Article
Biology
Taimur Hassan, Muhammad Shafay, Bilal Hassan, Muhammad Usman Akram, Ayman ElBaz, Naoufel Werghi
Summary: A novel knowledge distillation-based instance segmentation scheme is proposed for segmenting and classifying prostate tissues effectively to aid in the timely treatment of PCa. Through testing on multiple datasets, the scheme demonstrates excellent performance in identifying prostate tissues and grading PCa, as well as achieving significant results in testing with expert pathologists.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Engineering, Electrical & Electronic
R. Karthik, R. Menaka, M. V. Siddharth, Sameeha Hussain, P. Siddharth, Daehan Won
Summary: This study proposes a novel lightweight convolutional neural network for accurate scoring of Gleason grades in prostate cancer. With multiple levels of attention mechanisms and a channel attention module, the network is able to focus on more significant features, thus improving the accuracy of scoring.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Oncology
Yali Qiu, Yujin Hu, Peiyao Kong, Hai Xie, Xiaoliu Zhang, Jiuwen Cao, Tianfu Wang, Baiying Lei
Summary: Prostate biopsy histopathology and immunohistochemistry are important for the differential diagnosis of diseases and assessing cancer differentiation. Growing demand for experienced uropathologists increases pressure, and fluctuating grades can lead to inadequate or excessive treatment. An AI system was developed to address these issues.
FRONTIERS IN ONCOLOGY
(2022)
Article
Urology & Nephrology
Ohad Kott, Drew Linsley, Ali Amin, Andreas Karagounis, Carleen Jeffers, Dragan Golijanin, Thomas Serre, Boris Gershman
Summary: A state-of-the-art deep learning algorithm was developed for the histopathologic diagnosis and Gleason grading of prostate biopsy specimens, achieving 91.5% accuracy in coarse classification and 85.4% accuracy in fine classification. The algorithm showed excellent performance with high sensitivity and specificity, though limitations include the small sample size and the need for external validation.
EUROPEAN UROLOGY FOCUS
(2021)
Article
Biology
Jose M. Marron-Esquivel, L. Duran-Lopez, A. Linares-Barranco, Juan P. Dominguez-Morales
Summary: Prostate cancer is a commonly diagnosed cancer in men, and although its mortality has decreased, it remains a leading cause of death. Current diagnosis relies on biopsy tests and the Gleason scale, but there is variability in assignment of scores among pathologists. Applying deep learning-based automatic diagnosis systems can help reduce this inter-observer variability and provide a second opinion for medical centers.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Urology & Nephrology
Felicia Marginean, Ida Arvidsson, Athanasios Simoulis, Niels Christian Overgaard, Kalle Astrom, Anders Heyden, Anders Bjartell, Agnieszka Krzyzanowska
Summary: The study aimed to develop an artificial intelligence algorithm for improved standardisation in Gleason grading in prostate cancer biopsies, using machine learning and convolutional neural networks. The algorithm showed high accuracy in detecting cancer areas and assigning Gleason patterns correctly, achieving similar results as pathologists with low intraobserver variability.
EUROPEAN UROLOGY FOCUS
(2021)
Article
Engineering, Biomedical
Xu Lu, Shulian Zhang, Zhiyong Liu, Shaopeng Liu, Jun Huang, Guoquan Kong, Mingzhu Li, Yinying Liang, Yunneng Cui, Chuan Yang, Shen Zhao
Summary: A novel Automatic Region-based Gleason Grading (ARGG) network for prostate cancer based on deep learning is proposed in this study. The experimental results show that the proposed grading model outperforms manual diagnosis by physicians on prostate ultrasound images.
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
(2022)
Review
Urology & Nephrology
Kimmo Kartasalo, Wouter Bulten, Brett Delahunt, Po-Hsuan Cameron Chen, Hans Pinckaers, Henrik Olsson, Xiaoyi Ji, Nita Mulliqi, Hemamali Samaratunga, Toyonori Tsuzuki, Johan Lindberg, Mattias Rantalainen, Carolina Wahlby, Geert Litjens, Pekka Ruusuvuori, Lars Egevad, Martin Eklund
Summary: Artificial intelligence has shown promise in cancer detection and Gleason grading, but more work is needed to ensure accuracy and applicability in diverse clinical settings.
EUROPEAN UROLOGY FOCUS
(2021)
Article
Biology
Santiago Toledo-Cortes, Diego H. Useche, Henning Muller, Fabio A. Gonzalez
Summary: This paper proposes a quantum-inspired deep probabilistic learning ordinal regression model for medical image diagnosis, which takes advantage of deep learning and the ordinal information of disease stages. Experimental results show that the method improves the diagnosis performance and interpretability of the results.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Biochemistry & Molecular Biology
Wouter Bulten, Kimmo Kartasalo, Po-Hsuan Cameron Chen, Peter Strom, Hans Pinckaers, Kunal Nagpal, Yuannan Cai, David F. Steiner, Hester van Boven, Robert Vink, Christina Hulsbergen-van de Kaa, Jeroen van der Laak, Mahul B. Amin, Andrew J. Evans, Theodorus van der Kwast, Robert Allan, Peter A. Humphrey, Henrik Gronberg, Hemamali Samaratunga, Brett Delahunt, Toyonori Tsuzuki, Tomi Hakkinen, Lars Egevad, Maggie Demkin, Sohier Dane, Fraser Tan, Masi Valkonen, Greg S. Corrado, Lily Peng, Craig H. Mermel, Pekka Ruusuvuori, Geert Litjens, Martin Eklund
Summary: The PANDA challenge is the largest histopathology competition to date, aiming to catalyze the development of reproducible AI algorithms for Gleason grading in prostate cancer. The submitted algorithms achieved pathologist-level performance and their diversity and generalization were validated through cross-continental cohorts.
Article
Oncology
Mizuho Nishio, Hidetoshi Matsuo, Yasuhisa Kurata, Osamu Sugiyama, Koji Fujimoto
Summary: This study aimed to develop and evaluate an automatic prediction system for grading histopathological images of prostate cancer using a deep learning model and label distribution learning (LDL). Results showed that LDL improved the diagnostic performance of the automatic prediction system for cancer grading.
Article
Biology
Lizhi Shao, Zhenyu Liu, Jiangang Liu, Ye Yan, Kai Sun, Xiangyu Liu, Jian Lu, Jie Tian
Summary: Magnetic resonance imaging (MRI) is the best imaging modality for non-invasive observation of prostate cancer. However, existing quantitative analysis methods face challenges in patient-level prediction. Therefore, a new method called GMINet is proposed to improve the accuracy and interpretability of MRI in prostate cancer diagnosis and treatment. The method achieves state-of-the-art performance and enables tumor detection through attention analysis.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Oncology
Kamal Hammouda, Fahmi Khalifa, Norah Saleh Alghamdi, Hanan Darwish, Ayman El-Baz
Summary: In this study, an automated diagnostic deep learning system for the Gleason system was developed to grade and classify prostate cancer. The system showed high accuracy in classification and recognition, and achieved agreement with the consensus grade groups.
Article
Engineering, Electrical & Electronic
Tojo Mathew, B. Ajith, Jyoti R. Kini, Jeny Rajan
Summary: Cancer grade is an important indicator for prognosis and treatment decisions in cancer. This paper proposes a new method to address the challenges of limited datasets and class imbalance in automated cancer grading. By combining datasets from different sources and applying color-normalization, the high training data requirement of deep neural networks is met. Class imbalance is addressed through context-preserving augmentation. A customized convolutional neural network classifier is used to classify candidate cells. The proposed method outperforms recent methods and offers adaptability to different datasets.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2022)
Article
Oncology
Leonard M. da Silva, Emilio M. Pereira, Paulo G. O. Salles, Ran Godrich, Rodrigo Ceballos, Jeremy D. Kunz, Adam Casson, Julian Viret, Sarat Chandarlapaty, Carlos Gil Ferreira, Bruno Ferrari, Brandon Rothrock, Patricia Raciti, Victor Reuter, Belma Dogdas, George DeMuth, Jillian Sue, Christopher Kanan, Leo Grady, Thomas J. Fuchs, Jorge S. Reis-Filho
Summary: AI-based system Paige Prostate shows high sensitivity and NPV in the diagnosis of prostate cancer, leading to potential improvement in patient care. It can accurately classify histopathology slides into benign or suspicious categories, reducing diagnostic time and improving efficiency.
JOURNAL OF PATHOLOGY
(2021)
Article
Robotics
Alaa Eldin Abdelaal, Jordan Liu, Nancy Hong, Gregory D. Hager, Septimiu E. Salcudean
Summary: This study investigates the benefits of leveraging robots' parallel capabilities in automating multilateral surgical tasks. By proposing and developing parallel execution models for different surgical subtasks, significant reductions in completion time and improved efficiency are achieved.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Computer Science, Artificial Intelligence
Davood Karimi, Simon K. Warfield, Ali Gholipour
Summary: This study critically evaluates the role of transfer learning in training fully convolutional networks for medical image segmentation. It highlights the importance of task and data dependency in improving segmentation accuracy, with observations on limited changes in convolutional filters during training and the potential for accurate FCNs by freezing the encoder section at random values. Additionally, the research challenges the common belief that the encoder section needs to learn data/task-specific representations, offering new insights and alternative training methods for FCNs.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Davood Karimi, Lana Vasung, Camilo Jaimes, Fedel Machado-Rivas, Shadab Khan, Simon K. Warfield, Ali Gholipour
Summary: A machine learning-based technique is proposed for accurately estimating the number and orientations of fascicles in a voxel, outperforming classical and machine learning methods in predicting crossing fascicles and leading to more accurate tractography. This method also shows better robustness to measurement down sampling and expert quality assessment of tractography results compared to other methods.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Automation & Control Systems
Amir Hossein Hadi Hosseinabadi, David Gregory Black, Septimiu E. Salcudean
Summary: The article introduces a novel hardware and software architecture for a smart optical force-torque sensor. The use of configurable, modular, and compact electronics leads to unparalleled performance characteristics, such as ultra-low noise and high resolution. Performance is achieved through oversampling of optical transducers and parallel hardware processing using FPGA, with additional features like an inertial measurement unit and temperature sensor integrated for improved compensation.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2021)
Article
Computer Science, Information Systems
Hongzhi Zhu, Yasmin Halwani, Robert Rohling, Sidney Fels, Septimiu Salcudean
Summary: This paper aims to study the HCI control logic used in ultrasound (US) machines through surveying sonographers. The research discovers that the control logic for the most frequently used functions in US machines is identical across different manufacturers. Using the unified modeling language (UML), the study visualizes and formulates control logic to optimize logical interactions and simplify suboptimal ones. This research provides insights into HCI approaches in US machines and establishes a UML-based framework for future machine design.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Computer Science, Information Systems
Yanan Shao, Hoda S. Hashemi, Paula Gordon, Linda Warren, Jane Wang, Robert Rohling, Septimiu Salcudean
Summary: In this study, a new processing pipeline for breast cancer classification using S-WAVE data was proposed and evaluated on a dataset of 40 patients. The best results outperformed the state-of-the-art reported S-WAVE breast cancer classification performance, and the sensitivity of the classification results to feature selection and changes in breast lesion contours was studied.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Review
Engineering, Electrical & Electronic
Septimiu E. Salcudean, Hamid Moradi, David G. Black, Nassir Navab
Summary: Robot-assisted medical imaging utilizes robots to acquire medical images, enabling precise and accurate trajectory control of the imaging system. This technology provides valuable information to physicians and facilitates the alignment of preoperative imaging with patients. This review article describes recent advancements in ultrasound, endoscopy, X-ray, optical coherence tomography, and nuclear medicine, and discusses research directions in autonomous scanning and physics-driven approaches for accurate placement and trajectory control of the imaging system.
PROCEEDINGS OF THE IEEE
(2022)
Article
Computer Science, Interdisciplinary Applications
Shahed Mohammed, Mohammad Honarvar, Qi Zeng, Hoda Hashemi, Robert Rohling, Piotr Kozlowski, Septimiu Salcudean
Summary: In this paper, we introduce two model-based iterative methods for obtaining shear modulus images of tissue using magnetic resonance elastography. The methods utilize sparsifying regularization, wave equation constraint, and have fast convergence and improved contrast to noise.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Robotics
Hyunwoo Song, Hamid Moradi, Baichuan Jiang, Keshuai Xu, Yixuan Wu, Russell H. H. Taylor, Anton Deguet, Jin U. U. Kang, Septimiu E. E. Salcudean, Emad M. M. Boctor
Summary: This study introduces an integrated real-time intraoperative surgical guidance system that co-registers an endoscope camera of the da Vinci surgical robot and a transrectal ultrasound (TRUS) transducer using photoacoustic markers. The system enables tracking of the laser spot illuminated by a pulsed-laser-diode attached to the surgical instrument, providing both fluorescence (FL) and photoacoustic (PA) images of the surgical region-of-interest (ROI). Quantitative evaluation shows that the average registration and tracking errors are 0.84 mm and 1.16°, respectively. This study demonstrates the effectiveness of co-registered photoacoustic marker tracking using TRUS+PA imaging for functional guidance of the surgical ROI.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Computer Science, Cybernetics
David Black, Yas Oloumi Yazdi, Amir Hossein Hadi Hosseinabadi, Septimiu Salcudean
Summary: Current teleultrasound methods trade off between precision and latency versus flexibility and cost. This paper presents a novel concept of human teleoperation, where an expert remotely operates a person wearing a mixed-reality headset by controlling a virtual ultrasound probe. The system shows promise with lower latency and measurement errors compared to audiovisual teleguidance.
HUMAN-COMPUTER INTERACTION
(2023)
Article
Acoustics
Hoda S. Hashemi, Shahed K. Mohammed, Qi Zeng, Reza Zahiri Azar, Robert N. Rohling, Septimiu E. Salcudean
Summary: Real-time ultrasound imaging is crucial in ultrasound-guided interventions, providing more spatial information compared to conventional 2-D frames. However, the long data acquisition time of 3-D imaging can be a bottleneck, reducing practicality and introducing artifacts. This article introduces a real-time volumetric acquisition method using a matrix array transducer, which addresses the issues and allows for accurate tissue elasticity estimation. The method has been validated on various phantoms and shows promising results. Rating: 9/10.
IEEE TRANSACTIONS ON ULTRASONICS FERROELECTRICS AND FREQUENCY CONTROL
(2023)
Article
Acoustics
Tajwar Abrar Aleef, Qi Zeng, Hamid Moradi, Shahed Mohammed, Tom Curran, Mohammad Honarvar, Robert Rohling, S. Sara Mahdavi, Septimiu E. Salcudean
Summary: This article presents a first-of-a-kind 3-D hand-operated endorectal shear wave absolute vibro-elastography system, which allows quantitative and volumetric assessment of tissue stiffness during systematic prostate biopsy. The system shows high correlations with 3-D magnetic resonance elastography and an existing 3-D shear wave absolute vibro-elastography system.
IEEE TRANSACTIONS ON ULTRASONICS FERROELECTRICS AND FREQUENCY CONTROL
(2023)
Article
Computer Science, Artificial Intelligence
Zhongliang Jiang, Septimiu E. Salcudean, Nassir Navab
Summary: Ultrasound (US) is widely used in clinical intervention and diagnosis due to its non-invasive and real-time imaging capabilities. However, free-hand US examinations are dependent on the operator's skills. Robotic US Systems (RUSS) aim to overcome this limitation by offering reproducibility and intelligent imaging. This paper categorizes RUSS as teleoperated or autonomous and reviews recent developments in both categories. Machine learning and artificial intelligence play a key role in enabling motion and deformation-aware robotic image acquisition. The recovery of the language of sonography through artificial intelligence research is valuable. The article provides a comprehensive understanding of RUSS and presents challenges in developing intelligent robotic sonographer colleagues.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Engineering, Biomedical
Tajwar Abrar Aleef, Qi Zeng, W. James Morris, S. Sara Mahdavi, Septimiu E. Salcudean
Summary: This study proposes a framework for registering trans-perineal template mapping biopsy (TTMB) cores to advanced volumetric ultrasound data, such as multi-parametric transrectal ultrasound (mpTRUS). The framework allows for quick localization of the cores and can be used to train cancer classifiers.
INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Tajwar Abrar Aleef, Ingrid T. Spadinger, Michael D. Peacock, Septimiu E. Salcudean, S. Sara Mahdavi
Summary: The study presents a method using conditional generative adversarial networks to generate consistent treatment plans for low-dose-rate prostate brachytherapy. By learning from a large pool of clinical data, the proposed method achieved comparable results to manual plans with significantly reduced planning time.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT IV
(2021)