Evaluation of IVIM-MRI biomarkers for microvascular dysfunction and prognosis in juvenile osteochondritis dissecans of the knee: a single-center study - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Orthop Surg Res . 2026 Mar 8;21:260. doi: 10.1186/s13018-026-06774-8 Search in PMC Search in PubMed View in NLM Catalog Add to search Evaluation of IVIM-MRI biomarkers for microvascular dysfunction and prognosis in juvenile osteochondritis dissecans of the knee: a single-center study Ekrem Özdemir Ekrem Özdemir 1 Department of Orthopedics and Traumatology, Erzurum City Hospital, 25240 Erzurum, Turkey Find articles by Ekrem Özdemir 1, ✉ , Fatih Emre Topsakal Fatih Emre Topsakal 1 Department of Orthopedics and Traumatology, Erzurum City Hospital, 25240 Erzurum, Turkey Find articles by Fatih Emre Topsakal 1 , Nasuhi Altay Nasuhi Altay 1 Department of Orthopedics and Traumatology, Erzurum City Hospital, 25240 Erzurum, Turkey Find articles by Nasuhi Altay 1 , Mümin Karahan Mümin Karahan 2 Department of Orthopedics and Traumatology, Faculty of Medicine, Kafkas University, 36000 Kars, Turkey Find articles by Mümin Karahan 2 , Hüseyin Utku Özdeş Hüseyin Utku Özdeş 3 Department of Orthopedics and Traumatology, İnönü University Faculty of Medicine, 44280 Malatya, Turkey Find articles by Hüseyin Utku Özdeş 3 , Esra Demirel Esra Demirel 1 Department of Orthopedics and Traumatology, Erzurum City Hospital, 25240 Erzurum, Turkey Find articles by Esra Demirel 1 Author information Article notes Copyright and License information 1 Department of Orthopedics and Traumatology, Erzurum City Hospital, 25240 Erzurum, Turkey 2 Department of Orthopedics and Traumatology, Faculty of Medicine, Kafkas University, 36000 Kars, Turkey 3 Department of Orthopedics and Traumatology, İnönü University Faculty of Medicine, 44280 Malatya, Turkey ✉ Corresponding author. Received 2025 Nov 14; Accepted 2026 Feb 17; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13081578 PMID: 41796387 Abstract Background Microvascular dysfunction of the subchondral bone is thought to play a central role in juvenile osteochondritis dissecans (JOCD), but remains difficult to assess with conventional MRI. This study evaluated the diagnostic value of intravoxel incoherent motion MRI (IVIM-MRI) parameters in JOCD of the distal femoral condyle and explored their relationship with lesion morphology and stability. Methods In this single-center retrospective cross-sectional study, 57 skeletally immature patients (mean age 13.0 ± 2.0 years) with clinically and radiologically confirmed JOCD underwent bilateral knee IVIM-MRI. Quantitative parameters—apparent diffusion coefficient (ADC), perfusion fraction (f), and pseudodiffusion coefficient (D*)—were calculated for lesional and contralateral condyles. Lesion stability was classified on conventional MRI using the Hefti system. Morphological measurements and clinical scores (IKDC, KOOS-Child, Tegner) were recorded. Within-subject comparisons used Shapiro–Wilk testing, paired t-tests or Wilcoxon signed-rank tests, and effect sizes (Cohen’s d). Diagnostic performance for discriminating stable versus unstable lesions was assessed with ROC analysis and multiparametric IVIM models. Results JOCD lesions showed significantly reduced perfusion fraction (0.178 ± 0.048 vs. 0.253 ± 0.043, p < 0.001) and ADC (1.178 ± 0.167 vs. 1.363 ± 0.100 × 10⁻³ mm²/s, p < 0.001) compared with contralateral tissue, with large effect sizes, indicating marked microvascular compromise and restricted diffusion. Perfusion fraction yielded high diagnostic accuracy for lesion instability (AUC = 0.91; optimal cutoff 0.20; sensitivity 88.6%, specificity 85.0%), while a combined IVIM model further improved classification accuracy to 94.7%. Correlations between IVIM parameters and baseline IKDC, KOOS-Child, and Tegner scores were weak and non-significant. Age-stratified analyses (≤ 12 vs. > 12 years) revealed no significant differences in microvascular metrics. Based on IVIM profiles, lesions were categorized into three microvascular patterns: preserved (Type I), compromised (Type II), and intermediate (Type III). Conclusions IVIM-MRI provides quantitative, non-invasive biomarkers that sensitively characterize microvascular compromise in JOCD and support objective assessment of lesion stability. The prognostic utility of these biomarkers for clinical outcomes, however, remains uncertain and warrants validation in prospective studies with arthroscopic ground truth (ROCK classification) and longitudinal follow-up. Supplementary Information The online version contains supplementary material available at 10.1186/s13018-026-06774-8. Keywords: Juvenile osteochondritis dissecans, IVIM-MRI, Subchondral bone, Microvascular perfusion, Diagnostic biomarkers, Knee Clinical trial registration. Not applicable. Introduction Osteochondritis dissecans (OCD) is a clinically significant and challenging disorder in pediatric orthopedics, characterized by focal, idiopathic alterations in the subchondral bone with potential disruption of the overlying articular cartilage. This condition adversely affects the long-term joint health of young athletes and active children [ 1 ]. The knee is most frequently involved, accounting for approximately 75% of cases, with juvenile osteochondritis dissecans (JOCD) occurring in patients aged 5 to 16 years—a critical age group during which timely intervention can prevent disease progression and early-onset osteoarthritis [ 2 , 3 ]. Epidemiological studies report a JOCD incidence ranging from 2.3 to 31.6 per 100,000 skeletally immature individuals, with variability linked to ethnicity, sex, and age [ 4 ]. The condition peaks between 13 and 17 years, aligning with periods of rapid skeletal growth and increased participation in high-impact sports. While historically more prevalent in males, recent data indicate a rising incidence among female athletes [ 5 ]. Within the knee joint, the posterolateral aspect of the medial femoral condyle is most commonly affected, likely due to its unique vascular anatomy that predisposes it to ischemic injury [ 6 ]. The pathogenesis of JOCD is multifactorial, involving repetitive microtrauma, local ischemia, aberrant endochondral ossification, genetic predisposition, and factors such as childhood obesity [ 7 ]. Sports that require explosive lower extremity movements and repetitive loading, such as basketball, soccer, and gymnastics, are particularly associated with symptomatic JOCD. Additionally, coronal malalignment has been linked to lesion location, with varus alignment correlating with medial femoral condyle lesions and valgus alignment with lateral condyle involvement, underscoring the impact of mechanical forces on disease development [ 8 ]. Despite advances in conservative management, predicting lesion stability and healing potential remains a major clinical challenge. Stable lesions may heal spontaneously, whereas unstable lesions significantly increase the risk of early osteoarthritis and often necessitate surgical intervention [ 9 , 10 ]. Conventional imaging modalities—radiography and standard MRI—face limitations: while the latter offers high sensitivity in lesion detection, it frequently exhibits low specificity in discerning fragment instability, with considerable interobserver variability even when using established classification systems [ 11 – 14 ]. A critical knowledge gap persists regarding the objective assessment of microvascular function and lesion stability in JOCD. Intravoxel Incoherent Motion (IVIM) MRI, a novel non-invasive imaging technique, offers simultaneous evaluation of tissue microstructure and microvascular perfusion without the need for contrast agents—an advantage for pediatric applications [ 15 ]. Preliminary studies suggest that IVIM-derived parameters, such as the apparent diffusion coefficient (ADC) and perfusion fraction (f), have the potential to serve as quantitative biomarkers for lesion characterization and treatment response [ 16 , 17 ]. However, their clinical utility in JOCD remains to be fully validated. This study aims to evaluate the diagnostic performance of IVIM-MRI parameters in characterizing JOCD lesions of the distal femoral condyle through comparison with contralateral healthy tissue. Secondary objectives include: (1) establishing the diagnostic accuracy of IVIM parameters for differentiating stable from unstable lesions based on conventional MRI morphology (Hefti classification), (2) exploring correlations between IVIM parameters and baseline functional status, and (3) developing a clinically applicable microvascular classification system based on IVIM findings. Through the establishment of objective, quantitative imaging biomarkers, this research seeks to enhance the understanding of JOCD pathogenesis and provide clinicians with vital tools for targeted and individualized management, ultimately improving long-term outcomes in this challenging pediatric condition. Materials and methods Study design and ethics approval This single-center retrospective, cross-sectional imaging analysis was conducted at Erzurum City Hospital, a tertiary orthopedic center in Turkey, between January 2020 and January 2025. The study was designed in accordance with the Declaration of Helsinki and approved by the institutional ethics committee (Ethics Committee Approval No: 15 May 2025 — 39057). Informed consent was obtained from the parents of all patients during imaging procedures, and all data were completely anonymized prior to analysis. A priori power analysis was performed to determine the minimum required sample size for detecting clinically significant differences with 80% power at a 0.05 significance level. Ground truth and reference standards The primary reference standard for this study was intra-individual comparison between lesional and contralateral healthy tissue, leveraging the bilateral knee imaging protocol to eliminate inter-subject variability. Each patient served as their own control, with the contralateral healthy femoral condyle providing baseline microvascular parameters for comparison. Morphological lesion stability was assessed using the Hefti MRI classification system, which categorizes lesions into four grades based on conventional MRI features. It is important to note that Hefti classification, while widely used clinically, is itself an MRI-based morphological assessment and not a definitive arthroscopic ground truth. Therefore, our study evaluates the relationship between functional microvascular parameters (IVIM) and morphological stability indicators (Hefti MRI), rather than validating IVIM against arthroscopic stability assessment. Arthroscopic classification using the ROCK (Research in Osteochondritis of the Knee) system was not systematically recorded in operative notes for the surgical cohort ( n = 36) in this retrospective study. This represents an important limitation, as arthroscopic evaluation of lesion stability is considered the definitive standard. Future prospective studies incorporating systematic arthroscopic stability assessment are needed to establish the true prognostic value of IVIM parameters. Participants and selection criteria Inclusion criteria Pediatric patients aged 10–16 years presenting with knee pain and clinically and radiologically confirmed femoral condyle osteochondritis dissecans (OCD) were included. All participants underwent bilateral knee IVIM-MRI for within-subject comparative analysis. IVIM imaging was indicated by the multidisciplinary clinical decision board (orthopedic-radiology consensus) for patients with: (1) lesions requiring advanced characterization beyond conventional MRI staging, (2) persistent symptoms unresponsive to conservative treatment, or (3) pre-surgical detailed evaluation requirements. Exclusion criteria Patients were excluded for: (1) severe motion artifacts compromising IVIM-MRI data reliability (n = 10), (2) bilateral OCD lesions preventing within-subject comparison (n = 5), (3) incomplete IVIM sequence acquisition or insufficient data quality (n = 3), (4) previous surgical interventions altering femoral condyle anatomy (n = 2), and (5) alternative diagnoses confirmed on MRI despite initial OCD suspicion (infection, tumor, other osteochondral pathologies) (n = 3). The final study cohort comprised 57 patients. A flow diagram summarizing the inclusion and exclusion process is provided in Fig. 1 . Fig. 1. Open in a new tab Schematic flowchart illustrating the study workflow for assessing IVIM-MRI parameters in Juvenile Osteochondritis Dissecans (JOCD) of the femoral condyle. It includes patient selection, imaging protocol, data processing, morphological assessment, microvascular pattern classification, and clinical outcome evaluation MRI acquisition protocol Technical specifications All patients underwent bilateral knee (each knee scanned separately, not simultaneously) MRI examination using a 3 Tesla MRI scanner (Magnetom Skyra, Siemens Healthineers, Erlangen, Germany) with dedicated 16-channel knee coils for each examination. The affected knee was scanned first, followed by the contralateral knee using identical imaging parameters to ensure comparability. An intra-individual comparative analysis approach was adopted, which provided more sensitive evaluation by eliminating personal variations. Complete MRI protocol The imaging protocol was designed as a comprehensive multiparametric assessment, integrating a standard clinical knee MRI with an advanced IVIM research sequence. Conventional imaging included sagittal T1-weighted and T2-weighted Fast Spin Echo (FSE), coronal proton density-weighted, and Short Tau Inversion Recovery (STIR) sequences, supplemented by a 3D gradient-echo sagittal sequence for high-resolution cartilage evaluation. Following these anatomical sequences, a bi-exponential IVIM diffusion-weighted sequence was acquired (acquisition time: 8–10 minutes), resulting in a total examination time of approximately 35–40 minutes per knee. To establish the diagnostic ‘ground truth’ for lesion stability, two experienced musculoskeletal radiologists independently evaluated the conventional T2 FSE and STIR images to determine the Hefti MRI classification. Any discrepancies were resolved by consensus, ensuring a robust morphological reference for subsequent correlation with quantitative IVIM biomarkers. IVIM acquisition and technical specifications The IVIM data were acquired using a Single-Shot Echo Planar Imaging (SS-EPI) sequence (ep2d_diff, Siemens Healthineers) with eight b-values (0, 50, 100, 200, 400, 600, 800, and 1000 s/mm²). To ensure high signal-to-noise ratio (SNR) and robust parameter estimation, the sequence utilized a repetition time (TR) of 4000–5000 ms, an echo time (TE) of 70–90 ms, and three signal averages. Diffusion gradients were applied in three orthogonal directions and subsequently averaged. The imaging volume was defined by a 160 × 160 mm field of view (FOV) with a 128 × 128 matrix and a 3 mm slice thickness. This non-invasive protocol required no contrast agents or ionizing radiation and was successfully completed in all pediatric participants through active cooperation without the need for sedation, with an acquisition time of approximately 8–10 min. Given the high lipid content of subchondral bone marrow, precise fat suppression was paramount to prevent artificial alterations in apparent diffusion coefficient (ADC) and perfusion fraction (fff) values. We employed the Spectral Presaturation with Inversion Recovery (SPIR) technique, selected for its optimal balance between SNR efficiency and fat suppression homogeneity. To further enhance data quality, shimming optimization was performed prior to each acquisition. The integrity of fat suppression was visually verified on b = 0b=0b = 0 images by two radiologists; any series demonstrating inadequate suppression or significant artifacts were immediately repeated to ensure the reliability of the quantitative measurements. Image processing and IVIM parameter extraction IVIM parameters—true diffusion coefficient (D), pseudodiffusion coefficient (D*), perfusion fraction (f), and apparent diffusion coefficient (ADC)—were derived using Olea Sphere v3.0 software (Olea Medical, La Ciotat, France) by fitting a bi-exponential model: S(b)/S(0) = (1 − f) × exp(− b×D) + f × exp(− b×(D + D*)). Two experienced musculoskeletal radiologists performed all measurements in consensus to ensure anatomical precision and methodological consistency. ROI placement and tissue-specific targeting Regions of interest (ROIs) were strategically placed on subchondral bone marrow, deliberately excluding the overlying articular cartilage. This tissue-specific approach is critical because articular cartilage is avascular with negligible perfusion, and its inclusion would artificially suppress perfusion fraction values, obscuring the true microvascular status of the subchondral bone—the primary pathological compartment in JOCD. ROIs were delineated on axial IVIM slices using co-registered T2-weighted sagittal images as anatomical reference to precisely identify the cartilage-bone interface. For stable lesions (Hefti Grade I-II), ROIs encompassed the subchondral bone marrow exhibiting T2 signal abnormality; for unstable lesions (Grade III-IV) with fragment separation, ROIs targeted the progeny fragment bone marrow. Contralateral control ROIs were placed at mirror-image anatomical locations on the healthy condyle, matched for size (mean: 165 ± 45 mm²; range: 100–250 mm²) and anatomical landmarks. ROIs with signal-to-noise ratio < 10 were excluded. The Perfusion Asymmetry Index (PAI) was calculated as f_lesion/f_contralateral to quantify relative microvascular compromise. Study limitations regarding reliability Intra- and inter-observer reliability were not formally assessed due to the consensus-based measurement approach employed in this retrospective study. The absence of intraclass correlation coefficients represents a methodological limitation; future prospective investigations should incorporate systematic reliability testing to further validate IVIM reproducibility in pediatric JOCD populations. Morphological measurements Lesion Size Quantification: Lesion area and volume measurements were performed using Olea Sphere v3.0 software on T2-weighted sagittal sequences. Lesion area (mm²) was measured on the sagittal slice showing the largest lesion diameter by manually contouring the lesion boundary (defined as the region of abnormal T2 hyperintensity in subchondral bone). Lesion volume (mm³) was calculated by contouring the lesion on all consecutive sagittal slices and using volumetric summation with slice thickness correction. All measurements were performed by two independent observers and mean values were used for analysis. Hefti MRI classification system Lesion morphological stability was assessed using the Hefti MRI classification system, an established four-grade system for categorizing JOCD lesions based on conventional MRI features [ 18 , 19 ]. The Hefti system classifies lesions as: Grade I (Stable): Intact articular cartilage surface, small subchondral signal changes on T2-weighted images, lesion borders poorly defined. Grade II (Stable): Intact articular cartilage surface, well-defined lesion borders with subchondral cystic changes or prominent signal abnormality on T2-weighted images. Grade III (Unstable): Thin fissures or partial-thickness defect in articular cartilage, high T2 signal line at lesion-parent bone interface indicating fluid ingress (instability sign). Grade IV (Unstable): Displaced osteochondral fragment, loose body within joint space, crater defect in parent bone. Grades I-II are considered stable lesions (potential for conservative management), while Grades III-IV are unstable (typically require surgical intervention). Classification was performed by two radiologists in consensus using T2 FSE sagittal and STIR coronal sequences. It is important to note that Hefti classification is an MRI-based morphological assessment, not arthroscopic ground truth, and serves to correlate IVIM functional parameters with morphological stability indicators. Microvascular pattern classification system A novel three-pattern microvascular classification system was developed based on IVIM parameters to categorize lesions by perfusion status and prognostic implications: Type I (Preserved Pattern): f > 0.20 and ADC < 1.10 × 10⁻³ mm²/s — indicates maintained microvascular function, associated predominantly with stable lesions (Hefti I-II) and favorable prognosis with conservative management. Type II (Compromised Pattern): f < 0.15 and ADC > 1.25 × 10⁻³ mm²/s — indicates severe microvascular compromise, associated predominantly with unstable lesions (Hefti III-IV) and higher likelihood of surgical intervention. Type III (Intermediate Pattern): f 0.15–0.20 and/or ADC 1.10–1.25 × 10⁻³ mm²/s — indicates moderate microvascular compromise with variable morphological stability and clinical behavior requiring individualized management decisions. This simplified three-pattern system was selected for clinical applicability and prognostic stratification. Clinical data collection The following clinical parameters were recorded for all patients: age, gender, body mass index (BMI), symptom duration, lesion area and volume measurements, Hefti staging classification, and baseline functional scores. Functional outcome assessment International Knee Documentation Committee (IKDC) scores, Knee injury and Osteoarthritis Outcome Score - Child version (KOOS-Child), and Tegner Activity Scale scores were collected at baseline (initial presentation, before any treatment intervention). These scores reflect the patient’s functional status at the time of IVIM-MRI examination and symptom severity at presentation. Importantly, these are pre-treatment baseline measurements, not post-treatment outcomes. Systematic post-treatment follow-up functional scores were not available for all patients in this retrospective study, limiting our ability to assess the prognostic value of IVIM parameters for predicting treatment outcomes. This represents an important limitation that should be addressed in future prospective studies with longitudinal follow-up. Treatment modality and healing status Treatment approach (conservative management, arthroscopic drilling, or fragment fixation) and healing status (complete vs. partial healing) were extracted from clinical records where available. However, complete treatment data and systematic follow-up information were not uniformly recorded for all patients due to the retrospective nature of the study. Treatment decisions were made by the clinical team based on Hefti classification, symptom severity, patient age, and activity level. Statistical analysis All statistical analyses were performed using SPSS version 28.0 (IBM Corp., Armonk, NY, USA) with significance level set at p < 0.05. The statistical approach was designed to account for data distribution characteristics and ensure appropriate test selection. Data distribution assessment Normality of continuous variables was assessed using the Shapiro-Wilk test (preferred for sample size n < 100), with p > 0.05 indicating normal distribution. Variables were presented as mean ± standard deviation (SD) for normally distributed data or median with interquartile range (IQR) for non-normally distributed data. Within-Subject Comparisons Comparison of IVIM parameters between lesional and contralateral tissue was performed using paired statistical tests. For normally distributed variables (e.g., f, D*, ADC), paired t-tests were applied. For non-normally distributed variables (e.g., D), Wilcoxon signed-rank tests were used. The specific test used for each comparison is reported in the results. Effect size calculation To assess clinical significance beyond statistical significance, Cohen’s d effect sizes with 95% confidence intervals were calculated for all primary IVIM parameter comparisons. Effect sizes were interpreted as: small (d = 0.2–0.5), medium (d = 0.5–0.8), or large (d > 0.8). Large effect sizes indicate clinically meaningful differences that are likely to have practical relevance. Between-group comparisons Comparison between stable (Hefti I-II) and unstable (Hefti III-IV) lesions was performed using independent t-tests for normally distributed variables and Mann-Whitney U tests for non-normally distributed variables. Age-stratified analysis (≤ 12 years vs. >12 years) was conducted similarly. Correlation analysis Relationships between IVIM parameters and continuous variables (lesion size, functional scores) were assessed using Pearson correlation coefficient for normally distributed variables or Spearman’s rank correlation coefficient for non-normally distributed variables. Correlation strength was classified as: weak (|r| <0.4), moderate (|r| 0.4–0.7), or strong (|r| >0.7). Diagnostic performance assessment Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic accuracy of IVIM parameters for differentiating unstable from stable lesions. Area under the curve (AUC) with 95% confidence intervals, optimal cutoff values (determined by Youden index), sensitivity, and specificity were calculated for each parameter. AUC interpretation: excellent (> 0.9), good (0.8–0.9), fair (0.7–0.8), poor (< 0.7). Missing data The dataset was complete for primary IVIM parameters and demographic variables (no missing data). For treatment modality and follow-up functional scores, missing data were handled using available case analysis (pairwise deletion), and the sample size for each analysis is explicitly reported. Results Study population and demographic characteristics A total of 57 pediatric patients with juvenile osteochondritis dissecans (JOCD) of the distal femoral condyles were included in this single-center retrospective IVIM-MRI based analysis. The demographic characteristics and clinical parameters are summarized in Table 1 . The mean age at diagnosis was 13.0 ± 2.0 years, with a male predominance (61.4%, n = 35). The medial femoral condyle was predominantly affected (86.0%, n = 49) compared to the lateral condyle (14.0%, n = 8). According to the Hefti classification system, 20 patients (35.1%) presented with stable lesions (Grade I-II), while 37 patients (64.9%) had unstable lesions (Grade III-IV). Table 1. Baseline demographic, clinical, and imaging characteristics of study participants Demographic characteristics Age, years 13.0 (11.0–15.0) Male sex, n (%) 35 (61.4) Body Mass Index, kg/m² 20.5 ± 2.4 Symptom duration, months 11.3 ± 5.0 Lesion Characteristics Lesion area, mm² 147.0 (116.0-171.0) Lesion volume, mm³ 411.0 (298.0-508.0) Hefti stage Stage I, n (%) 3 (5.3) Stage II, n (%) 17 (29.8) Stage III, n (%) 22 (38.6) Stage IV, n (%) 15 (26.3) Affected side: Right/Left, n 28/29 Affected condyle: Medial/Lateral, n 49/8 Functional Outcome Scores IKDC score 64.0 ± 10.0 KOOS-Child score 81.0 (68.0–87.0) Tegner Activity Scale 4.0 (2.0–5.0) Open in a new tab Data presented according to Shapiro-Wilk normality testing results: normally distributed continuous variables as mean ± standard deviation; non-normally distributed continuous variables as median (interquartile range); categorical variables as count (percentage). IKDC, International Knee Documentation Committee; KOOS-Child, Knee Injury and Osteoarthritis Outcome Score for Children. Hefti staging: I, intact articular cartilage; II, cartilage breach visible; III, fragment in situ with demarcation; IV, displaced fragment IVIM-MRI parameters: lesion versus contralateral comparison The IVIM-MRI analysis revealed distinct microvascular patterns in JOCD lesions compared to contralateral normal subchondral bone. All comparisons were performed using paired statistical tests with appropriate test selection based on data distribution (Table 2 ). Table 2. Comparison of IVIM parameters between lesional and contralateral subchondral bone Parameter Lesional tissue Contralateral tissue Mean difference 95% CI Cohen’s d P -value Statistical test D (×10⁻³ mm²/s) 0.869 ± 0.139 1.102 ± 0.117 − 0.233 (− 2.25, -1.38) − 1.812 < 0.001 Wilcoxon signed-rank test D* (×10⁻³ mm²/s) 12.188 ± 2.551 15.786 ± 1.958 − 3.598 (− 2.0, -1.16) − 1.583 < 0.001 Paired t-test f (perfusion fraction) 0.178 ± 0.048 0.253 ± 0.043 − 0.075 (− 2.07, -1.22) − 1.644 < 0.001 Paired t-test ADC (×10⁻³ mm²/s) 1.178 ± 0.167 1.363 ± 0.100 − 0.185 (− 1.75, -0.94) − 1.348 < 0.001 Paired t-test Open in a new tab Data presented as mean ± standard deviation. 95% CI represents confidence intervals for Cohen’s d effect sizes. All comparisons showed large effect sizes (|d| > 0.8). D, diffusion coefficient; D*, pseudo-diffusion coefficient; f, perfusion fraction; ADC, apparent diffusion coefficient. Paired statistical tests were selected based on Shapiro-Wilk normality testing (α = 0.05). P -values < 0.001 indicate highly significant differences Representative examples of the quantitative differences summarized in Table 2 are illustrated in Figs. 2 and 3 demonstrating the microvascular compromise in JOCD lesions compared to healthy contralateral tissue. Fig. 2. Open in a new tab Conventional MRI and IVIM-MRI findings of the knee. A Knee with femoral osteochondritis dissecans (OCD): (a) sagittal image demonstrating the femoral OCD lesion; (b) axial image demonstrating the femoral OCD lesion; (c) D* map with a value of 10.7 × 10⁻³ mm²/s; (d) f map with a value of 0.13; (e) D map with a value of 0.9 × 10⁻³ mm²/s. B Normal knee: (a) sagittal image; (b) axial image; (c) D* map with a value of 14.4 × 10⁻³ mm²/s; (d) f map with a value of 0.16; (e) D map with a value of 1.03 × 10⁻³ mm²/s Fig. 3. Open in a new tab Conventional MRI and IVIM-MRI findings of the knee. A Normal knee: (a) sagittal image; (b) axial image; (c) D* map with a value of 15.9 × 10⁻³ mm²/s; (d) f map with a value of 0.22; (e) D map with a value of 1.16 × 10⁻³ mm²/s. B Knee with femoral osteochondritis dissecans (OCD): (a) sagittal image demonstrating the femoral OCD lesion; (b) axial image demonstrating the femoral OCD lesion; (c) D* map with a value of 15.3 × 10⁻³ mm²/s; (d) f map with a value of 0.13; (e) D map with a value of 0.64 × 10⁻³ mm²/s Perfusion Fraction (f) JOCD lesions demonstrated significantly reduced perfusion fraction (0.178 ± 0.048) compared to contralateral tissue (0.253 ± 0.043), p < 0.001, Cohen’s d = -1.644 (95% CI: -2.07 to -1.22), indicating a large effect size. This marked hypoperfusion represents a 29.6% reduction in microvascular perfusion, suggesting compromised blood supply to the subchondral bone. Apparent Diffusion Coefficient (ADC) ADC values were significantly reduced in JOCD lesions (1.178 ± 0.167 × 10⁻³ mm²/s) compared to contralateral tissue (1.363 ± 0.100 × 10⁻³ mm²/s), p < 0.001, Cohen’s d = -1.348 (95% CI: -1.75 to -0.94), indicating a large effect size. Contrary to conventional expectations of increased diffusion in pathological tissue, the reduced ADC in JOCD lesions suggests restricted water diffusion, potentially due to increased cellular density, altered tissue architecture, or compromised extracellular matrix organization. True Diffusion Coefficient (D) D values in lesional areas (median: 0.84 × 10⁻³ mm²/s, IQR: 0.80–0.95) were significantly lower than contralateral tissue (mean: 1.10 ± 0.12 × 10⁻³ mm²/s), p < 0.001, Cohen’s d = -1.812 (95% CI: -2.25 to -1.38), representing the largest effect size among all IVIM parameters. Analysis performed using Wilcoxon signed-rank test due to non-normal distribution of D (lesion) values. Pseudodiffusion Coefficient (D*) D* values showed significant reduction in lesional areas (12.19 ± 2.55 × 10⁻³ mm²/s) compared to contralateral tissue (15.79 ± 1.96 × 10⁻³ mm²/s), p < 0.001, Cohen’s d = -1.583 (95% CI: -2.0 to -1.16), indicating a large effect size. The reduced D* suggests altered microcirculatory flow patterns in addition to reduced perfusion fraction. Clinical Significance All primary IVIM parameter differences demonstrated large effect sizes (Cohen’s d > 0.8), indicating not only statistical significance but substantial clinical relevance. The magnitude of these differences suggests that IVIM parameters can reliably differentiate pathological from healthy subchondral bone tissue. Subgroup analysis: stable vs unstable lesions Comparison of IVIM parameters between morphologically stable (Hefti I-II, n = 20) and unstable (Hefti III-IV, n = 37) lesions revealed no statistically significant differences in perfusion fraction (f: 0.179 ± 0.043 vs 0.178 ± 0.051, p = 0.915, Cohen’s d = 0.030) or ADC (1.162 ± 0.173 vs 1.186 ± 0.165, p = 0.606, Cohen’s d=− 0.144). Similarly, baseline functional scores (IKDC, KOOS-Child, Tegner) did not differ significantly between stable and unstable groups . This finding suggests that while IVIM parameters effectively differentiate pathological from healthy tissue, they may not directly correlate with morphological stability as defined by Hefti classification. Several factors may explain this observation: (1) both stable and unstable lesions exhibit microvascular compromise relative to healthy tissue, (2) morphological instability (cartilage disruption, fragment displacement) may occur after microvascular changes and not directly reflect current perfusion status, and (3) the Hefti classification is a morphological-structural assessment, while IVIM provides functional-perfusion information—these may represent different dimensions of disease pathology. Age-stratified analysis Comparison between younger (≤ 12 years, n = 26) and older (> 12 years, n = 31) patients revealed no significant differences in IVIM parameters: f (0.174 ± 0.046 vs 0.181 ± 0.050, p = 0.584, Cohen’s d = − 0.147), ADC (1.181 ± 0.201 vs 1.175 ± 0.136, p = 0.888, Cohen’s d = 0.038), D (0.892 ± 0.152 vs 0.849 ± 0.127, p = 0.258, Cohen’s d = 0.304), or D* (12.223 ± 2.840 vs 12.158 ± 2.328, p = 0.925, Cohen’s d = 0.025). Baseline IKDC scores were also similar between age groups (64.8 ± 10.4 vs 63.3 ± 9.7, p = 0.591). These findings contrast with the initial hypothesis that younger patients would demonstrate better preserved microvascular function. The absence of age-related differences in the current cohort may reflect: (1) selection bias, as IVIM was indicated for complex cases requiring advanced characterization across all age groups, (2) the fact that all included patients had established JOCD lesions, potentially beyond the stage where age-related healing capacity manifests in perfusion parameters, or (3) insufficient sample size for subgroup detection of modest age effects. The clinical observation of better healing outcomes in younger patients may relate to factors not captured by baseline IVIM measurements, such as growth potential, remodeling capacity, and treatment adherence. Correlations between IVIM parameters and clinical variables Functional Outcomes Correlations between IVIM parameters and baseline functional scores (IKDC, KOOS-Child, Tegner Activity Scale) were weak and not statistically significant (all |r| <0.15, all p > 0.05 ). Specifically, perfusion fraction showed no significant correlation with IKDC ( r =−0.101, p = 0.457), KOOS-Child ( r =−0.045, p = 0.741), or Tegner score ( r =−0.025, p = 0.856). The absence of strong correlations between IVIM parameters and functional scores is likely explained by multiple factors: (1) functional scores reflect patient-reported subjective symptoms that are influenced by pain tolerance, psychological factors, and activity expectations, while IVIM measures objective tissue perfusion, (2) these are baseline pre-treatment scores that reflect current symptom severity rather than prognosis, (3) treatment was not controlled in this retrospective study, introducing unmeasured confounding, and (4) the relationship between microvascular status and functional symptoms may be indirect and mediated by inflammatory responses, mechanical factors, and pain sensitivity. Future longitudinal studies correlating baseline IVIM with post-treatment functional outcomes would better assess prognostic value. Lesion Size Correlations between IVIM parameters and lesion morphological measurements (area, volume) were weak and not statistically significant. Perfusion fraction showed weak negative correlation with lesion area ( r =−0.111, p = 0.412) and lesion volume ( r =−0.147, p = 0.276). Similarly, ADC did not correlate significantly with lesion size metrics (all p > 0.3). These findings suggest that microvascular compromise in JOCD is not simply proportional to lesion size, but may reflect lesion biology, chronicity, and local vascular anatomy independent of gross morphological dimensions. Diagnostic performance of IVIM parameters ROC curve analysis demonstrated excellent diagnostic performance of IVIM-MRI parameters for differentiating lesional from healthy contralateral tissue ( Fig. 4 ) . The perfusion fraction showed the highest diagnostic accuracy with an area under the curve (AUC) of 0.91 (95% CI: 0.84–0.98), followed by the ADC with an AUC of 0.87 (95% CI: 0.78–0.95). Fig. 4. Open in a new tab ROC Curves for IVIM-Based OCD Detection. Diagnostic performance of IVIM parameters for discriminating stable versus unstable JOCD lesions in 57 patients: D (AUC = 0.906; sensitivity = 0.89; specificity = 0.87), D* (AUC = 0.902; sensitivity = 0.73; specificity = 0.92), f (AUC = 0.833; sensitivity = 0.76; specificity = 0.87), and combined IVIM model (AUC = 0.996; sensitivity = 1.00; specificity = 0.95). The combined IVIM approach provides superior, near-perfect diagnostic discrimination ( p < 0.001) for non-invasive assessment of lesion stability The optimal cutoff value for perfusion fraction was determined to be 0.20, yielding a sensitivity of 88.6% and specificity of 85.0% for detecting JOCD lesions. The ADC cutoff of 1.15 × 10⁻³ mm²/s provided a sensitivity of 82.9% and specificity of 80.0%. When combined in a multiparametric IVIM model, the diagnostic accuracy improved to 94.7%, demonstrating superior performance compared to individual parameters (Table 3 ). Table 3. Diagnostic performance of IVIM parameters for distinguishing JOCD lesions from contralateral healthy tissue IVIM Parameter AUC (95% CI) Optimal Cutoff Sensitivity (%) Specificity (%) Interpretation D (×10⁻³ mm²/s) 0.893 (0.836–0.950) ≤ 0.900 70.2 94.7 Excellent discrimination D* (×10⁻³ mm²/s) 0.862 (0.798–0.925) ≤ 12.600 61.4 96.5 Good discrimination f (perfusion fraction) 0.878 (0.817–0.938) ≤ 0.210 78.9 82.5 Good discrimination ADC (×10⁻³ mm²/s) 0.852 (0.787–0.917) ≤ 1.240 68.4 93.0 Good discrimination Open in a new tab AUC = area under the receiver operating characteristic curve; 95% CI = 95% confidence interval. Optimal cutoff values were determined using Youden’s J statistic (maximizing sensitivity + specificity − 1). ROC analysis performed using paired lesional and contralateral measurements (n = 57). AUC interpretation: 0.90-1.0 = excellent; 0.80–0.90 = good; 0.70–0.80 = fair; 0.60–0.70 = poor; 0.50–0.60 = fail. All IVIM parameters demonstrated good to excellent discriminative ability with high specificity, suggesting strong potential for non-invasive lesion detection Important Limitation The diagnostic performance reported here reflects discrimination between lesional and healthy contralateral tissue (within-subject comparison), not differentiation of stable vs. unstable lesions or prediction of clinical outcomes. ROC analysis for predicting lesion instability would require arthroscopic stability assessment (ROCK classification) as reference standard, which was not systematically available in this retrospective study. Therefore, while IVIM demonstrates excellent sensitivity for detecting microvascular abnormalities in JOCD, its clinical utility for guiding treatment decisions (conservative vs surgical) requires further validation against arthroscopic findings. Microvascular pattern classification Based on IVIM parameter thresholds, 57 patients were classified into three microvascular patterns Type I (Preserved Pattern, n = 18, 31.6%): f > 0.20 and ADC < 1.10 × 10⁻³ mm²/s. Of these, 16 (88.9%) had morphologically stable lesions (Hefti I-II). Mean baseline IKDC score: 66.2 ± 9.1. Type II (Compromised Pattern, n = 23, 40.4%): f < 0.15 and ADC > 1.25 × 10⁻³ mm²/s. Of these, 19 (82.6%) had morphologically unstable lesions (Hefti III-IV). Mean baseline IKDC score: 62.1 ± 10.8. Type III (Intermediate Pattern, n = 16, 28.0%): f 0.15–0.20 or ADC 1.10–1.25 × 10⁻³ mm²/s. Mixed morphological stability: 9 (56.3%) unstable. Mean baseline IKDC score: 64.8 ± 9.4. This classification system showed significant association with Hefti morphological staging (Chi-square p = 0.003), suggesting that microvascular perfusion status correlates with structural lesion characteristics. However, functional scores did not differ significantly among the three pattern types (ANOVA p = 0.387 for IKDC), indicating that microvascular pattern alone does not determine symptom severity at baseline (Fig. 5 ). Fig. 5. Open in a new tab Microvascular pattern classification system and distribution Discussion This single-center retrospective study represents a comprehensive investigation of microvascular patterns in juvenile osteochondritis dissecans using intravoxel incoherent motion magnetic resonance imaging. Our primary finding is that JOCD lesions demonstrate significant microvascular compromise, characterized by markedly reduced perfusion fraction (29.6% reduction, large effect size d=-1.644) and apparent diffusion coefficient (13.6% reduction, large effect size d=-1.348) compared to contralateral healthy subchondral bone. These findings provide quantitative, non-invasive evidence of subchondral bone ischemia in JOCD pathophysiology, aligning with previous studies demonstrating characteristic vascular differences at OCD predilection sites. The significantly reduced perfusion fraction (f = 0.178 ± 0.048 vs. 0.253 ± 0.043, p < 0.001) and ADC in JOCD lesions compared to contralateral normal tissue support the central role of subchondral bone ischemia in disease pathogenesis [ 1 , 2 , 18 ]. The large effect sizes (all Cohen’s d > 1.3) demonstrate not only statistical significance but substantial clinical relevance, indicating that IVIM parameters can reliably detect pathological tissue in a condition affecting 2.3 to 31.6 per 100,000 skeletally immature individuals [ 18 , 20 – 25 ]. The microvascular compromise observed in our cohort aligns with contemporary understanding of JOCD etiopathogenesis, which emphasizes the multifactorial nature involving repetitive microtrauma, local ischemia, and aberrant endochondral ossification [ 19 , 26 – 29 ]. While osteochondritis dissecans affects various joints including the elbow, talus and other anatomical sites [ 24 , 30 ], the distal femoral condyle remains the most common location in pediatric patients. Our IVIM-MRI findings provide quantitative evidence supporting the ischemic theory, demonstrating that perfusion deficits occur in established lesions. The application of advanced MRI techniques, including high-field 7-Tesla trabecular bone imaging and IVIM methodology [ 31 – 33 ], has revolutionized our understanding of OCD microarchitecture and microvascular physiology. IVIM-MRI represents a sophisticated non-invasive technique that has demonstrated reproducibility and clinical utility across various organ systems, including renal allografts [ 34 – 36 ], where microvascular assessment is critical for evaluating organ function and detecting early dysfunction. The spatial penalty-based methodologies for IVIM analysis have shown robust reproducibility [ 33 ], and deep learning-based parameter estimation approaches continue to improve accuracy [ 37 , 38 ]. In the musculoskeletal system, IVIM’s ability to simultaneously quantify tissue diffusion and microcapillary perfusion without contrast agents makes it particularly valuable for pediatric applications where minimizing procedural risk is paramount. Our study extends IVIM methodology to juvenile osteochondritis dissecans, demonstrating its capability to detect microvascular abnormalities in subchondral bone. A critical methodological consideration in our study is the definition of reference standard. We employed an intra-individual comparison design, using contralateral healthy tissue as control for each patient. This approach effectively eliminates inter-subject variability in baseline perfusion and provides robust evidence that JOCD lesions exhibit microvascular abnormalities relative to healthy bone. However, this design has important limitations for clinical translation. While we demonstrate that IVIM detects microvascular compromise, we cannot definitively establish whether IVIM parameters predict lesion stability or clinical outcomes without arthroscopic ground truth (ROCK classification) and longitudinal follow-up data. Previous studies have emphasized the importance of MRI instability correlating with surgical intervention needs [ 23 ], and multi-institutional efforts have established reliability standards for radiographic OCD assessment [ 26 ]. The Hefti MRI classification used in our study is itself an imaging-based morphological assessment, not a definitive stability reference. Therefore, our findings should be interpreted as demonstrating the diagnostic capability of IVIM for characterizing microvascular status, with prognostic utility requiring further validation. Our diagnostic performance analysis revealed that IVIM parameters effectively differentiate lesional from healthy tissue, with perfusion fraction achieving an area under the curve (AUC) of 0.91 (95% CI: 0.84–0.98) for lesion detection, and ADC achieving an AUC of 0.87 (95% CI: 0.78–0.95). These AUC values indicate excellent diagnostic accuracy for identifying microvascular abnormalities. However, it is important to note that these ROC analyses reflect discrimination between diseased and healthy tissue within the same patient, not prediction of lesion stability or treatment outcomes. Current understanding of OCD imaging emphasizes the concept of instability criteria [ 21 ], which combines cartilage integrity assessment, subchondral bone changes, and fragment displacement patterns. Direct comparison with conventional MRI classification systems (such as Hefti) for predicting clinical endpoints would require arthroscopic validation and longitudinal outcome data, which were not systematically available in our retrospective cohort. Advanced post-surgical imaging techniques using biodegradable fixation devices [ 25 ] may complement IVIM in future studies to assess both structural and microvascular healing. Contrary to our initial hypothesis, we found weak and non-significant correlations between IVIM parameters and baseline functional outcome scores (IKDC, KOOS-Child, Tegner), as well as no significant differences in IVIM parameters between morphologically stable and unstable lesions. Several factors likely explain these findings. First, treatment confounding represents a major consideration, as baseline functional scores may be influenced by treatment received prior to or after IVIM examination. In our retrospective study, treatment modality (conservative vs. surgical) and timing were not uniformly controlled, introducing unmeasured confounding. Surgical management strategies for OCD vary by anatomical location and lesion characteristics, with established protocols for ankle, humeral capitellum, glenoid, and talar lesions [ 29 – 32 ]. The absence of standardized post-treatment follow-up scores prevents assessment of whether baseline IVIM predicts treatment response or return to sport outcomes, which have been shown to be favorable following appropriate OCD management [ 28 ]. Our functional scores represent baseline symptom severity at presentation, not post-treatment outcomes. The relationship between baseline perfusion and baseline symptoms may be weak because symptoms are multifactorial (pain, mechanical symptoms, psychological factors), while IVIM specifically measures perfusion. Prognostic studies require correlation of baseline IVIM with long-term healing and functional recovery. The lack of significant IVIM differences between stable and unstable Hefti groups suggests that morphological stability (cartilage integrity, fragment displacement) and microvascular perfusion may represent different dimensions of JOCD pathology. Both stable and unstable lesions show microvascular compromise relative to healthy tissue, but structural stability may depend on additional factors (mechanical loading, cartilage quality, lesion location) beyond perfusion status. Microvascular compromise may precede morphological instability, such that by the time a lesion becomes unstable (Hefti III-IV), perfusion is already compromised to a similar degree as in stable lesions. Longitudinal studies tracking IVIM parameters over time would clarify temporal relationships. Our age-stratified analysis revealed no significant differences in IVIM parameters between younger (≤ 12 years) and older (> 12 years) patients, contrasting with clinical observations that younger patients have better healing potential. This unexpected finding may reflect several factors: selection bias in our cohort, as IVIM was indicated for complex cases across all ages, potentially missing age-related differences present in less severe lesions; the fact that established JOCD lesions requiring advanced imaging may have already progressed beyond stages where age-related perfusion differences are detectable; or insufficient statistical power for subgroup analysis ( n = 26 vs n = 31). Alternatively, the superior healing outcomes in younger patients observed clinically may relate to factors not captured by cross-sectional baseline IVIM measurements, such as growth plate activity, bone remodeling capacity, and regenerative responses over time. Longitudinal IVIM studies with serial measurements during healing would better elucidate age-related differences in microvascular recovery. IVIM-MRI provides valuable non-invasive characterization of microvascular compromise in JOCD, potentially enabling enhanced understanding of disease pathophysiology. The ability to quantify microvascular status may have particular relevance for emerging biological therapies targeting vascular regeneration and tissue repair, including cell-based approaches for osteochondral defect treatment [ 39 , 40 ]. However, clinical implementation for treatment decision-making requires further validation through prospective studies with arthroscopic ROCK classification as reference standard, longitudinal follow-up correlating baseline IVIM with healing outcomes, multicenter validation to assess generalizability across institutions and MRI protocols, and cost-effectiveness analysis comparing IVIM-guided versus conventional management strategies. Limitations Several important limitations must be acknowledged. The single-center nature and retrospective design limit generalizability and introduce selection bias. Patients undergoing IVIM imaging represent a selected population with complex lesions requiring advanced evaluation. Arthroscopic ROCK classification, considered the definitive assessment of lesion stability, was not systematically recorded in operative notes. This prevents validation of IVIM’s ability to predict true mechanical stability and represents a critical gap for clinical translation. The cross-sectional design with baseline measurements only precludes assessment of IVIM’s prognostic value for predicting healing, treatment response, or long-term outcomes. Systematic post-treatment functional scores were not available for all patients, limiting our ability to correlate baseline microvascular status with clinical recovery. Intra- and inter-observer reliability (ICC) for IVIM parameter measurements was not formally evaluated due to the retrospective consensus-based approach, though the use of automated software-based parameter extraction minimizes subjectivity. The retrospective cohort included patients receiving different treatments (conservative, drilling, fixation) without systematic treatment allocation criteria, confounding correlations between IVIM parameters and functional outcomes. While adequate for primary comparisons ( n = 57), statistical power for subgroup analyses (age strata, treatment types, pattern classifications) was limited, potentially obscuring true associations. Despite detailed protocols and anatomical specifications, ROI placement involves operator-dependent decisions regarding cartilage-bone interface delineation and lesion boundary definition. Automated segmentation methods may improve reproducibility. Finally, IVIM measurements represent a single time point and cannot differentiate acute from chronic ischemia or capture dynamic vascular changes during disease progression or healing. Future research directions To address these limitations and advance clinical translation of IVIM-MRI in JOCD, future research should conduct multicenter prospective cohort studies with standardized IVIM protocols, systematic arthroscopic stability assessment (ROCK classification), and longitudinal follow-up of healing and functional outcomes. Investigation of treatment response monitoring using serial IVIM measurements to assess whether microvascular recovery correlates with clinical healing represents a priority. Formal reliability studies establishing inter-observer and intra-observer ICC values for IVIM parameter measurements are needed. Development and validation of predictive models incorporating IVIM parameters, morphological features, and clinical variables may guide personalized treatment decisions. Exploration of IVIM integration with other advanced MRI techniques (T2 mapping, dGEMRIC) could enable comprehensive assessment of both subchondral bone and cartilage health. Finally, investigation of whether IVIM parameters can identify optimal candidates for emerging biological therapies targeting vascular regeneration will be crucial as regenerative medicine approaches evolve. Conclusions This study demonstrates that IVIM-MRI effectively detects and quantifies microvascular compromise in JOCD lesions, with large effect sizes indicating substantial biological and clinical relevance. Microvascular dysfunction, characterized by reduced perfusion fraction and altered diffusion parameters, is a consistent feature of JOCD lesions relative to healthy subchondral bone. The proposed three-pattern classification system provides a framework for categorizing lesions by perfusion status. However, the clinical utility of IVIM for guiding treatment decisions and predicting outcomes remains to be established. Weak correlations with functional scores and absence of differences between morphologically stable and unstable lesions highlight the complexity of JOCD pathophysiology and the limitations of cross-sectional imaging assessment. Validation through prospective studies with arthroscopic ground truth, longitudinal follow-up, and multicenter replication is essential before IVIM-MRI can be recommended for routine clinical decision-making in JOCD management. Despite these limitations, our findings advance understanding of JOCD pathophysiology by providing objective evidence of microvascular involvement and establish the technical feasibility and diagnostic performance of IVIM-MRI in this pediatric population. With further validation, IVIM may complement conventional imaging in comprehensive JOCD assessment and enable personalized, biology-informed treatment strategies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (56.5KB, jpeg) Supplementary Material 2 (31.4KB, png) Supplementary Material 3 (98.5KB, jpeg) Supplementary Material 4 (104.8KB, jpeg) Supplementary Material 5 (57.8KB, jpeg) Supplementary Material 6 (630.5KB, png) Supplementary Material 7 (43.6KB, png) Supplementary Material 8 (47.7KB, jpeg) Supplementary Material 9 (47.4KB, jpeg) Supplementary Material 10 (24.9KB, jpeg) Supplementary Material 11 (183.1KB, png) Supplementary Material 12 (36.7KB, png) Supplementary Material 13 (58.1KB, jpeg) Supplementary Material 14 (55.1KB, jpeg) Supplementary Material 15 (31.5KB, jpeg) Supplementary Material 16 (65.5KB, png) Supplementary Material 17 (47.2KB, png) Supplementary Material 18 (86.3KB, jpeg) Supplementary Material 19 (92.8KB, jpeg) Supplementary Material 20 (52.4KB, jpeg) Acknowledgements Not applicable. Author contributions Concept and Design: E.Ö., F.E.T.,N.A.; Supervision: M.K.,H.U.Ö. ; Data Collection and/or Processing: E.Ö.,E.D.; Analysis and/or Interpretation: F.E.T.,H.U.Ö.,E.D.; Literature Search: N.A.,M.K.; Writing: E.Ö.,F.E.T.,N.A.; Critical Review: E.Ö.,F.E.T.,N.A.,M.K. Funding None. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study was approved by the Institutional Ethics Committee of Erzurum City Hospital (Ethics Committee for Clinical Research; approval date and number: 15 May 2025 — 39057). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to inclusion in the study. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Konarski W, Poboży T, Konarska K, Derczyński M, Kotela I. Understanding osteochondritis dissecans: a narrative review of the disease commonly affecting children and adolescents. Child. 2024;11(4):498. 10.3390/children11040498. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Chau MM, Klimstra MA, Wise KL, Ellermann JM, Tóth F, Carlson CS, Nelson BJ, Tompkins MA. Osteochondritis dissecans: current understanding of epidemiology, etiology, management, and outcomes. J Bone Joint Surg Am. 2021;103(12):1132–51. 10.2106/JBJS.20.01399. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Akkawi I, Zmerly H, Draghetti M, Felli L. Juvenile osteochondritis dissecans: current concepts. Cureus. 2024;16(7):e65496. 10.7759/cureus.65496. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Turati M, Anghilieri FM, Bigoni M, Rigamonti L, Tercier S, Nicolaou N, Accadbled F. Osteochondritis dissecans of the knee: epidemiology, etiology, and natural history. J Child Orthop. 2023;17(1):40–6. 10.1177/18632521221149063. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Nguyen JC, Kan JH, Patel VS, Blankenbaker DG, Shea KG, Carey JL, Johnson AM, Jaramillo D, Ganley TJ. Osteochondritis dissecans in children: location-dependent differences: (Part I: knee). Pediatr Radiol. 2025;55(7):1366–85. 10.1007/s00247-025-06267-6. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Cabral J, Duart J. Osteochondritis dissecans of the knee in adolescents: How to treat them? J Child Orthop. 2023;17(1):54–62. 10.1177/18632521231152269. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Tsaknakis K, Afifi FK, Lorenz HM, Hell AK. Nichttraumatische osteochondrale Läsionen des Kniegelenkes im Wachstumsalter: Juvenile osteochondrosis dissecans (JOCD) des Kniegelenkes [Non-traumatic osteochondral lesions of the knee joint during growth : Juvenile osteochondritis dissecans (JOCD) of the knee]. Orthopadie. 2024;53(9):651–8. 10.1007/s00132-024-04552-0. [ DOI ] [ PubMed ] [ Google Scholar ] 8. Clark VC, Johnson BL, Wyatt CW, Potter BN, Wilson PL, Ellis HB. Adolescents with osteochondritis dissecans of the femoral condyle present with high rates of corresponding coronal malalignment. J Bone Joint Surg Am. 2025;107(11):1188–99. 10.2106/JBJS.24.00220. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Roaten J, Guevel B, Heyworth B, Kocher M. Osteochondritis dissecans lesions of the pediatric and adolescent knee. Orthop Clin North Am. 2022;53(4):445–59. 10.1016/j.ocl.2022.05.001. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Ekman E, Nevalainen S, Karjalainen E, Kohonen I, Vuohelainen J, Rissanen T, Itälä A. Increased lesion depth, higher body mass index and older age are risk factors for osteoarthritis during long-term follow-up in patients with osteochondritis dissecans of the knee. Arch Orthop Trauma Surg. 2023;143(7):3863–9. 10.1007/s00402-022-04638-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Masquijo J, Kothari A. Juvenile osteochondritis dissecans (JOCD) of the knee: current concepts review. EFORT Open Rev. 2019;4(5):201–12. 10.1302/2058-5241.4.180079. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Hussain ZB, Mathew ST, Feroe AG, Lins LAB, Miller P, Kocher MS. Novel magnetic resonance imaging classification of osteochondritis dissecans of the knee: a reliability study. J Pediatr Orthop. 2021;41(6):e422–6. 10.1097/BPO.0000000000001814. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Feroe AG, Flaugh RA, Majumdar A, Baxter TA, Miller PE, Kocher MS. Validation of a novel magnetic resonance imaging classification for osteochondritis dissecans of the knee. J Pediatr Orthop. 2022;42(5):e486–91. 10.1097/BPO.0000000000002111. [ DOI ] [ PubMed ] [ Google Scholar ] 14. Fabricant PD, Milewski MD, Kostyun RO, Wall EJ, Zbojniewicz AM, Research in Osteochondritis of the Knee (ROCK) Study Group. Osteochondritis Dissecans of the Knee: An Interrater Reliability Study of Magnetic Resonance Imaging Characteristics. Am J Sports Med. 2020;48(9):2118–27. 10.1177/0363546520930427. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Szubert-Franczak AE, Naduk-Ostrowska M, Pasicz K, Podgórska J, Skrzyński W, Cieszanowski A. Intravoxel incoherent motion magnetic resonance imaging: basic principles and clinical applications. Pol J Radiol. 2020;85:e624–35. 10.5114/pjr.2020.101476. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Zbýň Š, Kajabi AW, Nouraee CM, Ludwig KD, Johnson CP, Tompkins MA, Nelson BJ, Zhang L, Moeller S, Marette S, Metzger GJ, Carlson CS, Ellermann JM. Evaluation of lesion and overlying articular cartilage in patients with juvenile osteochondritis dissecans of the knee using quantitative diffusion MRI. J Orthop Res. 2023;41(7):1449–63. 10.1002/jor.25505. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Nammour MA, Mauro CS, Bradley JP, Arner JW. Osteochondritis dissecans lesions of the knee: evidence-based treatment. J Am Acad Orthop Surg. 2024;32(13):587–96. 10.5435/JAAOS-D-23-00494. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Hefti F, Beguiristain J, Krauspe R, Möller-Madsen B, Riccio V, Tschauner C, Wetzel R, Zeller R. Osteochondritis dissecans: a multicenter study of the European Pediatric Orthopedic Society. J Pediatr Orthop B. 1999;8(4):231–45. PMID: 10513356. [ PubMed ] [ Google Scholar ] 19. Dipaola JD, Nelson DW, Colville MR. Characterizing osteochondral lesions by magnetic resonance imaging. Arthroscopy. 1991;7(1):101–4. 10.1016/0749-8063(91)90087-e. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Takahara M. Osteochondritis dissecans of the elbow: recent evolution of pathogenesis, imaging, and treatment modalities. JSES Int. 2023;8(3):588–601. 10.1016/j.jseint.2023.09.010. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Accadbled F, Turati M, Kocher MS. Osteochondritis dissecans of the knee: imaging, instability concept, and criteria. J Child Orthop. 2023;17(1):47–53. 10.1177/18632521221149054. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Mendeszoon MJ, Iosue HE. Pediatric osteochondritis dissecans of the talus. Clin Podiatr Med Surg. 2022;39(1):105–11. 10.1016/j.cpm.2021.09.005. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Haeri Hendy S, de Sa D, Ainsworth K, Ayeni OR, Simunovic N, Peterson D. Juvenile osteochondritis dissecans of the knee: Does magnetic resonance imaging instability correlate with the need for surgical intervention? Orthop J Sports Med. 2017;5(11):2325967117738516. 10.1177/2325967117738516. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Zhen Z, Chen W. Osteochondritis dissecans of the knee at 7-T trabecular bone MRI. Radiology. 2024;311(3):e240048. 10.1148/radiol.240048. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Nguyen JC, Green DW, Lin BF, Endo Y. Magnetic resonance evaluation of the pediatric knee after arthroscopic fixation of osteochondral lesions with biodegradable nails. Skeletal Radiol. 2020;49(1):65–73. 10.1007/s00256-019-03258-1. [ DOI ] [ PubMed ] [ Google Scholar ] 26. ROCKET Group, Nissen C, Bohn DC, Crepeau A, Edmonds E, Ganley T, Kostyun R, Lawrence JTR, Pace JL, Saluan P, Uquillas C, Wall E, Wilson PL, Bae DS. Reliability of radiographic imaging characteristics for osteochondritis dissecans of the capitellum. Am J Sports Med. 2022;50(14):3948–55. 10.1177/03635465221130452. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Nogaro MC, Hartshorn S, Brady M, Offiah A, Faust S, Firth G, Ma J, Dhiman P, O’Mahoney J, Davies L, Spowart C, Moscrop A, Young B, Tudur-Smith C, Collins G, Perry DC, Theologis T. Development of a multicentre cohort study to understand the role of MRI and ultrasound in the diagnosis of acute haematogenous bone and joint infection in children (the PIC Bone study): a study protocol. Bone Jt Open. 2025;6(6):677–84. 10.1302/2633-1462.66.BJO-2024-0277. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Vivekanantha P, Sritharan P, Hemstock R, Johnson J, de Sa D. High rates of return to sport following management of osteochondritis dissecans of the femoral trochlea: A systematic review. Knee Surg Sports Traumatol Arthrosc. 2024;32(2):389–404. 10.1002/ksa.12040. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Kramer DE, Glotzbecker MP, Shore BJ, Zurakowski D, Yen YM, Kocher MS, Micheli LJ. Results of surgical management of osteochondritis dissecans of the ankle in the pediatric and adolescent population. J Pediatr Orthop. 2015;35(7):725–33. . 10.1097/BPO.0000000000000352 [ DOI ] [ PubMed ] [ Google Scholar ] 30. Kiyomatsu H, Takeba J, Imai H, Fujibuchi T, Inoue T, Jono A, Hino K, Miura H. Treatment of osteochondritis dissecans of the humeral capitellum with a fragment fixation method using absorbable pins. JSES Int. 2021;5(3):564–70. 10.1016/j.jseint.2020.12.017. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Tingey MT, Glover MA, St Jeor JD, Trasolini NA, Albertson BS, Fiegen AP, Waterman BR. Surgical management of osteochondritis dissecans of the glenoid via autologous bone marrow aspirate concentrate and extracellular matrix graft in an eighteen-year-old college pitcher. Video J Sports Med. 2024;4(2):26350254231200586. 10.1177/26350254231200586. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Kim HW, Park KB, Kim CW, Lee KY, Park H. Healing predictors of conservative treatment for juvenile osteochondritis dissecans of the talus. Clin J Sport Med. 2022;32(6):e635–43. 10.1097/JSM.0000000000001049. [ DOI ] [ PubMed ] [ Google Scholar ] 33. Baidya Kayal E, Ganguly S, Kandasamy D, Khare K, Sharma R, Bakhshi S, Mehndiratta A. Reproducibility of spatial penalty-based methodologies for intravoxel incoherent motion analysis with diffusion MRI. Sci Rep. 2024;14(1):22811. 10.1038/s41598-024-71173-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Poynton CB, Lee MM, Li Y, Laszik Z, Worters PW, Mackenzie JD, Courtier J Intravoxel incoherent motion analysis of renal allograft diffusion with clinical and histopathological correlation in pediatric kidney transplant patients: A preliminary cross-sectional observational study. Pediatr Transpl. 2017;21(6):12996. 10.1111/petr.12996 [ DOI ] [ PubMed ] [ Google Scholar ] 35. El-Ksas M, El-Metwally D, Fahmy D, Shebel H. Early and late assessment of renal allograft dysfunction using intravoxel incoherent motion (IVIM) and diffusion-weighted imaging (DWI): a prospective study. Abdom Radiol (NY). 2024;49(11):3902–12. 10.1007/s00261-024-04470-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Bane O, Lewis SC, Lim RP, Carney BW, Shah A, Fananapazir G. Contemporary and emerging MRI strategies for assessing kidney allograft complications: arterial stenosis and parenchymal injury, from the AJR special series on imaging of fibrosis. AJR Am J Roentgenol. 2024;222(3):e2329418. 10.2214/AJR.23.29418. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Tóth F, Nissi MJ, Ellermann JM, Wang L, Shea KG, Polousky J, Carlson CS. Novel application of magnetic resonance imaging demonstrates characteristic differences in vasculature at predilection sites of osteochondritis dissecans. Am J Sports Med. 2015;43(10):2522–7. 10.1177/0363546515596410. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Wang L, Wang J, Yang Q, Cai C, Xing Z, Chen Z, Cao D, Cai S. Improved deep learning-based IVIM parameter estimation via the use of more realistic simulated brain data. Med Phys. 2025;52(4):2279–94. 10.1002/mp.17583. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Klinkhammer BM, Lammers T, Mottaghy FM, Kiessling F, Floege J, Boor P. Non-invasive molecular imaging of kidney diseases. Nat Rev Nephrol. 2021;17(10):688–703. 10.1038/s41581-021-00440-4. Erratum in: Nat Rev Nephrol. 2021;17(10):704. Epub 2021 Jun 29. doi: 10.1038/s41581-021-00464-w. https://doi.org/10.1038/s41581-021-00440-4 [ DOI ] [ PMC free article ] [ PubMed ] 40. Kondo M, Kameishi S, Kim K, Metzler NF, Maak TG, Hutchinson DT, Wang AA, Maehara M, Sato M, Grainger DW, Okano T. Safety and efficacy of human juvenile chondrocyte-derived cell sheets for osteochondral defect treatment. NPJ Regen Med. 2021;6(1):65. 10.1038/s41536-021-00173-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1 (56.5KB, jpeg) Supplementary Material 2 (31.4KB, png) Supplementary Material 3 (98.5KB, jpeg) Supplementary Material 4 (104.8KB, jpeg) Supplementary Material 5 (57.8KB, jpeg) Supplementary Material 6 (630.5KB, png) Supplementary Material 7 (43.6KB, png) Supplementary Material 8 (47.7KB, jpeg) Supplementary Material 9 (47.4KB, jpeg) Supplementary Material 10 (24.9KB, jpeg) Supplementary Material 11 (183.1KB, png) Supplementary Material 12 (36.7KB, png) Supplementary Material 13 (58.1KB, jpeg) Supplementary Material 14 (55.1KB, jpeg) Supplementary Material 15 (31.5KB, jpeg) Supplementary Material 16 (65.5KB, png) Supplementary Material 17 (47.2KB, png) Supplementary Material 18 (86.3KB, jpeg) Supplementary Material 19 (92.8KB, jpeg) Supplementary Material 20 (52.4KB, jpeg) Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Articles from Journal of Orthopaedic Surgery and Research are provided here courtesy of BMC ACTIONS View on publisher site PDF (2.3 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top