1
|
Mu X, Ge Z, Lu D, Li T, Liu L, Chen C, Song S, Fu W, Jin G. Deep learning model using planar whole-body bone scintigraphy for diagnosis of skull base invasion in patients with nasopharyngeal carcinoma. J Cancer Res Clin Oncol 2024; 150:449. [PMID: 39379746 PMCID: PMC11461747 DOI: 10.1007/s00432-024-05969-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/22/2024] [Accepted: 09/23/2024] [Indexed: 10/10/2024]
Abstract
PURPOSE This study assesses the reliability of deep learning models based on planar whole-body bone scintigraphy for diagnosing Skull base invasion (SBI) in nasopharyngeal carcinoma (NPC) patients. METHODS In this multicenter study, a deep learning model was developed using data from one center with a 7:3 allocation to training and internal test sets, to diagnose SBI in patients newly diagnosed with NPC using planar whole-body bone scintigraphy. Patients were diagnosed based on a composite reference standard incorporating radiologic and follow-up data. Ten different convolutional neural network (CNN) models were applied to both whole-image and partial-image input modes to determine the optimal model for each analysis. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), calibration, decision curve analysis (DCA), and compared with expert assessments by two nuclear medicine physicians. RESULTS The best-performing model using partial-body input achieved AUCs of 0.80 (95% CI: 0.73, 0.86) in the internal test set, 0.84 (95% CI: 0.77, 0.91) in the external cohort, and 0.78 (95% CI: 0.73, 0.83) in the treatment test cohort. Calibration curves and DCA confirmed the models' excellent discrimination, calibration, and potential clinical utility across internal and external datasets. The AUCs of both nuclear medicine physicians were lower than those of the best-performing deep learning model in external test set (AUC: 0.75 vs. 0.77 vs. 0.84). CONCLUSION Deep learning models utilizing partial-body input from planar whole-body bone scintigraphy demonstrate high discriminatory power for diagnosing SBI in NPC patients, surpassing experienced nuclear medicine physicians.
Collapse
Affiliation(s)
- Xingyu Mu
- Department of Radiology, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, 530021, People's Republic of China
- Department of Nuclear Medicine, Affiliated Hospital of Guilin Medical University, Guilin, Guangxi Zhuang Autonomous Region, 541001, People's Republic of China
| | - Zhao Ge
- Department of Nuclear Medicine, Affiliated Hospital of Guilin Medical University, Guilin, Guangxi Zhuang Autonomous Region, 541001, People's Republic of China
| | - Denglu Lu
- Department of Nuclear Medicine, Liuzhou Workers' Hospital, Liuzhou, Guangxi Zhuang Autonomous Region, 545000, People's Republic of China
| | - Ting Li
- Department of Radiology, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, 530021, People's Republic of China
| | - Lijuan Liu
- Department of Radiology, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, 530021, People's Republic of China
| | - Cheng Chen
- Department of Nuclear Medicine, Affiliated Hospital of Guilin Medical University, Guilin, Guangxi Zhuang Autonomous Region, 541001, People's Republic of China
| | - Shulin Song
- Department of Radiology, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, 530021, People's Republic of China
- Department of Radiology, The Fourth People's Hospital of Nanning, Nanning, Guangxi Zhuang Autonomous Region, 530023, People's Republic of China
| | - Wei Fu
- Department of Nuclear Medicine, Affiliated Hospital of Guilin Medical University, Guilin, Guangxi Zhuang Autonomous Region, 541001, People's Republic of China.
| | - Guanqiao Jin
- Department of Radiology, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, 530021, People's Republic of China.
| |
Collapse
|
2
|
Lecouvet FE, Chabot C, Taihi L, Kirchgesner T, Triqueneaux P, Malghem J. Present and future of whole-body MRI in metastatic disease and myeloma: how and why you will do it. Skeletal Radiol 2024; 53:1815-1831. [PMID: 39007948 PMCID: PMC11303436 DOI: 10.1007/s00256-024-04723-2] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/20/2024] [Revised: 06/05/2024] [Accepted: 06/05/2024] [Indexed: 07/16/2024]
Abstract
Metastatic disease and myeloma present unique diagnostic challenges due to their multifocal nature. Accurate detection and staging are critical for determining appropriate treatment. Bone scintigraphy, skeletal radiographs and CT have long been the mainstay for the assessment of these diseases, but have limitations, including reduced sensitivity and radiation exposure. Whole-body MRI has emerged as a highly sensitive and radiation-free alternative imaging modality. Initially developed for skeletal screening, it has extended tumor screening to all organs, providing morphological and physiological information on tumor tissue. Along with PET/CT, whole-body MRI is now accepted for staging and response assessment in many malignancies. It is the first choice in an ever increasing number of cancers (such as myeloma, lobular breast cancer, advanced prostate cancer, myxoid liposarcoma, bone sarcoma, …). It has also been validated as the method of choice for cancer screening in patients with a predisposition to cancer and for staging cancers observed during pregnancy. The current and future challenges for WB-MRI are its availability facing this number of indications, and its acceptance by patients, radiologists and health authorities. Guidelines have been developed to optimize image acquisition and reading, assessment of lesion response to treatment, and to adapt examination designs to specific cancers. The implementation of 3D acquisition, Dixon method, and deep learning-based image optimization further improve the diagnostic performance of the technique and reduce examination durations. Whole-body MRI screening is feasible in less than 30 min. This article reviews validated indications, recent developments, growing acceptance, and future perspectives of whole-body MRI.
Collapse
Affiliation(s)
- Frederic E Lecouvet
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium.
| | - Caroline Chabot
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium
| | - Lokmane Taihi
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium
| | - Thomas Kirchgesner
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium
| | - Perrine Triqueneaux
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium
| | - Jacques Malghem
- Department of Medical Imaging, Institut de Recherche Expérimentale et Clinique (IREC), Institut du Cancer Roi Albert II, Cliniques Universitaires Saint Luc, Université Catholique de Louvain (UCL), Avenue Hippocrate, 10, B-1200, Brussels, Belgium
| |
Collapse
|
3
|
Thakur U, Ramachandran S, Mazal AT, Cheng J, Le L, Chhabra A. Multiparametric whole-body MRI of patients with neurofibromatosis type I: spectrum of imaging findings. Skeletal Radiol 2024:10.1007/s00256-024-04765-6. [PMID: 39105762 DOI: 10.1007/s00256-024-04765-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 03/27/2024] [Revised: 07/20/2024] [Accepted: 07/22/2024] [Indexed: 08/07/2024]
Abstract
Neurofibromatosis (NF) type I is a neuroectodermal and mesodermal dysplasia caused by a mutation of the neurofibromin tumor suppressor gene. Phenotypic features of NF1 vary, and patients develop benign peripheral nerve sheath tumors and malignant neoplasms, such as malignant peripheral nerve sheath tumor, malignant melanoma, and astrocytoma. Multiparametric whole-body MR imaging (WBMRI) plays a critical role in disease surveillance. Multiparametric MRI, typically used in prostate imaging, is a general term for a technique that includes multiple sequences, i.e. anatomic, diffusion, and Dixon-based pre- and post-contrast imaging. This article discusses the value of multiparametric WBMRI and illustrates the spectrum of whole-body lesions of NF1 in a single imaging setting. Examples of lesions include those in the skin (tumors and axillary freckling), soft tissues (benign and malignant peripheral nerve sheath tumors, visceral plexiform, and diffuse lesions), bone and joints (nutrient nerve lesions, non-ossifying fibromas, intra-articular neurofibroma, etc.), spine (acute-angled scoliosis, dural ectasia, intraspinal tumors, etc.), and brain/skull (optic nerve glioma, choroid plexus xanthogranuloma, sphenoid wing dysplasia, cerebral hamartomas, etc.). After reading this article, the reader will gain knowledge of the variety of lesions encountered with NF1 and their WBMRI appearances. Timely identification of such lesions can aid in accurate diagnosis and appropriate patient management.
Collapse
Affiliation(s)
- Uma Thakur
- Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX, 75235, USA
| | - Shyam Ramachandran
- Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX, 75235, USA
| | - Alexander T Mazal
- Department of Radiology, Stanford University School of Medicine, Palo Alto, CA, USA
| | - Jonathan Cheng
- Department of Plastic Surgery, UT Southwestern Medical Center, Dallas, TX, USA
| | - Lu Le
- Department of Dermatology and Simmons Comprehensive Cancer Center, UT Southwestern Medical Center, Dallas, TX, USA
| | - Avneesh Chhabra
- Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX, 75235, USA.
- Department of Orthopedic Surgery, UT Southwestern Medical Center, Dallas, TX, USA.
| |
Collapse
|
4
|
Vulasala SS, Virarkar M, Karbasian N, Calimano-Ramirez LF, Daoud T, Amini B, Bhosale P, Javadi S. Whole-body MRI in oncology: A comprehensive review. Clin Imaging 2024; 108:110099. [PMID: 38401295 DOI: 10.1016/j.clinimag.2024.110099] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/09/2023] [Revised: 01/28/2024] [Accepted: 01/31/2024] [Indexed: 02/26/2024]
Abstract
Whole-Body Magnetic Resonance Imaging (WB-MRI) has cemented its position as a pivotal tool in oncological diagnostics. It offers unparalleled soft tissue contrast resolution and the advantage of sidestepping ionizing radiation. This review explores the diverse applications of WB-MRI in oncology. We discuss its transformative role in detecting and diagnosing a spectrum of cancers, emphasizing conditions like multiple myeloma and cancers with a proclivity for bone metastases. WB-MRI's capability to encompass the entire body in a singular scan has ushered in novel paradigms in cancer screening, especially for individuals harboring hereditary cancer syndromes or at heightened risk for metastatic disease. Additionally, its contribution to the clinical landscape, aiding in the holistic management of multifocal and systemic malignancies, is explored. The article accentuates the technical strides achieved in WB-MRI, its myriad clinical utilities, and the challenges in integration into standard oncological care. In essence, this review underscores the transformative potential of WB-MRI, emphasizing its promise as a cornerstone modality in shaping the future trajectory of cancer diagnostics and treatment.
Collapse
Affiliation(s)
- Sai Swarupa Vulasala
- Department of Radiology, University of Florida College of Medicine, Jacksonville, FL, United States.
| | - Mayur Virarkar
- Department of Radiology, University of Florida College of Medicine, Jacksonville, FL, United States
| | - Niloofar Karbasian
- Department of Radiology, McGovern Medical School at University of Texas Health Houston, Houston, TX, United States
| | - Luis F Calimano-Ramirez
- Department of Radiology, University of Florida College of Medicine, Jacksonville, FL, United States
| | - Taher Daoud
- Division of Diagnostic Imaging, University of Texas MD Anderson Cancer Center, Houston, TX, United States
| | - Behrang Amini
- Division of Diagnostic Imaging, University of Texas MD Anderson Cancer Center, Houston, TX, United States
| | - Priya Bhosale
- Division of Diagnostic Imaging, University of Texas MD Anderson Cancer Center, Houston, TX, United States
| | - Sanaz Javadi
- Division of Diagnostic Imaging, University of Texas MD Anderson Cancer Center, Houston, TX, United States
| |
Collapse
|
5
|
Bauer F, Sauer S, Weinhold N, Delorme S, Wennmann M. (Smoldering) multiple myeloma: mismatch between tumor load estimated from bone marrow biopsy at iliac crest and tumor load shown by MRI. Skeletal Radiol 2023; 52:2513-2518. [PMID: 37300710 PMCID: PMC10582145 DOI: 10.1007/s00256-023-04383-8] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/05/2023] [Revised: 05/23/2023] [Accepted: 06/01/2023] [Indexed: 06/12/2023]
Abstract
In multiple myeloma and its precursor stages, precise quantification of tumor load is of high importance for diagnosis, risk assessment, and therapy response evaluation. Both whole-body MRI, which allows to investigate the complete bone marrow of a patient, and bone marrow biopsy, which is commonly used to assess the histologic and genetic status, are relevant methods for tumor load assessment in multiple myeloma. We report on a series of striking mismatches between the plasma cell infiltration estimating the tumor load from unguided biopsies of the bone marrow at the posterior iliac crest and the tumor load assessment from whole-body MRI.
Collapse
Affiliation(s)
- Fabian Bauer
- Division of Radiology, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120, Heidelberg, Germany.
| | - Sandra Sauer
- Department of Medicine V, Multiple Myeloma Section, University Hospital Heidelberg, Im Neuenheimer Feld 410, 69120, Heidelberg, Germany
| | - Niels Weinhold
- Department of Medicine V, Multiple Myeloma Section, University Hospital Heidelberg, Im Neuenheimer Feld 410, 69120, Heidelberg, Germany
| | - Stefan Delorme
- Division of Radiology, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120, Heidelberg, Germany
| | - Markus Wennmann
- Division of Radiology, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120, Heidelberg, Germany
| |
Collapse
|
6
|
Özgül HA, Akin IB, Mutlu U, Balci A. Diagnostic value of machine learning-based computed tomography texture analysis for differentiating multiple myeloma from osteolytic metastatic bone lesions in the peripheral skeleton. Skeletal Radiol 2023; 52:1703-1711. [PMID: 37014470 DOI: 10.1007/s00256-023-04333-4] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/28/2022] [Revised: 03/25/2023] [Accepted: 03/26/2023] [Indexed: 04/05/2023]
Abstract
OBJECTIVES To report the diagnostic performance of machine learning-based CT texture analysis for differentiating multiple myeloma from osteolytic metastatic bone lesions in the peripheral skeleton. METHODS We retrospectively evaluated 172 patients with multiple myeloma (n = 70) and osteolytic metastatic bone lesions (n = 102) in the peripheral skeleton. Two radiologists individually used two-dimensional manual segmentation to extract texture features from non-contrast CT. In total, 762 radiomic features were extracted. Dimension reduction was performed in three stages: inter-observer agreement analysis, collinearity analysis, and feature selection. Data were randomly divided into training (n = 120) and test (n = 52) groups. Eight machine learning algorithms were used for model development. The primary performance metrics were the area under the receiver operating characteristic curve and accuracy. RESULTS In total, 476 of the 762 texture features demonstrated excellent interobserver agreement. The number of features was reduced to 22 after excluding those with strong collinearity. Of these features, six were included in the machine learning algorithms using the wrapper-based classifier-specific technique. When all eight machine learning algorithms were considered for differentiating multiple myeloma from osteolytic metastatic bone lesions in the peripheral skeleton, the area under the receiver operating characteristic curve and accuracy were 0.776-0.932 and 78.8-92.3%, respectively. The k-nearest neighbors model performed the best, with the area under the receiver operating characteristic curve and accuracy values of 0.902 and 92.3%, respectively. CONCLUSION Machine learning-based CT texture analysis is a promising method for discriminating multiple myeloma from osteolytic metastatic bone lesions.
Collapse
Affiliation(s)
- Hakan Abdullah Özgül
- Department of Radiology, Kemalpaşa State Hospital, Kırovası Küme Street, Kemalpaşa, 35730, Izmir, Turkey.
| | - Işıl Başara Akin
- Department of Radiology, Dokuz Eylul University, Faculty of Medicine, Izmir, Turkey
| | - Uygar Mutlu
- Department of Radiology, Yozgat State Hospital, Yozgat, Turkey
| | - Ali Balci
- Department of Radiology, Dokuz Eylul University, Faculty of Medicine, Izmir, Turkey
| |
Collapse
|
7
|
Santoni A, Simoncelli M, Franceschini M, Ciofini S, Fredducci S, Caroni F, Sammartano V, Bocchia M, Gozzetti A. Functional Imaging in the Evaluation of Treatment Response in Multiple Myeloma: The Role of PET-CT and MRI. J Pers Med 2022; 12:jpm12111885. [PMID: 36579605 PMCID: PMC9696713 DOI: 10.3390/jpm12111885] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/14/2022] [Revised: 11/01/2022] [Accepted: 11/08/2022] [Indexed: 11/12/2022] Open
Abstract
Bone disease is among the defining characteristics of symptomatic Multiple Myeloma (MM). Imaging techniques such as fluorodeoxyglucose positron emission tomography-computed tomography (FDG PET/CT) and magnetic resonance imaging (MRI) can identify plasma cell proliferation and quantify disease activity. This function renders these imaging tools as suitable not only for diagnosis, but also for the assessment of bone disease after treatment of MM patients. The aim of this article is to review FDG PET/CT and MRI and their applications, with a focus on their role in treatment response evaluation. MRI emerges as the technique with the highest sensitivity in lesions' detection and PET/CT as the technique with a major impact on prognosis. Their comparison yields different results concerning the best tool to evaluate treatment response. The inhomogeneity of the data suggests the need to address limitations related to these tools with the employment of new techniques and the potential for a complementary use of both PET/CT and MRI to refine the sensitivity and achieve the standards for minimal residual disease (MRD) evaluation.
Collapse
|