPCCT provided higher image quality and better conspicuity for both ground-glass nodules and solid nodules at different iDose(4) levels.
EUROPEAN RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yu-Kun Pan, Ming-Hua Sun, Jia-Jia Wang, Xing-Biao Chen, Xiao-Jing Kan, Ying-Hui Ge, Zhi-Ping Guo
Summary: The study found that the RRD CAC scoring scan using the IMR reconstruction algorithm is clinically feasible, and a correction factor can help reduce the AS underestimation effect.
QUANTITATIVE IMAGING IN MEDICINE AND SURGERY
(2021)
Article
Medicine, General & Internal
Pil-Hyun Jeon, Sang-Hyun Jeon, Donghee Ko, Giyong An, Hackjoon Shim, Chuluunbaatar Otgonbaatar, Kihong Son, Daehong Kim, Sung Min Ko, Myung-Ae Chung
Summary: This study compared the image quality of coronary computed tomography angiography (CCTA) using deep learning-based reconstruction (DLR), filtered back projection (FBP), and iterative reconstruction (IR). The results showed that DLR effectively reduced noise and improved signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) compared to FBP and IR. Therefore, DLR may be useful for CCTA examinations.
Article
Radiology, Nuclear Medicine & Medical Imaging
Alexia Rossi, Antonio G. Gennari, Dominik Etter, Dominik C. Benz, Thomas Sartoretti, Andreas A. Giannopoulos, Nidaa Mikail, Susan Bengs, Alexander Maurer, Catherine Gebhard, Ronny R. Buechel, Philipp A. Kaufmann, Tobias A. Fuchs, Michael Messerli
Summary: Deep learning image reconstruction (DLIR) systematically underestimates Agatston coronary artery calcium (CAC) score, suggesting caution in its use for cardiovascular risk assessment.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Changjing Feng, Rui Chen, Siting Dong, Wei Deng, Shushen Lin, Xiaomei Zhu, Wangyan Liu, Yi Xu, Xiaohu Li, Yinsu Zhu
Summary: By analyzing clinical data and CCTA images of 400 patients, it was found that FAI and NCPB were independent risk factors for coronary plaque progression. Combining conventional parameters with radiomics features derived from CCTA can better predict plaque progression.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Xiuxiu He, Bang Jun Guo, Yang Lei, Tonghe Wang, Walter J. Curran, Tian Liu, Long Jiang Zhang, Xiaofeng Yang
Summary: The proposed deep learning-based segmentation method demonstrated high accuracy in quantifying myocardium and pericardial fat from CCTA data, with a median Dice similarity coefficient of 0.88 for pericardial fat and 0.96 for myocardium.
EUROPEAN RADIOLOGY
(2021)
Review
Engineering, Biomedical
Dong Zeng, Cuidie Zeng, Zhixiong Zeng, Sui Li, Zhen Deng, Sijin Chen, Zhaoying Bian, Jianhua Ma
Summary: This paper provides an insight into computed tomography perfusion (CTP) imaging by covering its basics, current state, technical applications, and future potential. It focuses on the fundamentals of CTP imaging, including image acquisition and parameter estimation techniques. Various clinical applications of CTP imaging are discussed, along with the radiation dose effect. The paper also reviews the challenges and methods for reducing radiation dose in CTP imaging. Standardized performance metrics for evaluating CTP images are listed, as well as the determination of infarct and penumbra. The popularity and future trend of CTP imaging are revealed.
PHYSICS IN MEDICINE AND BIOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Ji He, Shilin Chen, Hua Zhang, Xi Tao, Wuhong Lin, Shanli Zhang, Dong Zeng, Jianhua Ma
Summary: Accurate CT image reconstruction can be achieved through downsampling imaging geometric modeling using deep-learning techniques. The proposed DSigNet combines geometric modeling knowledge of the CT imaging system with data-driven training for accurate CT image reconstruction, potentially improving image quality and speeding up reconstruction for modern CT systems.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jingyu Zhong, Yihan Xia, Yong Chen, Jianying Li, Wei Lu, Xiaomeng Shi, Jianxing Feng, Fuhua Yan, Weiwu Yao, Huan Zhang
Summary: This study compared the image quality between a deep learning image reconstruction (DLIR) algorithm and conventional iterative reconstruction (IR) algorithms in dual-energy CT (DECT) and assessed their impact on radiomics robustness. The results showed that DLIR significantly improved the image quality of DECT, but may alter radiomics features compared to conventional IR. Nine robust DECT radiomics features were identified.
EUROPEAN RADIOLOGY
(2023)
Article
Cardiac & Cardiovascular Systems
Camilla Nordheim Solli, Sandra Chamat-Hedemand, Hanne Elming, Anh Ngo, Lasse Kjaer, Vibe Skov, Anders Lindholm Sorensen, Christina Ellervik, Andreas Fuchs, Per Ejlstrup Sigvardsen, Jorgen Tobias Kuhl, Klaus Fuglsang Kofoed, Borge G. Nordestgaard, Hans Hasselbalch, Niels Eske Bruun
Summary: This study investigated whether patients with Philadelphia-negative Myeloproliferative Neoplasms (MPNs) have an increased burden of cardiac calcification and found that MPNs patients have a higher prevalence of coronary artery calcium score (CACS) and aortic valve calcification (AVC) compared to the general population. This association remains significant after adjusting for cardiovascular risk factors.
INTERNATIONAL JOURNAL OF CARDIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Xiaonan Cui, Sunyi Zheng, Marjolein A. Heuvelmans, Yihui Du, Grigory Sidorenkov, Shuxuan Fan, Yanju Li, Yongsheng Xie, Zhongyuan Zhu, Monique D. Dorrius, Yingru Zhao, Raymond N. J. Veldhuis, Geertruida H. de Bock, Matthijs Oudkerk, Peter M. A. van Ooijen, Rozemarijn Vliegenthart, Zhaoxiang Ye
Summary: This study evaluated the performance of a deep learning-based computer-aided detection (DL-CAD) system in a Chinese low-dose CT (LDCT) lung cancer screening program. The results showed that the DL-CAD system accurately detected pulmonary nodules with higher sensitivity and lower false-positive rate compared to double reading.
EUROPEAN JOURNAL OF RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yeshaswini Nagaraj, Hendrik Joost Wisselink, Mieneke Rook, Jiali Cai, Sunil Belur Nagaraj, Grigory Sidorenkov, Raymond Veldhuis, Matthijs Oudkerk, Rozemarijn Vliegenthart, Peter van Ooijen
Summary: The objective of this study is to evaluate the feasibility of a disease-specific deep learning model based on minimum intensity projection for automated emphysema detection in low-dose computed tomography scans. The study found that the DL model using minIP can automatically detect emphysema in LDCT scans, and thicker minIP slabs perform better.
JOURNAL OF DIGITAL IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Baoqiang Ma, Jiapan Guo, Tian-Tian Zhai, Arjen van der Schaaf, Roel J. H. M. Steenbakkers, Lisanne V. van Dijk, Stefan Both, Johannes A. Langendijk, Weichuan Zhang, Bingjiang Qiu, Peter M. A. van Ooijen, Nanna M. Sijtsema
Summary: This study developed a deep learning model to predict multiple and associated efficacy endpoints in oropharyngeal squamous cell carcinoma (OPSCC) patients based on computed tomography (CT). The multi-label learning models outperformed the single endpoint models, especially for 2-year regional control, distant metastasis-free survival, disease-specific survival, overall survival, and disease-free survival, with high AUC values.
Article
Engineering, Biomedical
Alessia De Biase, Nanna M. Sijtsema, Lisanne van Dijk, Johannes A. Langendijk, Peter M. A. van Ooijen
Summary: This study proposes a novel deep learning-based method that generates probability maps to capture the model uncertainty in tumor segmentation. The method was evaluated on 138 oropharyngeal cancer patients and showed promising results, offering guidance for radiation oncologists in slice-by-slice adaptive GTVp segmentation.
PHYSICS IN MEDICINE AND BIOLOGY
(2023)
Article
Oncology
Sunyi Zheng, Jiapan Guo, Johannes A. Langendijk, Stefan Both, Raymond N. J. Veldhuis, Matthijs Oudkerk, Peter M. A. van Ooijen, Robin Wijsman, Nanna M. Sijtsema
Summary: This study aimed to develop and evaluate a prediction model for 2-year overall survival (OS) in stage I-IIIA non-small cell lung cancer (NSCLC) patients who received definitive radiotherapy. The hybrid model, which integrated clinical variables and image features from pre-treatment CT scans, achieved reasonable performance and has the potential to identify high mortality risk patients and guide clinical decision making.
RADIOTHERAPY AND ONCOLOGY
(2023)
Article
Multidisciplinary Sciences
Jingxuan Wang, Nikos Sourlos, Sunyi Zheng, Nils Van der Velden, Gert Jan Pelgrim, Rozemarijn Vliegenthart, Peter van Ooijen
Summary: This paper emphasizes the importance of deep learning in the automatic detection, segmentation, and classification of pulmonary nodules in CT images. It provides a detailed guide on the data preparation steps, including permission, access, annotation, and preprocessing. Four popular datasets are used in the preparation process. Researchers should carefully select datasets, annotation methods, and preprocessing techniques based on their specific research questions.
Article
Radiology, Nuclear Medicine & Medical Imaging
I. Iris Hamelink, E. Erik Jan de Heide, G. J. Gert Jan Pelgrim, T. C. Thomas Kwee, P. M. A. Peter van Ooijen, G. H. Truuske de Bock, R. Rozemarijn Vliegenthart
Summary: This study evaluated the performance of AI software for automatic thoracic aortic diameter assessment in low-dose, non-contrast chest CT. The results showed that the AI software can accurately measure aortic diameters and is comparable to human readers in terms of measurement consistency.
EUROPEAN JOURNAL OF RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Xueping Jing, Mirjam Wielema, Andrea G. Monroy-Gonzalez, Thom R. G. Stams, Shekar V. K. Mahesh, Matthijs Oudkerk, Paul E. Sijens, Monique D. Dorrius, Peter M. A. van Ooijen
Summary: This study evaluates the feasibility of reducing inter-observer variability in breast density assessment through AI-assisted interpretation. Deep learning and radiomics models were developed and tested against a reference standard on an independent test set.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Oncology
Baoqiang Ma, Jiapan Guo, Hung Chu, Lisanne V. van Dijk, Peter M. A. van Ooijen, Johannes A. Langendijk, Stefan Both, Nanna M. Sijtsema
Summary: This study compared the prediction performance of radiomics, self-supervised learning, and end-to-end deep learning for OPSCC patients after (chemo)radiotherapy. The results showed that features extracted using self-supervised learning had the best internal prediction performance, while radiomics features had better external generalizability.
PHYSICS & IMAGING IN RADIATION ONCOLOGY
(2023)
Article
Engineering, Multidisciplinary
Lu Liu, Runlei Ma, Peter M. A. van Ooijen, Matthijs Oudkerk, Rozemarijn Vliegenthart, Raymond N. J. Veldhuis, Christoph Brune
Summary: This study aimed to automatically assess epicardial adipose tissue (EAT) on non-contrast low-dose CT calcium score images using advanced U-net methods. The findings showed that using labels representing the region inside the pericardium improved the accuracy of EAT segmentation, and 3D convolutional neural networks did not consistently outperform 2D networks. In conclusion, deep learning-based methods have potential for robust EAT segmentation and quantification.
Article
Radiology, Nuclear Medicine & Medical Imaging
Dewinda Julianensi Rumala, Peter van Ooijen, Reza Fuad Rachmadi, Anggraini Dwi Sensusiati, I. Ketut Eddy Purnama
Summary: This paper presents Deep-Stacked CNN, a deep heterogeneous model based on stacked generalization, which uses multiple CNNs to improve the robustness and accuracy of multi-class brain disease classification.
JOURNAL OF DIGITAL IMAGING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Tugba Akinci D'Antonoli, Armando Ugo Cavallo, Federica Vernuccio, Arnaldo Stanzione, Michail E. Klontzas, Roberto Cannella, Lorenzo Ugga, Agah Baran, Salvatore Claudio Fanni, Ekaterina Petrash, Ilaria Ambrosini, Luca Alessandro Cappellini, Peter van Ooijen, Elmar Kotter, Daniel Pinto dos Santos, Renato Cuocolo
Summary: This study investigated the reliability of the total radiomics quality score (RQS) and the reproducibility of individual RQS items' score in a large multireader study. The results showed low inter-rater reliability for total RQS and moderate to good intra-rater reliability for individual RQS items' score. There is a need for a robust and reproducible assessment method to improve the quality of radiomics research.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Andrea Ponsiglione, Arnaldo Stanzione, Gaia Spadarella, Agah Baran, Luca Alessandro Cappellini, Kevin Groot Lipman, Peter Van Ooijen, Renato Cuocolo
Summary: The overall methodological rigor of radiomics studies in the ovarian field is not ideal, with a lack of prospective design and formal validation of results. This limits the reproducibility of results and potential translation to clinical setting.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yeshaswini Nagaraj, Gonda de Jonge, Anna Andreychenko, Gabriele Presti, Matthias A. Fink, Nikolay Pavlov, Carlo C. Quattrocchi, Sergey Morozov, Raymond Veldhuis, Matthijs Oudkerk, Peter M. A. van Ooijen
Summary: An automatic COVID-19 Reporting and Data System (CO-RADS)-based classification was developed in a multi-demographic setting. The study achieved clinically acceptable performance and can serve as a standardized tool for automated COVID-19 assessment. Inter-observer agreement for CO-RADS scoring was significant, and suspected COVID-19 CT scans were identified with an accuracy of 84%.
EUROPEAN RADIOLOGY
(2022)
Article
Health Care Sciences & Services
L. B. van den Oever, D. S. Spoor, A. P. G. Crijns, R. Vliegenthart, M. Oudkerk, R. N. J. Veldhuis, G. H. de Bock, P. M. A. van Ooijen
Summary: This study developed an automatic cardiac structure segmentation pipeline for use in low-dose non-contrast planning CT. The pipeline achieved good results in contouring the whole heart and ventricles, indicating the feasibility of robust automatic contouring with deep learning methods in local centers with small datasets.
JOURNAL OF MEDICAL SYSTEMS
(2022)