Diffusion-weighted imaging of bone and soft tissue tumors: a systematic review

Abstract

Diffusion-weighted magnetic resonance imaging (DWI) analyzes water diffusion in tissues, indirectly reflecting cellular density and aiding tumor characterization. Our objective was to explore the utility of DWI in three topics: (i) differentiation of bone and soft tissue sarcomas with respect to tumor type/grade; (ii) correlation with chemotherapy response in Ewing's sarcoma and osteosarcoma; and (iii) differentiation of benign from malignant soft tissue and bone tumors. The aim of the review article was to assess the existing literature detailing the insight that can be provided by DWI when addressing bone and soft tissue tumors. A search was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines using Pubmed, Scopus, Embase, and CENTRAL databases from 1 January 2013 to 30 November 2023. Eligible studies were assessed for quality using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) scoring system. In total, 22 studies met the inclusion criteria. DWI proved effective in select scenarios when distinguishing among bone and soft tissue sarcomas (sensitivity = 48%–89%, specificity = 75%–100%), correlating chemotherapy response with histopathology results (sensitivity = 25%–85%, specificity = 50%–100%), and differentiating between benign and malignant bone tumors (sensitivity = 54%–92%, specificity = 39%–92%) and soft tissue tumors (sensitivity = 68%–91%, specificity = 60%–81%). DWI is a valuable tool for the diagnosis and prognosis of bone and soft tissue sarcomas, treatment effect of primary bone tumors, and distinguishing benign from malignant bone/soft tissue tumors. Though promising, the technology has shown mixed results, warranting further research.

Keywords

Magnetic resonance diffusion/perfusion primary neoplasms soft tissues/skin skeletal-appendicular skeletal-axial muscular

Introduction

The evolution of magnetic resonance imaging (MRI) techniques has allowed for detailed insights into the landscape of bone and soft tissue tumors. Although primary malignant bone tumors account for <1% of malignancies and soft tissue sarcomas represent approximately 0.5%–1.0% of malignancies, they are a top five cause of cancer deaths for those aged under 20 years (1 –3). With a wide range of histopathology, they require comprehensive characterization, as management can vary tremendously, depending on the diagnosis. The gold standard to establish a tumor's diagnosis is an open or core needle biopsy (4). Although core needle biopsy has a substantially lower complication rate than incisional biopsy (<1% vs. 16%), these procedures all carry inherent anatomic and anesthetic risks (5,6). Furthermore, traditional MRI sequences, although useful in detecting tumor necrosis and hemorrhage, are limited in their ability to quantify the oncologic response to radiation and systemic therapies (7 –14). Diffusion-weighted imaging (DWI) adds another dimension to the ability of MRI characterization of bone and soft tissue tumors (15).

Since its inception in 1985, DWI has undergone continuous development and assumes a pivotal role in characterizing tissue functionalities within many organs such as the liver and brain. DWI provides imaging contrast by analyzing the diffusion property of water in tissues, illustrating the cellular density (16). When pathological processes occur in the human body, disturbances in water distribution between cells and extracellular compartments alter these diffusion properties. DWI provides insights into these changes, particularly in cases like high-grade malignancies where intracellular proportion increases, leading to more restricted diffusion (17). Images obtained through DWI can be utilized to calculate apparent diffusion coefficient (ADC) values, typically derived through non-linear regression applied to a series of images obtained with varying levels of diffusion weighting (diffusion moments or b-values) (18).

The aim of this systematic review was to assess the existing literature detailing the insight that can be provided by DWI when addressing bone and soft tissue tumors. To this end, this systematic review aimed to evaluate the collective body of evidence in three topics (Table 1) The three topics present in this table will be referenced as topic 1, topic 2, and topic 3 herein. This review aims to characterize the role that DWI plays in bone and soft tissue lesion diagnostics, ultimately improving clinical understanding of when and how to use this tool, and thereby improving clinical decision-making.

Table 1.

Topic summary with title, aim, and QUADAS-2 criteria.

Title, aim, and QUADAS-2 criteria by topic
	Title	Aim	Review question	Patient selection	Index test	Reference standard	Target condition
Topic 1	Evaluating the accuracy of DWI-MRI in differentiating among bone and soft tissue sarcomas	The authors aim to analyze DWI-MRI’s accuracy in distinguishing different bone and soft tissue sarcomas, exploring its potential as a diagnostic discriminator	In patients with bone and/or soft tissue sarcomas, how does DWI-MRI compare to histopathology when assessing sarcoma subtype or tumor grading?	Individuals who are known to have bone and/or soft tissue sarcomas	DWI-MRI	Histopathology	Characterization of lesions by sarcoma tumor subtype or grade
Topic 2	Evaluating the accuracy of DWI-MRI when correlating chemotherapy response in the treatment of Ewing's and osteosarcoma to histopathology results	The authors aim to examine DWI-MRI's accuracy in linking chemotherapy responses and histopathology outcomes, shedding light on its role in tracking treatment efficacy at a cellular level	In patients with Ewing's or osteosarcoma, how does DWI-MRI compare to histopathology when assessing adjuvant and neoadjuvant chemotherapy responses?	Individuals who are known to have osteosarcoma or Ewing's sarcoma	DWI-MRI	Histopathology	Evaluating chemotherapy response
Topic 3	Evaluating the accuracy of DWI-MRI when differentiating benign from malignant soft tissue and bone lesions	The authors aim to reveal DWI-MRI's accuracy in distinguishing between benign and malignant soft tissue and bone lesions to better understand its potential for refined diagnostic categorization	In patients with benign or malignant bone or soft tissue tumors, how does DWI-MRI compare to histopathology when characterizing lesions as benign or malignant?	Individuals who are known to have benign or malignant bone or soft tissue tumors	DWI-MRI	Histopathology	Characterization of lesions as benign or malignant

DWI, diffusion-weighted imaging; MRI, magnetic resonance imaging.

Material and Methods

Review protocol

This retrospective review followed the Preferred Reporting Items for Systematic Reviews and Meta Analysis (PRISMA) (19).

Search strategy

The authors searched Pubmed, Scopus, Embase, and The Cochrane Central Register of Controlled Trials (CENTRAL) for relevant studies. The exact search criteria are shown in Table 2. The date range of the query was set to 1 January 2013 to 30 November 2023.

Table 2.

Query criteria for database search.

	Query criteria for search	Results
Pubmed	((Diffusion Weighted Imaging) OR (Apparent Diffusion Coefficient)) AND ((Orthopedic Oncology) OR (Orthopaedic Oncology) OR (Musculoskeletal Oncology) OR (Malignant Bone Tumors) OR (Malignant Primary Bone Tumors) OR (Osteosarcoma) OR (Ewing Sarcoma) OR (Chondrosarcoma) OR (Benign Primary Bone Tumors) OR (Malignant Soft Tissue Sarcoma) OR (Soft Tissue Sarcoma) OR (Metastatic Bone Lesions) OR (Osseous Metastases))	857
Scopus	TITLE-ABS-KEY (((diffusion AND weighted AND imaging) OR (apparent AND diffusion AND coefficient)) AND ((orthopedic AND oncology) OR (orthopaedic AND oncology) OR (musculoskeletal AND oncology) OR (malignant AND bone AND tumors) OR (malignant AND primary AND bone AND tumors) OR (osteosarcoma) OR (ewing AND sarcoma) OR (chondrosarcoma) OR (benign AND primary AND bone AND tumors) OR (malignant AND soft AND tissue AND sarcoma) OR (soft AND tissue AND sarcoma) OR (metastatic AND bone AND lesions) OR (osseous AND metastases))) AND (LIMIT-TO (SUBJAREA, “MEDI”))	796
Embase	((Diffusion Weighted Imaging or Apparent Diffusion Coefficient) and (Orthopedic Oncology or Orthopaedic Oncology or Musculoskeletal Oncology or Malignant Bone Tumors or Malignant Primary Bone Tumors or Osteosarcoma or Ewing Sarcoma or Chondrosarcoma or Benign Primary Bone Tumors or Malignant Soft Tissue Sarcoma or Soft Tissue Sarcoma or Metastatic Bone Lesions or Osseous Metastases))	489
Cochrane	#1. MeSH descriptor: [Diffusion Magnetic Resonance Imaging] explode all trees #2. MeSH descriptor: [Neoplasms, Connective and Soft Tissue] explode all trees #3. #1 AND #2	1

Study eligibility and selection and quality of included studies

Our search strategy resulted in a total of 2143 articles requiring further analysis for inclusion. The exact breakdown of the number of articles from each database is shown in Table 2. Then, a title and abstract review were independently conducted by each of the authors. Articles that met criteria for further analysis included prospective cohort studies, retrospective cohort studies, and case-control studies that discussed DWI in the context of musculoskeletal oncology. We purposefully stayed broad at this step, trying to encompass all relevant studies analyzing DWI with respect to bone and soft tissue oncologic pathologies due to the paucity of literature available. This approach aimed to summarize the landscape of use cases for DWI in musculoskeletal oncology comprehensively. Case reports, animal studies, cadaveric studies, non-English studies, other systematic reviews, and author replies were excluded. Disagreements were resolved by the corresponding author. A total of 121 studies resulted after this due diligence. The full text of all eligible studies was then independently reviewed by the authors using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) for quality assessment , resulting in 22 qualifying studies (Figure 1).

Fig. 1.

Flow chart of literature review. After primarily excluding studies through title and abstract review and secondarily excluding additional studies with full-text review and applying QUADAS-2 criteria, 22 studies qualified for inclusion.

QUADAS-2 is widely used in systematic reviews as a guide to evaluate the methodological quality of diagnostic accuracy studies and can be helpful in identifying potential sources of bias and variations in quality across different studies. It consists of four domains: (i) patient selection, (ii) index test, (iii) reference standard, and (iv) flow and timing. Within each domain, specific items are assessed to determine the risk of bias and concerns regarding applicability (domain 4 [Flow and Timing] is only used to assess concerns regarding applicability). It is important to note that for the purposes of this systematic review, QUADAS-2 was used independently in the context of topics 1, 2, and 3 – each topic was assigned a separate review question, which was formulated using a patient selection, index test, reference standard, and target condition for the three aforementioned topics to properly address risk of bias and concerns regarding applicability (Table 1). This modification was made to ensure QUADAS-2 was applied appropriately for each topic discussed in this systematic review (20).

Fig. 2 provides a visualization of the assessment of all relevant articles with respect to risk of bias and concern of applicability. The articles that were determined to have low risk of bias and low concern of applicability were ultimately included. The evaluation of the included studies using QUADAS-2 alongside each study's location, population, and imaging technique can be found in Table for topics 1, 2 and 3, respectively.

Fig. 2.

Graphs of overall risk of bias and concern of applicability for all relevant studies by QUADAS-2 domain.

Table 3.

Location, population, and imaging details of studies included for topics 1, 2, and 3.

Title	Author (year)	Country	Diagnostic test used	MRI strength	No. of patients	Age (years)*	Sex
Topic 1: evaluating the accuracy of DWI-MRI in differentiating among bone and soft tissue sarcoma
Comparison of ADC values in different malignancies of the skeletal musculature: a multicentric analysis	Surov et al. (2015) (21)	USA	DWI	1.5 T	51	58.8 ± 16.1	23 M, 28 F
Osteosarcoma subtypes: magnetic resonance and quantitative diffusion weighted imaging criteria	Zeitoun et al. (2018) (22)	USA	Conventional, contrast-enhanced, and DWI	1.5 T	31	19 (7–46)	18 M, 13 F
Conventional MR and diffusion-weighted imaging of musculoskeletal soft tissue malignancy: correlation with histologic grading	Chhabra et al. (2019)	France	Conventional, contrast-enhanced, and DWI	27 scans on 1.5 T; 24 scans on 3.0 T	51	52 ± 16	22 M, 29 F
Can MRI diffusion-weighted imaging identify postoperative residual/recurrent soft-tissue sarcomas?	ElDaly et al. (2018) (23)	India	Conventional, contrast-enhanced, and DWI	1.5 T	36	38 (17–73)	20 M, 16 F
Does the apparent diffusion coefficient from diffusion-weighted MRI imaging aid in the characterization of malignant soft tissue tumors and sarcomas	Gowda et al. (2023)	USA	Conventional and DWI	49 scans on 1.5 T; 29 scans on 3.0 T	78	Inclusion criteria: patients aged 18–100 years	47 M, 53 F
Topic 2: Accuracy of DWI-MRI when correlating chemotherapy response to histopathology results
Prediction of poor responders to neoadjuvant chemotherapy in patients with osteosarcoma: additive value of diffusion-weighted MRI including volumetric analysis to standard MRI at 3T	Lee et al. (2020) (24)	Republic of Korea	DWI	3.0 T	17	17 (10–53)	13 M, 4 F
Diffusion-weighted imaging in differentiating mid-course responders to chemotherapy for long-bone osteosarcoma compared to the histologic response: an update	Habre et al. (2020) (25)	France	DWI	1.5 T	26	13 (4.7–19.6)	11 M, 15 F
The value of diffusion weighted imaging and apparent diffusion coefficient in primary osteogenic and Ewing sarcomas for the monitoring of response to treatment: Initial experience	Asmar et al. (2020) (26)	Lebanon	DWI	3.0 T	15	11.53 ± 3.52	9 M, 6 F
Multiparametric MRI with diffusion-weighted imaging in predicting response to chemotherapy in cases of osteosarcoma and Ewing's sarcoma	Saleh et al. (2020) (27)	Egypt	Conventional, contrast-enhanced, and DWI	1.5 T	104	20 ± 10.5 (2 to 49)	–
Correlation of histopathology and multimodal magnetic resonance imaging in childhood osteosarcoma: Predicting tumor response to chemotherapy	Teo et al. (2020) (28)	USA	DWI	3.0 T	15	13 (10–20)	7 M, 8 F
Topic 3: Accuracy of DWI-MRI when differentiating benign from malignant soft tissue and bone lesions
The role of diffusion-weighted MRI (DWI) in the differentiation of benign from malignant skeletal lesions of the pelvis	Douis et al. (2016) (29)	Ireland	DWI	3.0 T	33	48.6 (14–77)	16 M, 17 F
Solid bone tumors of the spine: Diagnostic performance of apparent diffusion coefficient measured using diffusion-weighted MRI using histology as a reference standard	Pozzi et al. (2018) (30)	Italy	Conventional and DWI	1.5 T	116	59.5 ± 14.1	62 M, 54 F
Assessing the relation between bone marrow signal intensity and apparent diffusion coefficient in diffusion-weighted MRI	Padhani et al. (2013) (31)	UK	Conventional and DWI	1.5 T	49	Cohort 1 = 47.7 (22–65); cohort 2 = 53.5 (37–72); cohort 3 = 61.1 (29–80)	16 M, 33 F
Role of apparent diffusion coefficients with diffusion-weighted magnetic resonance imaging in differentiating between benign and malignant bone tumors	Wang et al. (2014) (32)	China	Conventional, contrast-enhanced, and DWI	3.0 T	178	31.52 ± 28.31	81 M, 97 F
Correlation of quantitative diffusion weighted MR imaging between benign, malignant chondrogenic and malignant non-chondrogenic bone tumors with histopathologic type	Setiawati (2021) (33)	Italy	Conventional and DWI	3.0 T	84	32.7 (10–73)	44 M, 40 F
Differentiation of benign and malignant skeletal lesions with quantitative diffusion weighted MRI at 3T	Ahlawat et al. (2015) (34)	USA	Conventional, contrast-enhanced, and DWI	3.0 T	31	Benign lesions: 26.4 ± 19.7 Malignant lesions: 50.8 ± 22.1	15 M, 16 F
Apparent diffusion coefficient (ADC): A potential in vivo biological surrogate of the incidentally discovered bone lesions at 3T MRI	Nouh et al. (2021) (35)	Egypt, UK, Kuwait	DWI	3.0 T	44	39.2 (2–73)	24 M, 20 F
Characterization of soft tissue tumors by diffusion-weighted imaging	Pekcevik et al. (2015) (36)	Turkey	DWI	1.5 T	25	36.6 (3–77)	17 M, 18 F
Diagnostic performance of conventional MRI parameters and apparent diffusion coefficient values in differentiating between benign and malignant soft-tissue tumours	Song et al. (2017) (37)	Republic of Korea	Conventional and DWI	3.0 T	123	46.2 (7–81)	70 M, 53 F
Diagnostic performance of conventional MRI parameters and apparent diffusion coefficient values in differentiating between benign and malignant soft-tissue tumours	Robba et al. (2017) (38)	Italy	DWI	1.5 T	46	57 (12–85)	27 M, 19 F
Diffusion weighted imaging of extremity bone tumors-inter-reader analysis and incremental value over conventional MR imaging	Guirguis et al. (2023)(39)	USA	Conventional and DWI	53 scans on 1.5 T; 80 scans on 3.0 T	133	Benign lesions: 47.2 ± 16.7 Malignant lesions: 50.5 ± 19.3	69 M, 64 F
Osseous-tissue tumor reporting and data system with diffusion-weighted imaging of bone tumors—an interreader analysis and whether it adds incremental value on tumor grading over conventional magnetic resonance imaging	Guirguis et al. (2023)(39)	USA	Conventional and DWI	53 scans on 1.5 T; 80 scans on 3.0 T	133	Benign lesions: 47.2 ± 16.7 Malignant lesions: 50.5 ± 19.3	69 M, 64 F

*Values are given as mean ± SD or median (range).

DWI, diffusion-weighted imaging; MRI, magnetic resonance imaging.

Table 4.

Summary of results of topics 1, 2, and 3.

Author (Year)	Variable Being Evaluated	Number of Lesions	^⸶Group-specific ADC Values	Proposed ADC Cut Off (10^-3 mm^2/s)	Sensitivity	Specificity	P-Value
Topic 1 Results Summary
ElDaly et al (2018) (23)	Recurrent vs. Non-Recurrent Soft Tissue Sarcoma Lesions (ADC Mean)	27 vs. 9	-	<=1.4	85%	100%	<0.0001*
ElDaly et al (2018) (23)	Recurrent vs. Non-Recurrent Soft Tissue Sarcoma Lesions (ADC Minimum)	27 vs. 9	-	<=1.2	89%	89%	<0.0001*
Surov et al (2015) (21)	Muscle Lymphoma vs. Muscle Metastases (ADC Mean)	14 vs. 11	0.76 vs. 1.28	-	-%	-%	0.01*
Surov et al (2015) (21)	Muscle Lymphoma vs. Muscle Sarcoma (ADC Mean)	14 vs. 26	0.76 vs. 1.82	-	-%	-%	0.001*
Surov et al (2015) (21)	Muscle Metastases vs. Muscle Sarcoma (ADC Mean)	11 vs. 26	1.28 vs. 1.82	-	-%	-%	0.48
Zeitoun et al (2018) (22)	Osteoblastic Osteosarcoma (ADC Mean)	9	-	1.01 (a)	-%	-%	-
Zeitoun et al (2018) (21)	Fibroblastic Osteosarcoma (ADC Mean)	8	-	1.12 (a)	-%	-%	-
Zeitoun et al (2018) (21)	Chondroblastic Osteosarcoma (ADC Mean)	6	-	1.32 (a)	-%	-%	-
Chhabra et al (2019)	Grade III vs Grade I/II Tumor (ADC Mean)	26 vs. 25	-	0.88	62%	80%	-
Chhabra et al (2019)	Grade III vs Grade I/II Tumor (Excluding Myxoid and Chondroid) (ADC Mean)	-	-	0.74	48%	85%	-
Chhabra et al (2019)	Grade I vs Grade II/III Tumors (ADC Mean)	43 vs. 8	-	0.98	56%	88%	-
Chhabra et al (2019)	Grade I vs Grade II/III Tumors (Excluding Myxoid and Chondroid) (ADC Mean)	-	-	0.98	66%	75%	-
Gowda et al (2023)	Liposarcomas vs. Non-Liposarcomas (Whole Tumor ADC Mean)	67 vs. 11	-	-	-%	-%	Reader 1: 0.03* Reader 2: 0.04*
Gowda et al (2023)	Liposarcomas vs. Non-Liposarcomas (Whole Tumor ADC Minimum)	67 vs. 11	-	-	-%	-%	Reader 1: 0.17 Reader 2: 0.53
Gowda et al (2023)	Liposarcomas vs. Non-Liposarcomas (Darkest Area ADC Mean)	67 vs. 11	-	-	-%	-%	Reader 1: <0.01* Reader 2: 0.01*
Gowda et al (2023)	Liposarcomas vs. Non-Liposarcomas (Darkest Area ADC Minimum)	67 vs. 11	-	-	-%	-%	Reader 1: 0.02* Reader 2: <0.01*
Gowda et al (2023)	Myxoid Tumors vs. Non-Myxoid Tumors (Whole Tumor ADC Mean)	66 vs. 12	-	-	-%	-%	Reader 1: 0.03* Reader 2: 0.06
Gowda et al (2023)	Myxoid Tumors vs. Non-Myxoid Tumors (Whole Tumor ADC Minimum)	66 vs. 12	-	-	-%	-%	Reader 1: 0.10 Reader 2: 0.08
Gowda et al (2023)	Myxoid Tumors vs. Non-Myxoid Tumors (Darkest Area ADC Mean)	66 vs. 12	-	-	-%	-%	Reader 1: 0.07 Reader 2: <0.01*
Gowda et al (2023)	Myxoid Tumors vs. Non-Myxoid Tumors (Darkest Area ADC Minimum)	66 vs. 12	-	-	-%	-%	Reader 1: 0.35 Reader 2: <0.01*
Gowda et al (2023)	Synovial Tumors vs. Non-synovial Tumors (Whole Tumor ADC Mean)	66 vs. 12	-	-	-%	-%	Reader 1: 0.57 Reader 2: <0.01*
Gowda et al (2023)	Synovial Tumors vs. Non-synovial Tumors (Whole Tumor ADC Minimum)	66 vs. 12	-	-	-%	-%	Reader 1: <0.01* Reader 2: 0.02*
Gowda et al (2023)	Synovial Tumors vs. Non-synovial Tumors (Darkest Area ADC Mean)	66 vs. 12	-	-	-%	-%	Reader 1: 0.09 Reader 2: 0.04*
Gowda et al (2023)	Synovial Tumors vs. Non-synovial Tumors (Darkest Area ADC Minimum)	66 vs. 12	-	-	-%	-%	Reader 1: 0.12 Reader 2: 0.04*
Topic 2 Results Summary
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (2D ADC Mean) (Reader 1)	13 vs. 4	-	<=2.079	85%	75%	0.017*
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (2D ADC Mean) (Reader 2)	13 vs. 4	-	<=1.783	69%	75%	0.089
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (2D ADC Minimum) (Reader 1)	13 vs. 4	-	<=1.442	85%	100%	0.024*
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (2D ADC Minimum) (Reader 2)	13 vs. 4	-	<=1.481	77%	75%	0.089
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (3D ADC Mean)	13 vs. 4	-	<=2.039	85%	50%	0.042*
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (3D ADC Skewness)	13 vs. 4	-	>-0.82	85%	100%	0.017*
Lee et al (2020) (24)	Poor vs. Good Responder Post Neoadjuvant Treatment for Osteosarcoma (3D ADC Kurtosis)	13 vs. 4	-	<=4.9	77%	75%	0.258
Habre et al (2020) (25)	Poor vs. Good Responder Mid Course of Neoadjuvant Treatment for Long-Bone Osteosarcoma (ADC Mean Initial vs. ADC Mean Mid Course)	12 vs. 14	-	>0.07 (b)	25%	100%	0.046*
Habre et al (2020) (25)	Poor vs. Good Responder Post Neoadjuvant Treatment for Long-Bone Osteosarcoma (ADC Mean Initial vs. ADC Mean Final)	12 vs. 14	-	-	-%	-%	0.014*
Asmar et al (2020) (26)	Pre-Treatment vs. Post-Treatment Osteosarcoma and Ewing's Sarcoma (ADC Mean Initial vs. ADC Mean Final)	15	1.191 vs. 1.534	-	-%	-%	0.03*
Asmar et al (2020) (26)	Pre-Treatment vs. Post-Treatment Osteosarcoma and Ewing's Sarcoma (ADC Minimum Initial vs. ADC Minimum Final)	15	0.134 vs. 0.388	-	-%	-%	0.02*
Saleh et al (2020) (27)	Pre-Treatment vs. Post-Treatment Osteosarcoma Sarcoma (ADC Mean Initial vs. ADC Mean Final)	51	0.90 vs. 1.62	-	-%	-%	<0.001*
Saleh et al (2020) (27)	Pre-Treatment vs. Post-Treatment Osteosarcoma Sarcoma (ADC Minimum Initial vs. ADC Minimum Final)	51	0.71 vs. 1.23	-	-%	-%	<0.001*
Saleh et al (2020) (27)	Pre-Treatment vs. Post-Treatment Ewing's Sarcoma (ADC Mean Initial vs. ADC Mean Final)	51	0.71 vs. 1.60	-	-%	-%	<0.001*
Saleh et al (2020) (27)	Pre-Treatment vs. Post-Treatment Ewing's Sarcoma (ADC Minimum Initial vs. ADC Minimum Final)	51	0.42 vs. 1.07	-	-%	-%	<0.001*
Topic 3 Results Summary
Malignant vs. Benign Bone Lesions
Douis et al (2016) (29)	Malignant vs Benign Pelvic Skeletal Lesions (ADC Average)	13 vs. 20	1.422 vs. 1.264	-	-%	-%	0.29
Douis et al (2016) (29)	Malignant vs Benign Pelvic Skeletal Lesions (ADC Minimum)	13 vs. 20	0.780 vs. 0.772	-	-%	-%	0.94
Douis et al (2016) (29)	Malignant vs Benign Pelvic Skeletal Lesions (ADC Maximum)	13 vs. 20	1.969 vs. 1.677	-	-%	-%	0.149
Douis et al (2016) (29)	Malignant vs Benign Pelvic Skeletal Lesions (Qualitative Assessment)	13 vs. 20	-	-	54%	85%	0.026*
Pozzi et al (2018) (30)	Malignant (Primary + Metastatic) vs. Benign Spine Bone Tumors (ADC Mean)	100 vs. 16	-	0.904	90%	40%	-
Pozzi et al (2018) (30)	Malignant (Metastatic) vs. Benign Spine Bone Lesions (ADC Mean)	65 vs. 16	-	0.904	90%	39%	-
Pozzi et al (2018) (30)	Malignant (Primary) vs. Benign Spine Bone Lesions (ADC Mean)	35 vs. 16	-	0.902	90%	43%	-
Padhani et al (2013) (31)	Bone Marrow Signal Intensity	29	-	0.774	85%	90%	<0.001*
Wang et al (2014) (32)	Malignant vs. Benign Bone Lesions (ADC Mean)	114 vs. 84	-	>= 1.1	90%	85%	<0.05* (c)
Setiawati (2021) (33)	Malignant vs. Benign Bone Lesions (ADC Mean)	57 vs. 27	-	>=1.15	82%	92%	<0.001*
Ahlawat et al (2015) (34)	Malignant vs. Benign Bone Lesions (ADC Minimum)	13 vs. 18	-	>=0.9	92%	78%	<0.0001*
Ahlawat et al (2015) (34)	Malignant vs. Benign Bone Lesions (ADC Mean)	13 vs. 18	-	>=1.4	77%	72%	<0.0001*
Ahlawat et al (2015) (34)	Malignant vs. Benign Bone Lesions (ADC Maximum)	13 vs. 18	-	>=1.8	62%	78%	0.04*
Nouh et al (2021) (35)	Malignant vs. Benign Bone Lesions (ADC Mean)	36 vs. 16	-	>=1.1	86%	63%	<0.001*
Guirguis et al (2023) (Study 1 & 2)	Malignant vs. Benign Bone Lesions (Whole Lesion ADC Mean)	57 vs. 76	1.431 vs. 1.142	-	-%	-%	0.006*
Guirguis et al (2023) (Study 1 & 2)	Malignant vs. Benign Bone Lesions (Whole Lesion ADC Minimum)	57 vs. 76	0.813 vs. 0.474	-	-%	-%	0.001*
Guirguis et al (2023) (Study 1 & 2)	Malignant vs. Benign Bone Lesions (Darkest Area ADC Mean)	57 vs. 76	1.253 vs. 0.902	-	-%	-%	0.007*
Guirguis et al (2023) (Study 1 & 2)	Malignant vs. Benign Bone Lesions (Darkest Area ADC Minimum)	57 vs. 76	0.989 vs. 0.610	-	-%	-%	0.003*
Malignant vs. Benign Soft Tissue Lesions
Robba et al (2017) (38)	Malignant vs. Benign Soft Tissue Lesions (ADC Mean)	34 vs. 10	-	>=1.45	91%	60%	-
Pekcevik et al (2015) (36)	Malignant vs. Benign Soft Tissue Tumors (ADC Mean)	6 vs. 19	0.90 vs. 2.31	-	-%	-%	0.031*
Song et al (2017) (37)	Malignant vs. Benign Soft Tissue Tumors (ADC Mean)	54 vs. 69	-	>=1.35	68%	73%	<0.001*
Song et al (2017) (37)	Malignant vs. Benign Soft Tissue Tumors (ADC Minimum)	54 vs. 69	-	>=0.81	68%	67%	0.001*
Song et al (2017) (37)	Malignant vs. Benign Soft Tissue Tumors (Non-Myxoid & Non-Hemosiderin) (ADC Mean)	54 vs. 51	-	>=1.13	83%	81%	<0.001*
Song et al (2017) (37)	Malignant vs. Benign Soft Tissue Tumors (Non-Myxoid & Non-Hemosiderin) (ADC Minimum)	54 vs. 51	-	>=0.631	77%	76%	0.001*

Note: Teo et al was excluded due to different study design/parameters.

a) ADC Mean values were not correlated or compared statistically between lesion types

b) Represents the difference in ADC values before and mid-course/after treatment

c) Represents p-value related to the difference between the mean ADC values in malignant and benign bone lesion groups

* indicates statistical significance

^†

The specific ADC metric (e.g., mean) is not specified, as group-specific ADC values were reported using different summary measures across studies. The metric of interest is noted in parentheses in the “Variable Being Evaluated” column.

Results

Quality of the included studies

Among the studies assessed utilizing QUADAS-2, 22 were found to have low risk of bias with low concern of applicability, five were found to have low risk of bias with high concern of applicability, 10 were found have to high risk of bias with low concern of applicability, and three were found to have high risk of bias with high concern of applicability. The findings of the QUADAS-2 assessment are depicted in Fig. 2, while the results of all studies are displayed in Table 4.

Topic 1: Evaluating the accuracy of DWI-MRI in differentiating among bone and soft tissue sarcomas

ElDaly et al. identified an ADC_mean cutoff of ≤1.4 ×10⁻³ mm²/s, achieving a sensitivity of 85%, specificity of 100%, and P <0.0001 in detecting recurrent postoperative soft tissue sarcoma tumors (23). Surov et al. found that muscle lymphoma had significantly lower ADC_mean values than muscle metastases (P = 0.01) and muscle sarcoma (P = 0.001) (21). An excellent example of DWI characterizing lymphoma is shown in Fig. 3. Zeitoun et al. observed that osteoblastic osteosarcoma exhibited the lowest ADC_mean values among osteosarcoma subtypes (22). Chhabra et al. reported that ADC cutoffs between 0.74 to 0.98 ×10⁻³ mm²/s provided a sensitivity of 48%–66% and specificity of 75%–88% in differentiating tumor grades (39). Gowda et al. found significant ADC_mean differences in liposarcomas versus non-liposarcomas and synovial tumors versus non-synovial tumors, highlighting variability within soft tissue sarcomas (36).

Topic 2: Accuracy of DWI-MRI when correlating chemotherapy response in the treatment of Ewing's sarcoma and osteosarcoma to histopathology results

Fig. 3.

(a, b) A 45-year-old woman with a history of lymphoma with multiple enhancing lesions throughout the osseous structures: axial T1 post-contrast. (c) DWI B800 and (d) ADC sequences demonstrating the typical avid diffusion restriction associated with highly cellular tumors such as lymphoma. Reference the lesion in the left ilium (arrow), which has an ADC measurement of 0.66 ×10⁻³ mm²/s. ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Studies by Lee et al. and Habre et al. demonstrated that DWI could predict chemotherapy response in osteosarcoma, with post-treatment ADC_mean and ADC_skewness being strong indicators of poor treatment response (24,25). Saleh et al. showed that post-chemotherapy ADC_mean values significantly increased for osteosarcoma (0.90 to 1.62 ×10⁻³ mm²/s) and Ewing's sarcoma (0.71 to 1.6 ×10⁻³ mm²/s; P <0.001) (27). An example of ADC value in osteosarcoma correlating with partial chemotherapy response can be seen in Fig. 4. Asmar et al. confirmed these findings in pediatric populations, reporting significant increases in ADC_mean (1.191 to 1.534 ×10⁻³ mm²/s; P = 0.03) and ADC_min (0.134 to 0.388 ×10⁻³ mm²/s; P = 0.02) (26). However, Teo et al. noted that integrating DWI with conventional MRI did not improve tumor necrosis estimation accuracy (28).

Topic 3: Accuracy of DWI-MRI when differentiating benign from malignant soft tissue and bone tumors

Fig. 4.

A 27-year-old woman with a large, heterogenous left distal femur osteosarcoma with significant soft tissue mass: (a) sagittal STIR; (b) coronal T1 post-contrast). (c) Pre-treatment (axial) and (d) post-neoadjuvant chemotherapy (sagittal) DWI/ADC images obtained show significant decrease in diffusion restriction (ADC increase from 1.05 ×10⁻³ mm²/s to 1.62 ×10⁻³ mm²/s) after four cycles of neoadjuvant chemotherapy, suggesting partial response to therapy. Final pathology after surgical excision revealed 50% necrosis of the tumor. ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Malignant versus benign bone tumors

Douis et al. found ADC_mean effective in differentiating benign from malignant bone tumors, with restricted diffusion (P = 0.026) (29). Pozzi et al. reported that ADC_mean cutoffs of approximately 0.904 ×10⁻³ mm²/s had a sensitivity of 90% for differentiating primary from metastatic malignancies (30). Padhani et al. identified a cutoff of 0.774 ×10⁻³ mm²/s as consistent with malignancy (P <0.001) (31). Wang et al. showed that an ADC_mean ≥1.10 ×10⁻³ mm²/s provided a sensitivity and specificity of 90% and 85%, respectively, for distinguishing benign from malignant tumors (32). Setiawati et al. demonstrated that an ADC cutoff ≥1.15 ×10⁻³ mm²/s effectively differentiated benign from malignant cartilaginous tumors (P <0.001) (33). In Fig. 5, we show a pelvic dedifferentiated chondrosarcoma with ADC value measuring 1.01 ×10⁻³mm²/s. Ahlawat et al. found an ADC_min cutoff of 0.9 ×10⁻³ mm²/s with 92% sensitivity and 78% specificity for differentiating benign from malignant bone tumors (34). Nouh et al. found that an ADC_mean cutoff value ≥1.10 ×10⁻³ mm²/s effectively characterized a tumor as benign, resulting in a sensitivity and specificity of 86% and 63%, respectively (35). Guirguis et al. highlighted the utility of whole tumor and darkest area ADC measurements but noted that conventional MRI alone was sufficient for bone tumor characterization (37,38).

Fig. 5.

A 61-year-old woman with a large dedifferentiated chondrosarcoma involving the right pelvis. (a, b) T1 post-contrast axial (a) and coronal (b) images demonstrate a heterogeneously enhancing destructive mass with large soft tissue component. (c) DWI B800 and (d) ADC sequences depicting areas of diffusion restriction, with ADC measuring as low as 1.01 ×10⁻³ mm²/s. ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Malignant versus benign soft tissue tumors

Pekcevik et al. reported that benign cystic tumors had higher ADC_mean values than benign or malignant solid/mixed tumors (P <0.001 and P = 0.003, respectively). The authors also found that malignant solid or mixed tumors had lower ADC values than that of benign solid or mixed tumors (P = 0.02) (36). This concept is illustrated well by Fig. 6 (rhabdomyosarcoma with ADC value of 1.11 ×10⁻³ mm²/s and benign myxoma with an ADC value of 2.62 ×10⁻³ mm²/s). Song et al. identified ADC_mean and ADC_min cutoffs of ≥1.35 × 10⁻³ mm²/s and ≥0.81 × 10⁻³ mm²/s, respectively, as effective for differentiating malignant from benign soft tissue tumors (P <0.001 and P = 0.001) (37). Robba et al. confirmed these findings, with an ADC_mean cutoff ≥1.45 ×10⁻³ mm²/s showing a sensitivity and specificity of 91% and 60% in differentiating benign from malignant soft tissue tumors (38).

Fig. 6.

(a–c) A 60-year-old man with a large right thigh soft tissue rhabdomyosarcoma exhibiting predominantly peripheral irregular nodular enhancement with areas of central necrosis (panel a = axial T1 post-contrast image). (b, c) The enhancing nodular components of the mass demonstrate diffusion restriction, with ADC measuring 1.11 × 10⁻³ mm²/s. (d–f) In contrast, a 57-year-old woman with a right deltoid benign intramuscular myxoma. (d) Axial T1 post-contrast image shows mild thin peripheral and internal enhancement of the mass. (e) DWI B800 and (f) ADC sequences depicting “T2 shine through,” and thus the lack of diffusion restriction. The ADC_mean for this lesion measures 2.62 ×10⁻³ mm²/s. ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Discussion

DWI was originally developed for applications in neuroimaging, particularly for detecting acute ischemic stroke (40). DWI functions by analyzing stochastic Brownian motion of water molecules at a microscopic level. Given intracellular water molecules have less freedom to move than extracellular water molecules (due to intracellular organelles and membranes), DWI sequences provide highly detailed information regarding cellularity for an analyzed tissue. This sensitivity was thought to make it useful for detecting changes associated with pathological musculoskeletal conditions where tumor, inflammation, and degenerative processes impact the cellular composition of an analyzed tissue (41). However, the consensus surrounding DWI's utility in musculoskeletal oncology is mixed and nuanced. Promising outcomes have been observed in certain pathologies, such as distinguishing between benign and malignant vertebral fractures, but its use in other musculoskeletal conditions remains more controversial (42).

The application of DWI in distinguishing various bone and soft tissue sarcomas based on ADC values yields consistent trends. Notably, proposed ADC cutoff values demonstrate high sensitivity and specificity in detecting recurrent soft tissue sarcomas in routine surveillance imaging (23). Studies exploring DWI's effectiveness in distinguishing between muscle lymphoma, sarcoma, and metastases reveal statistically significant differences in ADC_mean values, supporting its role in differential diagnoses (21). The examination of different pathological subtypes of osteosarcoma further contributes to the consensus, highlighting distinctive ADC_mean values for specific subtypes (22). In addition, proposed ADC cutoff values demonstrate notable sensitivity and specificity when distinguishing between various malignant tumor grades (39). Lastly, ADC values from both whole tumor and darkest tumor areas were effective in differentiating various soft tissue sarcoma subtypes (36). These data collectively support that DWI holds promise as a valuable and non-invasive tool for characterizing and differentiating bone and soft tissue sarcomas.

In evaluating the predictive potential of DWI for chemotherapy responses in osteosarcoma and Ewing's sarcoma, multiple investigations highlight significant differences in ADC values during neoadjuvant chemotherapy for osteosarcoma, but they vary in the specific ADC parameters identified as predictors of treatment response. The observed discrepancies extend to distinctions in pre- and post-treatment ADC values for osteosarcoma and Ewing's sarcoma. Although DWI holds promise in predicting treatment responses through consistent trends in ADC value changes, Teo et al. caution about the limitations, particularly in its added value when compared to conventional MRI alone (24 –28). Nonetheless, DWI seems to prove itself as a clinically used technology for assessing chemotherapy response.

In the assessment of DWI's diagnostic potential in differentiating benign from malignant osseous tumors, quantitative analysis of mean and ADC_min values emerged as a generally effective method for differentiation. However, Douis et al., in contrast to other authors, reported that quantitative analysis was not useful in distinguishing between benign and malignant skeletal tumors (29). Despite this discrepancy, most authors identified specific ADC cutoff values that exhibit high sensitivity and specificity across various scenarios, including primary and metastatic malignancies, bone marrow, and chondrogenic tumors (30 –35,37,38). A similar consensus is reached regarding the role of DWI in differentiating malignant versus benign soft tissue tumors: lower ADC values are associated with more aggressive cellular disease, suggesting that DWI sequences may be useful to establish tumor grade in soft tissue tumors (36 –38).

The present study has some limitations. First, the retrospective nature of the included studies introduces biases and limits the establishment of causal relationships. The heterogeneity observed in how treatments are administered across different studies poses a challenge in drawing generalizable conclusions about DWI's effectiveness. In addition, variations in field strength and acquisition protocols may influence the comparability of results. Furthermore, the continued reliance on histologic diagnosis as the gold standard underscores the limitation that DWI, despite its promising results, is not a surrogate for obtaining a biopsy of suspicious bone and soft tissue tumors. Further, our analysis on the response to chemotherapy (topic 2) was limited to Ewing's sarcoma and osteosarcoma due to the current limitations in the literature. Osteosarcoma and Ewing’s sarcoma are the two most common primary bone tumors; as a result, they are the focus of the majority of studies examining DWI protocols in malignant bone tumors, whereas studies on other malignant bone tumors are lacking. Similarly, another significant limitation of our study is the broad focus of the topics, which may not fully reflect the overall biological behavior of bone and soft tissue tumors. This constraint is primarily due to the paucity of high-quality studies assessing DWI technology in musculoskeletal oncology. Our broad inclusion criteria aimed to encompass all relevant studies analyzing DWI with respect to bone and soft tissue pathologies. However, the limited availability of comprehensive studies hindered our ability to perform a more detailed assessment. These limitations emphasize the need for future prospective studies with standardized methodologies to validate the utility of DWI in musculoskeletal oncology.

In conclusion, DWI is a transformative technological innovation that offers significant utility within orthopedic oncology. By enabling a non-invasive and highly sensitive visualization of tissue cellularity, DWI has allowed further diagnostic accuracy in characterizing bone and soft tissue tumors. For instance, ADC_mean values of ≤1.1 ×10⁻³ mm²/s for bone tumors and ≤1.4 ×10⁻³ mm²/s for soft tissue sarcomas are suggested cutoffs indicating malignancy based on our literature review. However, continued research is needed to fully optimize its role within clinical practice.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Avneesh Chhabra

Robert C Weinschenk

References

Franchi

. Epidemiology and classification of bone tumors. Clin Cases Miner Bone Metab 2012;9:92–95.

Gage

Nagarajan

Ruck

, et al. Sarcomas in the United States: recent trends and a call for improved staging. Oncotarget 2019;10:2462–2474.

Siegel

Miller

Jemal

. Cancer statistics, 2018. CA Cancer J Clin. 2018;68:7–30.

Ponti

Arioli

Longo

, et al. Ultrasound-guided percutaneous bone biopsy: feasibility, diagnostic yield and technical notes. Diagnostics 2023;13:1773.

Meek

Mills

Hanrahan

, et al. Pearls and pitfalls for soft-tissue and bone biopsies: a cross-institutional review. Radiographics 2020;40:266–290.

Espinosa

Jamadar

Jacobson

, et al. CT-guided biopsy of bone: a radiologist’s perspective. AJR Am J Roentgenol 2008;190:W283–W289.

Verma

Cowperthwaite

Burnett

, et al. Differentiating tumor recurrence from treatment necrosis: a review of neuro-oncologic imaging strategies. Neuro-Oncol 2013;15:515–534.

Kumar

Leeds

Fuller

, et al. Malignant gliomas: mR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment. Radiology 2000;217:377–384.

Mullins

Barest

Schaefer

, et al. Radiation necrosis versus glioma recurrence: conventional MR imaging clues to diagnosis. AJNR Am J Neuroradiol 2005;26:1967–1972.

10.

Curnes

Laster

Ball

, et al. MRI Of radiation injury to the brain. AJR Am J Roentgenol 1986;147:119–124.

11.

Brismar

Roberson

Davis

. Radiation necrosis of the brain. Neuroradiological considerations with computed tomography. Neuroradiology 1976;12:109–113.

12.

Mikhael

. Radiation necrosis of the brain: correlation between computed tomography, pathology, and dose distribution. J Comput Assist Tomogr 1978;2:71–80.

13.

van Dellen

Danziger

. Failure of computerized tomography to differentiate between radiation necrosis and cerebral tumour. S Afr Med J 1978;53:171–172.

14.

Kingsley

Kendall

. CT Of the adverse effects of therapeutic radiation of the central nervous system. AJNR Am J Neuroradiol 1981;2:453–460.

15.

Bhojwani

Szpakowski

Partovi

, et al. Diffusion-weighted imaging in musculoskeletal radiology—clinical applications and future directions. Quant Imaging Med Surg 2015;5:740.

16.

Charles-Edwards

de Souza

. Diffusion-weighted magnetic resonance imaging and its application to cancer. Cancer Imaging 2006;6:135–143.

17.

Baliyan

Das

Sharma

, et al. Diffusion-weighted imaging: technique and applications. World J Radiol 2016;8:785–798.

18.

Kuhnke

Langner

Khaw

, et al. Diffusion-weighted MRI – how many B-values are necessary? Rofo 2012;184:303–310.

19.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71.

20.

Whiting

Rutjes

Westwood

, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529–536.

21.

Surov

Nagata

Razek

AAA

, et al. Comparison of ADC values in different malignancies of the skeletal musculature: a multicentric analysis. Skeletal Radiol 2015;44:995–1000.

22.

Zeitoun

Shokry

Ahmed Khaleel

, et al. Osteosarcoma subtypes: magnetic resonance and quantitative diffusion weighted imaging criteria. J Egypt Natl Cancer Inst 2018;30:39–44.

23.

ElDaly

Moustafa

AFI

Abdel-Meguid

SMS

, et al. Can MRI diffusion-weighted imaging identify postoperative residual/recurrent soft-tissue sarcomas? Indian J Radiol Imaging 2018;28:70–77.

24.

Lee

Jee

W-H

Jung

, et al. Prediction of poor responders to neoadjuvant chemotherapy in patients with osteosarcoma: additive value of diffusion-weighted MRI including volumetric analysis to standard MRI at 3T. PLoS One 2020;15:e0229983.

25.

Habre

Dabadie

Loundou

, et al. Diffusion-weighted imaging in differentiating mid-course responders to chemotherapy for long-bone osteosarcoma compared to the histologic response: an update. Pediatr Radiol 2021;51:1714–1723.

26.

Asmar

Saade

Salman

, et al. The value of diffusion weighted imaging and apparent diffusion coefficient in primary osteogenic and Ewing sarcomas for the monitoring of response to treatment: initial experience. Eur J Radiol 2020;124:108855.

27.

Saleh

Abdelrahman

Madney

, et al. Multiparametric MRI with diffusion-weighted imaging in predicting response to chemotherapy in cases of osteosarcoma and Ewing’s sarcoma. Br J Radiol 2020;93:20200257.

28.

Teo

Daescu

Cederberg

, et al. Correlation of histopathology and multi-modal magnetic resonance imaging in childhood osteosarcoma: predicting tumor response to chemotherapy. PLoS One 2022;17:e0259564.

29.

Douis

Davies

Sian

. The role of diffusion-weighted MRI (DWI) in the differentiation of benign from malignant skeletal lesions of the pelvis. Eur J Radiol 2016;85:2262–2268.

30.

Pozzi

Albano

Messina

, et al. Solid bone tumors of the spine: diagnostic performance of apparent diffusion coefficient measured using diffusion-weighted MRI using histology as a reference standard. J Magn Reson Imaging 2018;47:1034–1042.

31.

Padhani

van Ree

Collins

, et al. Assessing the relation between bone marrow signal intensity and apparent diffusion coefficient in diffusion-weighted MRI. AJR Am J Roentgenol 2013;200:163–170.

32.

Wang

Cui

, et al. Role of apparent diffusion coefficients with diffusion-weighted magnetic resonance imaging in differentiating between benign and malignant bone tumors. World J Surg Oncol 2014;12:365.

33.

Setiawati

Suarnata

Rahardjo

, et al. Correlation of quantitative diffusion weighted MR imaging between benign, malignant chondrogenic and malignant non-chondrogenic bone tumors with histopathologic type. Heliyon 2021;7:e06402.

34.

Ahlawat

Khandheria

Subhawong

, et al. Differentiation of benign and malignant skeletal lesions with quantitative diffusion weighted MRI at 3T. Eur J Radiol 2015;84:1091–1097.

35.

Nouh

Doweidar

Khalil

AM-E

. Apparent diffusion coefficient (ADC): a potential in vivo biological surrogate of the incidentally discovered bone lesions at 3T MRI. Eur J Radiol Open 2021;8:100386.

36.

Pekcevik

Kahya

Kaya

. Characterization of soft tissue tumors by diffusion-weighted imaging. Iran J Radiol 2015;12:e15478.

37.

Song

Yoon

Chong

, et al. Diagnostic performance of conventional MRI parameters and apparent diffusion coefficient values in differentiating between benign and malignant soft-tissue tumours. Clin Radiol 2017;72:691.e1–691.e10.

38.

Robba

Chianca

Albano

, et al. Diffusion-weighted imaging for the cellularity assessment and matrix characterization of soft tissue tumour. Radiol Med (Torino) 2017;122:871–879.

39.

Guirguis

Gupta

Thakur

, et al. Osseous–tissue tumor reporting and data system with diffusion–weighted imaging of bone tumors—an interreader analysis and whether it adds incremental value on tumor grading over conventional magnetic resonance imaging. J Comput Assist Tomogr 2023;47:255–263.

40.

Schaefer

Copen

Lev

, et al. Diffusion-weighted imaging in acute stroke. Magn Reson Imaging Clin N Am 2006;14:141–168.

41.

Khoo

MMY

Tyler

Saifuddin

, et al. Diffusion-weighted imaging (DWI) in musculoskeletal MRI: a critical review. Skeletal Radiol 2011;40:665–681.

42.

Baur

Stäbler

Brüning

, et al. Diffusion-weighted MR imaging of bone marrow: differentiation of benign versus pathologic compression fractures. Radiology 1998;207:349–356.