Article
Oncology
Shivani Kumar, Lois Holloway, Miriam Boxer, Mei Ling Yap, Phillip Chlap, Daniel Moses, Shalini Vinod
Summary: This study compared the differences in GTV delineation of lung cancer using MRI and PET versus CT and PET. The results showed similar interobserver variability between MRI and PET, but there were significant differences between MRI and CT delineated volumes.
RADIOTHERAPY AND ONCOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Cedric Draulans, Robin De Roover, Uulke A. van der Heide, Linda Kerkmeijer, Robert J. Smeenk, Floris Pos, Wouter V. Vogel, James Nagarajah, Marcel Janssen, Sofie Isebaert, Frederik Maes, Cindy Mai, Raymond Oyen, Steven Joniau, Martina Kunze-Busch, Karolien Goffin, Karin Haustermans
Summary: Threshold-based contouring using SOSTs trained with GTV(majority) achieved accuracy comparable to manual contours in delineating GTV(histo). Median SOSTs were found to be 41 SUV% for 68Ga-PSMA-11 PET and 44 SUV% for 18F-PSMA-1007 PET, forming a basis for tracer-specific window leveling.
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Joseph Harms, Yang Lei, Sibo Tian, Neal S. McCall, Kristin A. Higgins, Jeffrey D. Bradley, Walter J. Curran, Tian Liu, Xiaofeng Yang
Summary: The study presents a deep learning-based algorithm for automatic generation of cardiac substructure contours, aiming to investigate the relationship between radiation dose and toxicities. Results show that the proposed method outperforms traditional methods both quantitatively and qualitatively.
Article
Oncology
Aurora Rosvoll Groendahl, Yngve Mardal Moe, Christine Kiran Kaushal, Bao Ngoc Huynh, Espen Rusten, Oliver Tomic, Eivor Hernes, Bettina Hanekamp, Christine Undseth, Marianne Gronlie Guren, Eirik Malinen, Cecilia Marie Futsaether
Summary: This study evaluated the applicability of deep learning for automatic delineation of the gross tumor volume in ASCC patients, comparing the effects of single and multimodality image combinations. Training a 2D U-Net CNN model resulted in high Dice scores for both datasets of 86 and 36 patients. The highest performance was observed with a multimodal combination of PET and ceCT.
Article
Oncology
Thomas Weissmann, Yixing Huang, Stefan Fischer, Johannes Roesch, Sina Mansoorian, Horacio Ayala Gaona, Antoniu-Oreste Gostian, Markus Hecht, Sebastian Lettmaier, Lisa Deloch, Benjamin Frey, Udo S. Gaipl, Luitpold Valentin Distel, Andreas Maier, Heinrich Iro, Sabine Semrau, Christoph Bert, Rainer Fietkau, Florian Putz
Summary: This article investigates the use of deep learning for automatic segmentation of head and neck lymph node levels. The study shows that the nnU-net 3D-fullres/2D-ensemble model can achieve highly accurate autosegmentation with a limited training dataset. It is also found that geometric accuracy metrics are not a perfect representation of expert ratings.
FRONTIERS IN ONCOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yoshiyuki Fukugawa, Ryo Toya, Tomohiko Matsuyama, Takahiro Watakabe, Yoshinobu Shimohigashi, Yudai Kai, Tadashi Matsumoto, Natsuo Oya
Summary: The study evaluated the impact of the metal artifact reduction (MAR) algorithm on the quality of computed tomography (CT) simulator images for patients with tonsillar cancer (TC) and found that the MAR algorithm could reduce interobserver variations in delineating tumor volume.
BMC MEDICAL IMAGING
(2022)
Article
Oncology
Johannes Kraft, Paul Lutyj, Felix Grabenbauer, Serge-Peer Stroehle, Joerg Tamihardja, Gary Razinskas, Stefan Weick, Anne Richter, Henner Huflage, Andrea Wittig, Michael Flentje, Dominik Lisowski
Summary: This study evaluates the image quality of brain metastases on dual-energy computed tomography (DECT) virtual monoenergetic imaging (VMI) and its impact on target volume delineation. The results show that VMI provides superior image contrast, lesion demarcation, and target volume delineation compared to conventional CT. This has significant implications for stereotactic radiotherapy planning.
RADIOTHERAPY AND ONCOLOGY
(2023)
Review
Oncology
Jingjing Shen, Peihua Gu, Yun Wang, Zhongming Wang
Summary: Postoperative adjuvant radiotherapy is important for breast cancer patients, and the accuracy of target volume and organ delineation significantly affects the treatment effect. Automatic delineation software based on an atlas and deep learning has been developed to reduce workload, establish a uniform delineation standard, and reduce inter-observer and intra-observer differences.
Article
Oncology
Yin Gao, Chenyang Shen, Xun Jia, Yang Kyun Park
Summary: This study describes the clinical implementation and evaluation of a virtual treatment planner (VTP) based on an artificial intelligence robot. Using deep reinforcement learning and human knowledge guidance, the VTP can autonomously adjust relevant parameters in treatment plan optimization to generate high-quality plans for prostate cancer stereotactic body radiation therapy. The evaluation results showed that the performance of the VTP is comparable to human-generated plans.
RADIOTHERAPY AND ONCOLOGY
(2023)
Article
Oncology
Meng Jin, Xia Liu, Jiabin Ma, Xiansong Sun, Hongnan Zhen, Jing Shen, Zhikai Liu, Xin Lian, Zheng Miao, Ke Hu, Xiaorong Hou, Fuquan Zhang
Summary: This study evaluated the impact of MR and CT simulation on defining the postoperative tumor bed in breast conserving radiotherapy patients without surgical clips. The results showed that MR can improve the visualization of changes in the tumor bed compared to CT, providing a more precise definition and enhancing consistency in tumor bed contouring.
CANCER MANAGEMENT AND RESEARCH
(2021)
Article
Oncology
Stephanie E. Combs, Brigitta G. Baumert, Martin Bendszus, Alessandro Bozzao, Michael Brada, Laura Fariselli, Alba Fiorentino, Ute Ganswindt, Anca L. Grosu, Frank L. Lagerwaard, Maximilian Niyazi, Tufve Nyholm, Ian Paddick, Damien Charles Weber, Claus Belka, Giuseppe Minniti
Summary: This study established a joint guideline for target volume definition of skull base tumors, emphasizing the use of high-precision and tissue-contrast MRI images to improve target delineation accuracy. Experts agreed that radiation techniques, imaging techniques, and technical aspects specific to different tumor types are crucial factors affecting target delineation.
RADIOTHERAPY AND ONCOLOGY
(2021)
Article
Oncology
Cas Stefaan Dejonckheere, Anja Thelen, Birgit Simon, Susanne Greschus, Muemtaz Ali Koeksal, Leonard Christopher Schmeel, Timo Wilhelm-Buchstab, Christina Leitzen
Summary: Malignant brain tumours often require surgery and radiation treatment. Comparing magnetic resonance images (MRIs) before and after surgery, there were notable differences in the tumour cavity, affected brain tissue, midline position, and bleeding in the surgical area. Therefore, a second MRI before radiation treatment is necessary to accurately plan and target the tumour. Optimized treatment planning based on current anatomy can improve outcomes in patients with high-grade glioma undergoing radiation therapy.
Article
Oncology
Yang-Zi Zhang, Xiang-Gao Zhu, Ma-Xiaowei Song, Kai-Ning Yao, Shuai Li, Jian-Hao Geng, Hong-Zhi Wang, Yong-Heng Li, Yong Cai, Wei-Hu Wang
Summary: Education program can improve the accuracy and consistency of CTV delineation for rectal cancer, and reduce the interobserver variation.
WORLD JOURNAL OF GASTROINTESTINAL ONCOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Shanghai Peng, Cui Yang, Hongbo Guo, Lijun Shen, Min Zhang, Jiazhou Wang, Zhen Zhang, Bin Cai, Weigang Hu
Summary: The purpose of this study was to establish and evaluate a (quasi) real-time automated treatment planning strategy for rectal cancer radiotherapy. The approach utilized a one-step full 3D fluence map prediction model based on a nonorthogonal convolution operation. The results showed that the RTTP method improved planning efficiency and deliverability performance while maintaining a plan quality close to that of the manual plans.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Oncology
Seyed Ali Jalalifar, Hany Soliman, Arjun Sahgal, Ali Sadeghi-Naini
Summary: This study investigated the impact of using less accurate but automatically generated tumor outlines on the efficacy of imaging biomarkers for predicting the response of brain metastasis to radiotherapy. The findings suggest that the imaging biomarkers and prediction models are resilient to imperfections in the automatically generated tumor masks.
Article
Anesthesiology
Erwan L'Her, Souha Nazir, Victoire Pateau, Dimitris Visvikis
Summary: Monitoring tidal volume using time-of-flight camera shows high correlation with reference values during mechanical ventilation, and is also feasible for respiratory monitoring under high-flow nasal canula for patients.
JOURNAL OF CLINICAL MONITORING AND COMPUTING
(2022)
Article
Medicine, General & Internal
Shima Sepehri, Olena Tankyevych, Taman Upadhaya, Dimitris Visvikis, Mathieu Hatt, Catherine Cheze Le Rest
Summary: The study evaluated the potential benefit of combining different algorithms into an improved consensus for final prediction in the context of radiomics, showing that this consensus can significantly enhance performance, especially in predicting outcomes using radiomic features.
Article
Radiology, Nuclear Medicine & Medical Imaging
Esteban Lucas Solari, Andrei Gafita, Sylvia Schachoff, Borjana Bogdanovi, Alberto Villagran Asiares, Thomas Amiel, Wang Hui, Isabel Rauscher, Dimitris Visvikis, Tobias Maurer, Kristina Schwamborn, Mona Mustafa, Wolfgang Weber, Nassir Navab, Matthias Eiber, Mathieu Hatt, Stephan G. Nekolla
Summary: The study evaluated the performance of combined PET and mpMRI radiomics for predicting postsurgical Gleason scores in primary prostate cancer patients. The results showed that the PET + ADC double-modality model performed the best, significantly outperforming other models, and had better predictive ability for psGS compared to biopsy GS.
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Bogdan Badic, Ronrick Da-Ano, Karine Poirot, Vincent Jaouen, Benoit Magnin, Johan Gagniere, Denis Pezet, Mathieu Hatt, Dimitris Visvikis
Summary: This study evaluated the value of contrast-enhanced diagnostic CT scans characterized through radiomics as predictors of recurrence for patients with stage II and III colorectal cancer in two different French University Hospitals. The results showed improved predictive performance after harmonization using a statistical method. Combining clinical variables and radiomics shape descriptors could effectively predict disease-free survival.
EUROPEAN RADIOLOGY
(2022)
Article
Oncology
S. Sellami, V Bourbonne, M. Hatt, F. Tixier, D. Bouzid, F. Lucia, O. Pradier, G. Goasduff, D. Visvikis, U. Schick
Summary: This study aimed to predict progression to radiotherapy using CBCT, and identified the predictive radiomic feature Coarseness. It was found to improve clinical-based prediction models.
Article
Computer Science, Artificial Intelligence
Valentin Oreiller, Vincent Andrearczyk, Mario Jreige, Sarah Boughdad, Hesham Elhalawani, Joel Castelli, Martin Vallieres, Simeng Zhu, Juanying Xie, Ying Peng, Andrei Iantsen, Mathieu Hatt, Yading Yuan, Jun Ma, Xiaoping Yang, Chinmay Rao, Suraj Pai, Kanchan Ghimire, Xue Feng, Mohamed A. Naser, Clifton D. Fuller, Fereshteh Yousefirizi, Arman Rahmim, Huai Chen, Lisheng Wang, John O. Prior, Adrien Depeursinge
Summary: This paper presents the post-analysis of the first edition of the HEad and neCK TumOR (HECKTOR) challenge, which focused on automatic segmentation of head and neck tumors in combined FDG-PET and CT images. A total of 64 teams participated in the challenge, and the best method achieved a high Dice Score Coefficient (DSC) and outperformed human inter-observer agreement and other methods.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
M. Hatt, A. K. Krizsan, A. Rahmim, T. J. Bradshaw, P. F. Costa, A. Forgacs, R. Seifert, A. Zwanenburg, I El Naqa, P. E. Kinahan, F. Tixier, A. K. Jha, D. Visvikis
Summary: This guideline provides best practices for radiomics analyses, including study design, data collection, and feature standardization. Although radiomics is a rapidly evolving field, this guideline primarily focuses on hand-crafted methods.
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Francois Lucia, Thomas Louis, Francois Cousin, Vincent Bourbonne, Dimitris Visvikis, Carole Mievis, Nicolas Jansen, Bernard Duysinx, Romain Le Pennec, Malik Nebbache, Martin Rehn, Mohamed Hamya, Margaux Geier, Pierre-Yves Salaun, Ulrike Schick, Mathieu Hatt, Philippe Coucke, Roland Hustinx, Pierre Lovinfosse
Summary: The purpose of this study was to develop machine learning models to predict regional and/or distant recurrence in patients with early-stage non-small cell lung cancer (ES-NSCLC) after stereotactic body radiation therapy (SBRT) using [F-18]FDG PET/CT and CT radiomics combined with clinical and dosimetric parameters.
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2023)
Meeting Abstract
Oncology
V. Bourbonne, M. Morjani, F. Lucia, M. Hatt, V. Jaouen, S. Querellou, D. Visvikis, O. Pradier, U. Schick
INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS
(2023)
Meeting Abstract
Radiology, Nuclear Medicine & Medical Imaging
N. Abdallah, J. Marion, C. Tauber, T. Carlier, P. Chauvet, M. Hatt
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2023)
Meeting Abstract
Radiology, Nuclear Medicine & Medical Imaging
N. Abdallah, J. Marion, C. Tauber, T. Carlier, P. Chauvet, M. Hatt
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2023)
Meeting Abstract
Radiology, Nuclear Medicine & Medical Imaging
A. Iantsen, P. Lovinfosse, M. Ferreira, A. Jadoul, N. Withofs, C. Derwael, A. Frix, J. Guiot, D. Visvikis, M. Hatt, R. Hustinx
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2021)