Article
Health Care Sciences & Services
Te-Chun Hsieh, Chiung-Wei Liao, Yung-Chi Lai, Kin-Man Law, Pak-Ki Chan, Chia-Hung Kao
Summary: This study evaluated deep learning techniques to improve the efficacy of bone metastasis detection on bone scans, retrospectively examining 19,041 patients. The overall performance of all models improved with contrastive learning. The high negative predictive value (NPV) of the optimal model may help physicians safely exclude bone metastases, decreasing physician workload, and improving patient care.
JOURNAL OF PERSONALIZED MEDICINE
(2021)
Article
Oncology
Ilyass Moummad, Cyril Jaudet, Alexis Lechervy, Samuel Valable, Charlotte Raboutet, Zamila Soilihi, Juliette Thariat, Nadia Falzone, Joelle Lacroix, Alain Batalla, Aurelien Corroyer-Dulmont
Summary: Due to long waiting times for MRI, algorithms such as resampling and denoising models are developed to speed up image acquisition time. However, the impact of these algorithms on radiomics has been poorly studied. This study aimed to develop resampling and denoising DL models and evaluate their impact on radiomics. The findings showed that DL models were able to reconstruct low resolution and noised MRI images quickly into high quality images, restoring radiomic features.
Article
Oncology
Hishan Tharmaseelan, Abhinay K. Vellala, Alexander Hertel, Fabian Tollens, Lukas T. Rotkopf, Johann Rink, Piotr Woznicki, Isabelle Ayx, Soenke Bartling, Dominik Noerenberg, Stefan O. Schoenberg, Matthias F. Froelich
Summary: This study demonstrates the performance of radiomics and CNN-based classifiers in determining the primary origin of visually indistinguishable gastrointestinal liver metastases. The results show that the radiomics-based K-nearest neighbor classifier performed well on an independent test set, while the image-based DenseNet-121 classifier also achieved good accuracy.
Article
Multidisciplinary Sciences
Qiang Lin, Tongtong Li, Chuangui Cao, Yongchun Cao, Zhengxing Man, Haijun Wang
Summary: This paper presents deep classifiers based on deep networks for reliably classifying SPECT bone images in automated diagnosis of metastasis. The classifiers, including VGG, ResNet and DenseNet, perform well in identifying bone metastasis with SPECT imaging, achieving high accuracy, precision, recall, specificity, F-1 score and AUC.
SCIENTIFIC REPORTS
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jingwei Wei, Jin Cheng, Dongsheng Gu, Fan Chai, Nan Hong, Yi Wang, Jie Tian
Summary: This study developed and validated a deep learning-based radiomics model for predicting chemotherapy response in colorectal liver metastases. Results showed that the DL model outperformed the traditional radiomics model in both training and validation with AUC of 0.903 and 0.820, respectively. Combining the DL model with CEA levels slightly improved predictive performance.
Review
Computer Science, Artificial Intelligence
Sergei Bedrikovetski, Nagendra N. Dudi-Venkata, Gabriel Maicas, Hidde M. Kroon, Warren Seow, Gustavo Carneiro, James W. Moore, Tarik Sammour
Summary: The review assessed the diagnostic performance of deep learning algorithms and radiomics models for lymph node metastases in abdominopelvic malignancies. Results showed that radiomics models improved diagnostic accuracy compared to radiologist's assessment, while deep learning models may further enhance this, although data remains limited.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2021)
Review
Oncology
Iromi R. Paranavithana, David Stirling, Montserrat Ros, Matthew Field
Summary: With recent progress in radiation therapy, precise delineation of metastatic lesions is crucial for the curative treatment of bone metastases. However, automatic segmentation algorithms for tumor segmentation still need improvement due to limited contrast and the lack of a widely agreed gold standard. Many proposed methods also lack evidence for their suitability in clinical practice.
Article
Radiology, Nuclear Medicine & Medical Imaging
Lulu Peng, Zehong Yang, Jue Liu, Yi Liu, Jianwei Huang, Junwei Chen, Yun Su, Xiang Zhang, Ting Song
Summary: This study aimed to explore whether deep learning radiomics (DLR) from MRI can be used to identify pregnancies with placenta accreta spectrum (PAS). The analysis of the training and validation datasets showed that the MRI-based DLR model had better performance in diagnosing PAS compared to the clinical model and MRI morphologic model.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Engineering, Biomedical
Yang Chen, Zhenyu Yang, Jingtong Zhao, Justus Adamson, Yang Sheng, Fang-Fang Yin, Chunhao Wang
Summary: In this study, a deep ensemble learning (DEL) model with radiomics spatial encoding execution was developed for improved glioma segmentation accuracy using multi-parametric magnetic resonance imaging (mp-MRI). The results showed that the DEL model outperformed the mp-MRI-only model, indicating the significance of the adopted radiomics spatial encoding execution in medical image segmentation.
PHYSICS IN MEDICINE AND BIOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Qiang Lin, Runxia Gao, Mingyang Luo, Haijun Wang, Yongchun Cao, Zhengxing Man, Rong Wang
Summary: This study proposes a semi-supervised segmentation model for automated identification and delineation of skeletal metastasis lesions in bone scan images, aiding in clinical diagnosis.
FRONTIERS IN MOLECULAR BIOSCIENCES
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Chia-Ying Lin, Shu-Mei Guo, Jenn-Jier James Lien, Wen-Tsen Lin, Yi-Sheng Liu, Chao-Han Lai, I-Lin Hsu, Chao-Chun Chang, Yau-Lin Tseng
Summary: This study developed a combined model that integrated deep learning, radiomics, and clinical data to classify lung nodules into different categories and pathological subtypes, as well as assign Lung-RADS scores.
Article
Computer Science, Information Systems
Ahmad Chaddad, Paul Sargos, Christian Desrosiers
Summary: The study introduces a novel approach in radiomics, demonstrating that the GMM-CNN features with an RF classifier can significantly enhance the accuracy of prognostic assessment for PDAC patients before surgery.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yung-Chieh Chen, Yi-Tien Li, Po-Chih Kuo, Sho-Jen Cheng, Yi-Hsiang Chung, Duen-Pang Kuo, Cheng-Yu Chen
Summary: This study developed a diagnostic tool combining machine learning segmentation and radiomic texture analysis for bone density screening using chest low-dose computed tomography (LDCT). The tool showed high accuracy in classifying abnormal bone density and distinguishing osteoporosis. The combination of ML segmentation and RTA is a feasible approach for osteoporosis screening during lung cancer screening.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Aydin Demircioglu
Summary: This study investigated the influence of choices on the predictive performance of radiomic models. The results showed that no single choice consistently led to the best-performing models, and the optimal performance required optimization during cross-validation.
INSIGHTS INTO IMAGING
(2022)
Article
Oncology
Naoya Ishibashi, Toshiya Maebayashi, Yuki Kimura, Masahiro Okada
Summary: This study investigated the relationship between pre-radiation therapy bone scan index (BSI) and overall survival (OS) in patients with bone metastases from cancers other than breast or prostate cancer. The study found no significant association between BSI and OS, but identified age, type of bone metastases, and type of cancer as significant factors affecting OS.
JOURNAL OF CANCER RESEARCH AND THERAPEUTICS
(2022)
Review
Radiology, Nuclear Medicine & Medical Imaging
Alexey Surov, Maciej Pech, Alexander Eckert, Christoph Arens, Oliver Grosser, Andreas Wienke
Summary: Correlations between maximum standard uptake values (SUVmax) derived from 18F-FDG PET and histopathological features in HNSCC were investigated, with the results showing that SUVmax cannot reflect relevant histopathological features in HNSCC.
Article
Radiology, Nuclear Medicine & Medical Imaging
Dennis Kupitz, Eric Einspaenner, Heiko Wissel, Alexander Hohn, Michael C. Kreissl, Oliver S. Grosser
Summary: This study evaluates the feasibility of using standard measurement hardware for the measurement of [Lu-177m]Lu and [Lu-177]Lu at equilibrium. Device-specific detection limits for [Lu-177m]Lu were determined according to international standards and validated.
Review
Biochemistry & Molecular Biology
Linnea Hojer Wang, Markus Wehland, Petra M. M. Wise, Manfred Infanger, Daniela Grimm, Michael C. C. Kreissl
Summary: This manuscript investigates four tyrosine kinase inhibitors (TKIs), cabozantinib, vandetanib, pralsetinib, and selpercatinib, used for treating advanced and/or metastatic medullary thyroid cancer (MTC). The focus is on treatment-related hypertension, a well-known adverse effect (AE) of these TKIs. While TKI-induced hypertension is rarely a dose-limiting side effect, complications associated with hypertension can increase with longer patient survival without proper medication.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2023)
Article
Oncology
Amina Banda, Bastiaan M. Prive, Youssra Allach, Maike J. M. Uijen, Steffie M. B. Peters, Cato C. Loeff, Martin Gotthardt, Constantijn H. J. Muselaers, J. Alfred Witjes, Inge M. van Oort, J. P. Michiel Sedelaar, Harm Westdorp, Niven Mehra, Fadi Khreish, Samer Ezziddin, Amir Sabet, Michael C. Kreissl, Thomas Winkens, Philipp Seifert, Marcel J. R. Janssen, Willemijn A. M. van Gemert, James Nagarajah
Summary: Prostate-specific membrane antigen-directed radioligand therapy (PSMA-RLT) is a novel treatment option for castration-resistant prostate cancer. This study investigated the safety and efficacy of two PSMA-RLT schemes in early-stage metastatic prostate cancer patients. The treatment was found to be safe with limited side effects and showed promising efficacy.
Article
Otorhinolaryngology
Friederike Eilsberger, Michael C. Kreissl, Markus Luster, Andreas Pfestroff
Summary: Theranostics via the sodium iodide symporter (NIS) is an important option in differentiated thyroid carcinoma due to its similarity in uptake and kinetics for diagnostic and therapeutic nuclides. However, radioiodine refractory thyroid carcinomas (RRTC) lack NIS expression, necessitating the exploration of alternative theranostic targets such as somatostatin receptors (SSTR) or the prostate-specific membrane antigen (PSMA). The potential success of these approaches is yet to be fully evaluated.
LARYNGO-RHINO-OTOLOGIE
(2023)
Article
Medicine, General & Internal
Ken Kudura, Nando Ritz, Arnoud J. Templeton, Tim Kutzker, Robert Foerster, Kwadwo Antwi, Michael C. Kreissl, Martin H. K. Hoffmann
Summary: The predictive value of total metabolic tumor burden prior to treatment was assessed in advanced non-small-cell lung cancer patients receiving immune checkpoint inhibitors. The results showed that the total metabolic tumor burden had negligible impact on overall survival, progression-free survival, and clinical benefit, while the morphological and metabolic properties of the primary tumor showed strong predictive power.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Oncology
Ken Kudura, Nando Ritz, Arnoud J. Templeton, Marc Kissling, Tim Kutzker, Robert Foerster, Martin H. K. Hoffmann, Kwadwo Antwi, Michael C. Kreissl
Summary: The frequency of additional primary malignancies detected incidentally on FDG-PET/CT at staging in NSCLC patients was assessed, as well as their impact on patient management and survival. FDG-PET/CT performed for staging can be a valuable tool to identify additional primary tumors in NSCLC patients. The early detection of additional primary tumors and interdisciplinary patient management may lead to similar survival rates as patients with NSCLC only.
Review
Oncology
Rossella Elisei, Enrique Grande, Michael C. Kreissl, Sophie Leboulleux, Tarun Puri, Nicolas Fasnacht, Jaume Capdevila
Summary: The incidence of thyroid cancer is increasing globally, with Europe ranking second in disease burden after Asia. Over the past few decades, research has identified key molecular pathways and targetable kinases/drivers for each histologic subtype of thyroid cancer. Newer and selective RET inhibitors, such as selpercatinib and pralsetinib, have shown promising efficacy and favorable toxicity profiles in clinical trials for the treatment of RET-driven advanced thyroid cancer. Genetic testing is crucial to identify RET alterations and determine the optimal treatment approach.
FRONTIERS IN ONCOLOGY
(2023)
Article
Medicine, General & Internal
Philipp Genseke, Christoph Ferdinand Wielenberg, Jens Schreiber, Eva Luecke, Steffen Frese, Thorsten Walles, Michael Christoph Kreissl
Summary: This prospective study used PET/CT quantitative data to differentiate between benign and malignant thoracic lymph nodes in lung cancer patients. They established cut-off values and a scoring system for this purpose, which can improve diagnostic accuracy and reliability.
Article
Biochemistry & Molecular Biology
Friederike Eilsberger, Michael C. Kreissl, Christoph Reiners, Adrien Holzgreve, Markus Luster, Andreas Pfestroff
Summary: The ATA risk classification can be applied to a German population to estimate the response to RAI therapy in the short term. Changing the age cutoff value from 55 to 50 years does not significantly impact the treatment response.
Article
Geriatrics & Gerontology
Yeo-Jin Yi, Falk Luesebrink, Mareike Ludwig, Anne Maass, Gabriel Ziegler, Renat Yakupov, Michael C. Kreissl, Matthew Betts, Oliver Speck, Emrah Duezel, Dorothea Haemmerer
Summary: By using a new analysis pipeline, researchers have improved the spatial precision in the brainstem area, allowing for investigations of the structure and function of the locus coeruleus at the group level.
NEUROBIOLOGY OF AGING
(2023)
Article
Oncology
Akram Al-Ibraheem, Ula Al-Rasheed, Noor Mashhadani, Ahmed Saad Abdlkadir, Dhuha Ali Al-Adhami, Saad Ruzzeh, Feras Istatieh, Areen Mansour, Basem Hamdan, Reem Kheetan, Marwa Al-Shatti, Issa Mohamad, Malik E. Juweid, Areej Abu Sheikha, Kamal Al-Rabi, Gerasimos P. Sykiotis, Michael C. Kreissl, Taleb Ismael, Iyad Sultan, Hikmat Abdel-Razeq
Summary: This study examines the long-term outcomes of differentiated thyroid cancer in the Arab population and identifies influential factors. The analysis of 528 cases shows favorable survival outcomes, with age, gender, risk categorization, and tumor stage impacting disease progression and mortality. RAI treatment and adherence to clinical practice guidelines result in better disease control.
Review
Oncology
Petra Petranovic Ovcaricek, Alfredo Campenni, Bart de Keizer, Desiree Deandreis, Michael C. Kreissl, Alexis Vrachimis, Murat Tuncel, Luca Giovanella
Summary: Radioiodine therapy is the main treatment for metastatic differentiated thyroid cancer, but only half of the patients achieve remission or have stable disease during follow-up. Novel tracers like PSMA ligands and FAPI have potential for diagnosing and treating radioiodine-refractory disease. This review discusses the role of these radiopharmaceuticals in managing radioiodine-refractory disease.
Article
Medicine, General & Internal
Elmer Jeto Gomes Ataide, Mathews S. Jabaraj, Simone Schenke, Manuela Petersen, Sarvar Haghghi, Jan Wuestemann, Alfredo Illanes, Michael Friebe, Michael C. Kreissl
Summary: This study assessed the potential of a Decision Support System (DSS) in reducing observer variability in thyroid nodule detection and region estimation. The results showed notable disparities between physician evaluations and the DSS assessments, highlighting the need for supplementary decision-making tools.
Review
Radiology, Nuclear Medicine & Medical Imaging
Friederike Eilsberger, Michael C. Kreissl, Markus Luster, Andreas Pfestroff
Summary: Theranostics via the sodium iodide symporter (NIS) is a unique option in differentiated thyroid carcinoma, with NIS being the most important theranostic target. Radioiodine refractory thyroid carcinomas (RRTC) have reduced or absent NIS expression, prompting the need to explore new theranostic targets.
NUKLEARMEDIZIN-NUCLEAR MEDICINE
(2022)