Volume 34, Issue 1 (5-2026)
JSSU 2026, 34(1): 9759-9774 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Vaez E, Karimian A, Marateb H, Dadgar H, Arabi H. A Comprehensive Review of Machine and Deep Learning Approaches in PET and SPECT Images to Bone Metastasis Detection. JSSU 2026; 34 (1) :9759-9774
URL: http://jssu.ssu.ac.ir/article-1-6504-en.html
URL: http://jssu.ssu.ac.ir/article-1-6504-en.html
Abstract: (60 Views)
Introduction: Bone metastasis represents an advanced stage of various cancers and is a major cause of severe clinical complications. Accurate detection and classification of these lesions in nuclear medicine imaging play a critical role in evaluating disease burden and assessing its progression. The main objective of this review article was to categorize common machine learning and deep learning methods that have been applied for the segmentation and classification of bone metastases in nuclear medicine images.
Methods: This study was conducted in August 2025 and provided a comprehensive review of the literature using the keywords bone metastases, segmentation, classification, machine learning, and deep learning in the Web of Science (WoS) and PubMed databases. More than 500 articles were initially retrieved, and after applying the screening criteria, only those studies that aligned with the aims of this paper were selected. The included studies were then classified into four main categories based on the type of algorithm and imaging modality: machine learning–based studies, deep learning–based studies, studies utilizing positron emission tomography (PET), and those based on single-photon emission
computed tomography (SPECT).
Results: Machine learning and deep learning methods applied to SPECT and PET imaging have demonstrated high accuracy in detecting and classifying bone lesions, often outperforming
conventional diagnostic methods.
Conclusion: The integration of artificial intelligence into nuclear medicine imaging enables timely diagnosis, more precise clinical decision-making, and improved prognosis for patients with bone metastases.
Methods: This study was conducted in August 2025 and provided a comprehensive review of the literature using the keywords bone metastases, segmentation, classification, machine learning, and deep learning in the Web of Science (WoS) and PubMed databases. More than 500 articles were initially retrieved, and after applying the screening criteria, only those studies that aligned with the aims of this paper were selected. The included studies were then classified into four main categories based on the type of algorithm and imaging modality: machine learning–based studies, deep learning–based studies, studies utilizing positron emission tomography (PET), and those based on single-photon emission
computed tomography (SPECT).
Results: Machine learning and deep learning methods applied to SPECT and PET imaging have demonstrated high accuracy in detecting and classifying bone lesions, often outperforming
conventional diagnostic methods.
Conclusion: The integration of artificial intelligence into nuclear medicine imaging enables timely diagnosis, more precise clinical decision-making, and improved prognosis for patients with bone metastases.
Type of Study: Review article |
Subject:
other
Received: 2025/08/29 | Accepted: 2026/05/5 | Published: 2026/05/5
Received: 2025/08/29 | Accepted: 2026/05/5 | Published: 2026/05/5
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
