Article
Chemistry, Analytical
Eero Lehtonen, Jarmo Teuho, Juho Koskinen, Mojtaba Jafari Tadi, Riku Klen, Reetta Siekkinen, Joaquin Rives Gambin, Tuija Vasankari, Antti Saraste
Summary: A novel method for estimating respiratory motion using IMUs based on MEMS technology is presented, showing high accuracy in detecting respiratory cycles and estimating their lengths. Compared to clinically used techniques, the method achieved lower breathing rate and amplitude errors, demonstrating potential for simplifying logistics in PET imaging studies and enabling multi-position motion measurements.
Article
Optics
Pei-Ju Chiang, Chang-Hao Lin
Summary: This study proposes an active stereo vision system to improve the spatial resolution of scanning large objects. By projecting stripe patterns onto the object surface with different rotation angles, the system can accurately and completely detect matching points. The system also uses a two-step method to remove incorrect matching points and improve reconstruction results. Experimental results show that this system provides a low-cost, convenient, and effective approach for high spatial resolution 3D object reconstruction.
OPTICS AND LASERS IN ENGINEERING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Freba Grawe, Franziska Blom, Michael Winkelmann, Caroline Burgard, Christine Schmid-Tannwald, Lena M. Unterrainer, Gabriel T. Sheikh, Paulo L. Pfitzinger, Philipp Kazmierczak, Clemens C. Cyran, Jens Ricke, Christian G. Stief, Peter Bartenstein, Johannes Ruebenthaler, Matthias P. Fabritius, Thomas Geyer
Summary: PSMA-RADS 1.0 is a reliable method for assessing PSMA-PET/CT with strong consistency and agreement among readers. It shows great potential for establishing a standard approach to diagnosing and planning treatment for prostate cancer patients, and can be used confidently even by readers with less experience.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Freba Grawe, Ricarda Ebner, Thomas Geyer, Leonie Beyer, Michael Winkelmann, Gabriel T. Sheikh, Ralf Eschbach, Christine Schmid-Tannwald, Clemens C. Cyran, Jens Ricke, Peter Bartenstein, Maurice M. Heimer, Lorenzo Faggioni, Christine Spitzweg, Matthias P. Fabritius, Christoph J. Auernhammer, Johannes Ruebenthaler
Summary: This study aimed to determine the intra- and interreader agreement of SSTR-RADS 1.0. The results showed high consistency among readers in assessing target lesions using SSTR-RADS 1.0, indicating that the system can effectively standardize the diagnosis and treatment planning in NET patients.
EUROPEAN RADIOLOGY
(2023)
Article
Computer Science, Software Engineering
Byeongjoo Ahn, Ioannis Gkioulekas, Aswin C. Sankaranarayanan
Summary: The combination of structured light system and kaleidoscope enables precise and full surround 3D scanning of complex shapes, with accurate determination of projector and camera pixel labels. This system can serve as a multi-view structured light system with hundreds of virtual projectors and cameras, demonstrating advantages in scanning objects with a wide range of shapes and reflectances.
ACM TRANSACTIONS ON GRAPHICS
(2021)
Article
Engineering, Electrical & Electronic
Yeqi Hu, Wei Rao, Lin Qi, Junyu Dong, Jinzhen Cai, Hao Fan
Summary: In this article, a novel optical 3-D measurement system is proposed for objects immersed in a glass water tank. By using structured light projection and refractive stereo triangulation method, the proposed system can effectively reduce measurement errors and work well under different turbidity conditions.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Automation & Control Systems
Shengyang Chen, Yurong Feng, Chih-Yung Wen, Yajing Zou, Wu Chen
Summary: In this article, a stereo visual inertial pose estimation method based on feedforward and feedbacks is presented. The proposed method achieves fast processing by storing only the most recent pose and measurements. It introduces gradient decreased feedback, roll-pitch feedforward, and bias estimation feedback to fuse vision and inertial measurements. The system, called FVIS, demonstrates high accuracy and robustness compared to existing visual inertial SLAM approaches. FVIS has been implemented and tested on a UAV platform and the source code is publicly available.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Engineering, Civil
Hamed Haghighi, Mehrdad Dianati, Valentina Donzella, Kurt Debattista
Summary: Camera image simulation is crucial for virtual validation of autonomous vehicles and robots, as well as for creating image datasets for training vision models. To address the computational complexity, we propose a technique based on Stereo Super Resolution (SSR) to speed up the simulation of stereo images, achieving promising results in terms of speed and performance.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Tiantian Li, Mengxi Zhang, Wenyuan Qi, Evren Asma, Jinyi Qi
Summary: Respiratory motion is a significant factor affecting the quality of PET imaging. In this study, a robust joint estimation method combining deep learning with image registration was proposed. The method effectively estimated the emission image and patient motion from respiratory gated data, outperforming traditional registration-based methods. The proposed DL-ADMM algorithm showed promising results in both simulated and real data studies, reducing bias and improving image quality.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Tasmia Rahman Tumpa, Shelley N. Acuff, Jens Gregor, Sanghyeb Lee, Dongming Hu, Dustin R. Osborne
Summary: A new TOF-PEPT algorithm is introduced in this paper for specific respiratory motion estimation in PET/CT imaging. Results show high correlation between motion signals derived from TOF-PEPT and Anzai band, outperforming COM methods. Gated imaging based on TOF-PEPT improves image quality, with higher max SUVs and sharper images observed in clinical studies compared to Anzai and COM methods.
Article
Chemistry, Analytical
Viprav B. Raju, Edward Sazonov
Summary: This paper proposes a novel device called FOODCAM and an associated methodology for estimating food portion sizes without the need for dimensional reference. The device was able to accurately estimate food portion sizes and has the potential for application in diet and eating behavior studies.
Article
Engineering, Multidisciplinary
Yue Wang, Lei Zhou, Zihao Li, Jun Wang, Xuangou Wu, Xiangjun Wang, Lei Hu
Summary: This paper presents a 3-D reconstruction method for dynamic stereo vision of metal surface based on line structured light, overcoming the limitation of the measurement range of static stereo vision. The proposed method uses joint calibration and global optimization to accurately reconstruct the 3-D coordinates of the line structured light fringe, improving the reconstruction accuracy.
Article
Ophthalmology
Yiya Chen, Zhimo Yao, Zhifen He, Ziyun Cheng, Pi-Chun Huang, Seung Hyun Min, Fan Lu, Robert F. Hess, Jiawei Zhou
Summary: In this study, the researchers developed a stereo task to examine the role of depth in disambiguating motion direction. They found clear speed tuning of stereo sensitivity, which was not significantly affected by different spatial frequencies. Additionally, they observed a significant difference in stereo sensitivity between static and laterally moving stimuli. This study suggests that lateral motion modulates human global depth perception.
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE
(2022)
Article
Computer Science, Interdisciplinary Applications
Dimitrios Psychogyios, Evangelos Mazomenos, Francisco Vasconcelos, Danail Stoyanov
Summary: This paper proposes a learning-based framework for jointly estimating disparity and binary tool segmentation masks. The experimental results show that supervising the segmentation task improves the network's disparity estimation accuracy. The domain adaptation scheme enables domain adaptation of the adjacent disparity task. The best overall multi-task model performs well on the test sets.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Chemistry, Organic
Aaron Schuell, Lisa Grothe, Eduardo Rodrigo, Thomas Erhard, Siegfried R. Waldvogel
Summary: A novel electrosynthetic approach for the synthesis of aryl dibenzothiophenium salts, including the metal-free, electrochemical formation of a C-S bond under ambient conditions, has been reported. The broad applicability of this method has been demonstrated with 14 examples, achieving isolated yields of up to 72%, including nitrogen-containing heterocycles. The resulting sulfonium salts can serve as precursors for the synthesis of [F-18]fluoroarenes found in PET tracer ligands.
Article
Radiology, Nuclear Medicine & Medical Imaging
Johan Berglund, Adam van Niekerk, Henric Ryden, Tim Sprenger, Enrico Avventi, Ola Norbeck, Stefan L. Glimberg, Oline V. Olesen, Stefan Skare
Summary: This study utilized an optical markerless tracking system for real-time rigid body motion correction during scanning, enabling motion-robust diffusion weighted imaging of the brain. Prospective motion correction with dynamic ghost correction allowed for high quality DWI images in the presence of fast and continuous motion within a 10 degrees range.
MAGNETIC RESONANCE IN MEDICINE
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jakob M. Slipsager, Stefan L. Glimberg, Liselotte Hojgaard, Rasmus R. Paulsen, Paul Wighton, M. Dylan Tisdall, Camilo Jaimes, Borjan A. Gagoski, P. Ellen Grant, Andre van Der Kouwe, Oline V. Olesen, Robert Frost
Summary: Comparing prospective motion correction (PMC) and retrospective motion correction (RMC) in Cartesian 3D-encoded MPRAGE scans showed that PMC resulted in superior image quality compared to RMC, with increasing correction frequency reducing motion artifacts in RMC.
MAGNETIC RESONANCE IN MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Josefine Vilsboll Sundgaard, James Harte, Peter Bray, Soren Laugesen, Yosuke Kamide, Chiemi Tanaka, Rasmus R. Paulsen, Anders Nymark Christensen
Summary: This study proposes an automatic diagnostic algorithm for detecting otitis media based on deep metric learning, comparing different distance-based metric loss functions. The results are comparable to the best clinical experts, paving the way for more accurate and operator-independent diagnosis of otitis media.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Physiology
Xabier Morales Ferez, Jordi Mill, Kristine Aavild Juhl, Cesar Acebes, Xavier Iriart, Benoit Legghe, Hubert Cochet, Ole De Backer, Rasmus R. Paulsen, Oscar Camara
Summary: This study successfully predicted the endothelial cell activation potential (ECAP), an index linked to thrombosis risk, solely based on patient-specific LAA morphology using a deep learning model. Evaluation of various DL approaches showed that the graph-based DL model outperformed others and demonstrated adaptability in a more realistic dataset.
FRONTIERS IN PHYSIOLOGY
(2021)
Article
Otorhinolaryngology
Josefine Vilsboll Sundgaard, Maria Varendh, Franziska Nordstrom, Yosuke Kamide, Chiemi Tanaka, James Harte, Rasmus R. Paulsen, Anders Nymark Christensen, Peter Bray, Soren Laugesen
Summary: This study investigates the inter-rater reliability and agreement of the diagnosis of otitis media with effusion, acute otitis media, and no effusion cases based on otoscopy images and wideband tympanometry measurements. The results show high inter-rater reliability for diagnosing acute otitis media, lower for diagnosing otitis media with effusion. Wideband tympanometry can provide valuable information for diagnosis, and the diagnostic certainty increases when examined together with otoscopy images.
INTERNATIONAL JOURNAL OF PEDIATRIC OTORHINOLARYNGOLOGY
(2022)
Article
Computer Science, Information Systems
Umaer Hanif, Eileen B. Leary, Logan D. Schneider, Rasmus R. Paulsen, Anne Marie Morse, Adam Blackman, Paula K. Schweitzer, Clete A. Kushida, Stanley Y. Liu, Poul Jennum, Helge B. D. Sorensen, Emmanuel J. M. Mignot
Summary: This study analyzed 3D craniofacial scans of 1366 patients with obstructive sleep apnea (OSA) to predict apnea-hypopnea index (AHI) using a machine learning algorithm, achieving a mean absolute error of 11.38 events/hour and 67% accuracy for predicting OSA. The algorithm has potential to serve as an inexpensive and efficient screening tool for individuals with suspected OSA.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Computer Science, Information Systems
Josefine Vilsboll Sundgaard, Peter Bray, Soren Laugesen, James Harte, Yosuke Kamide, Chiemi Tanaka, Anders Nymark Christensen, Rasmus R. Paulsen
Summary: In this study, an automatic diagnostic algorithm based on wideband tympanometry measurements is proposed for detecting otitis media. A convolutional neural network is developed for classification of otitis media using the analysis of the wideband tympanogram. Results show high performance in overall otitis media detection, although specific types of otitis media cannot be distinguished.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Computer Science, Software Engineering
T. V. Christiansen, J. A. Baerentzen, R. R. Paulsen, M. R. Hannemose
Summary: Neural implicit surfaces are effective for representing shapes with arbitrary topology, but open surfaces are still challenging. The generalized winding number (GWN) is a promising approach for distinguishing points on 3D shapes. However, it lacks information about the distance to the surface, which is necessary for tasks like ray tracing. To address this, we propose the semi-signed distance field (SSDF) representation, which combines the GWN and surface distance. We compare the GWN and SSDF for various applications and find that both are suitable for neural representation of open surfaces.
COMPUTER GRAPHICS FORUM
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Ana-Teodora Radutoiu, Francois Patou, Jan Margeta, Rasmus R. Paulsen, Paula Lopez Diez
Summary: We propose a novel method for automatic ROI extraction in full head CT scans to isolate the inner ear. By utilizing state-of-the-art communicative multi-agent reinforcement learning and specifically designed landmarks, we are able to robustly extract ROIs with the same orientation and relevant anatomy.
MACHINE LEARNING IN MEDICAL IMAGING, MLMI 2022
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Morten Rieger Hannemose, Josefine Vilsboll Sundgaard, Niels Kvorning Ternov, Rasmus R. Paulsen, Anders Nymark Christensen
Summary: In this paper, methods for estimating the difficulty of diagnosing medical image cases by doctors are introduced, using embeddings obtained with deep metric learning. A practical method for obtaining ground truth human difficulty is also presented. Experimental results on two medical datasets show that our methods outperform existing ones.
APPLICATIONS OF MEDICAL ARTIFICIAL INTELLIGENCE, AMAI 2022
(2022)
Proceedings Paper
Neuroimaging
Paula Lopez Diez, Kristine Sorensen, Josefine Vilsboll Sundgaard, Khassan Diab, Jan Margeta, Francois Patou, Rasmus R. Paulsen
Summary: This paper proposes a framework for inner ear abnormality detection based on a deep reinforcement learning model. By training on normative data only, the framework extracts two abnormality measurements: the variability of the predicted configuration of landmarks in a subspace formed by shape alignment and projection, and the distribution of predicted Q-values before landmark localization. The implementation shows outstanding performance on artificial and real clinical CT datasets, suggesting potential applications in solving other complex anomaly detection problems.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT III
(2022)
Proceedings Paper
Engineering, Biomedical
Kristine Sorensen, Oscar Camara, Ole de Backer, Klaus F. Kofoed, Rasmus R. Paulsen
Summary: Medical image segmentation involves labelling each pixel or voxel as inside or outside a given anatomy. The proposed NUDF method learns Unsigned Distance Field directly from the image, allowing for high-resolution processing with small memory requirements. The method accurately predicts 3D mesh models of the complex and variable LAA shape, capturing its details.
2022 IEEE INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (IEEE ISBI 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Kristine Aavild Juhl, Xabier Morales, Ole de Backer, Oscar Camara, Rasmus Reinhold Paulsen
Summary: This study focuses on surface representations of complex anatomies using a novel Implicit Neural Distance Representation based on unsigned distance fields. The optimized latent space contains important global shape information useful for reconstructing anatomies and successfully separating different anatomies. The representation is also shown to be effective in gender classification of human face geometries.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Paula Lopez Diez, Josefine Vilsboll Sundgaard, Francois Patou, Jan Margeta, Rasmus Reinhold Paulsen
Summary: This study introduces a pipeline for characterizing facial and cochlear nerves in CT scans, utilizing deep reinforcement learning and key landmark definition. The method accurately locates these nerves and provides reliable measurements for computer-aided diagnosis and surgery planning.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT IV
(2021)
Proceedings Paper
Engineering, Biomedical
Josefine Vilsboll Sundgaard, Kristine Aavild Juhl, Klaus Fuglsang Kofoed, Rasmus R. Paulsen
Summary: Automated segmentation of cardiac CT scans using 2D convolutional neural networks with multi-planar approach and spatial propagation addresses the limitations of 3D networks, ensuring spatial consistency and achieving promising results on challenging datasets.
MEDICAL IMAGING 2020: IMAGE PROCESSING
(2021)
