BACKGROUND:Magnetic resonance imaging (MRI) has proven to be a valuable noninvasive tool to evaluate graft integrity after anterior cruciate ligament (ACL) reconstruction. However, MRI protocols and interpretation methodologies are quite diverse, preventing comparisons of signal intensity across subsequent scans and independent investigations. The purpose of this study was to create an artificial intelligence (AI)-based index (Thessaly Graft Index [TGI]) for the evaluation of graft integrity following ACL reconstruction. METHODS:The cohort study included 24 patients with an isolated ACL injury that had been treated with a hamstring tendon autograft and followed for 1 year. MRI was performed preoperatively and 1 year postoperatively. The clinical and functional evaluations were performed with use of the KT-1000 and with the following patient-reported outcome measures (PROMs): the Knee Injury and Osteoarthritis Outcome Score (KOOS), the International Knee Documentation Committee Subjective Knee Function form (IKDC), the Lysholm score, and the Tegner Activity Scale (TAS). An AI model, based on the YOLOv5 Nano version, was designed to compute the probability of accurately detecting, in the sagittal plane, a healthy ACL (on a percentage scale) and was trained on healthy and injured knees from the KneeMRI dataset. The model was used to assess the integrity of ACL grafts, with a maximum score of 100. The results were compared with the MRI assessment from an independent radiologist and were correlated with PROMs and KT-1000 laxity. RESULTS:The mean preoperative and postoperative TGI scores were 64.21 ± 8.96 and 82.37 ± 3.53, respectively. A mean increase of 15% in the TGI scores was observed between preoperative and postoperative images. The minimum threshold for TGI to categorize a graft as healthy on the postoperative MRI was 79.21%. Twenty-two grafts were characterized as intact and 2 as reruptured, with postoperative TGI scores of 71% and 42%. The radiologist's assessment was in total agreement with the TGI scores. The correlation of the TGI ranged from moderate to good with the TAS (0.668), IKDC (0.516), Lysholm (0.521), KOOS total (0.594), and KT-1000 (0.561). CONCLUSIONS:The TGI is an AI tool that is able to accurately recognize an ACL graft rupture. Moreover, the TGI correlated with the KT-1000 postoperative values and PROM scores. LEVEL OF EVIDENCE:Diagnostic Level IV. See Instructions for Authors for a complete description of levels of evidence.

Background: Magnetic resonance imaging (MRI) has proven to be a valuable noninvasive tool to evaluate graft integrity after anterior cruciate ligament (ACL) reconstruction. However, MRI protocols and interpretation methodologies are quite diverse, preventing comparisons of signal intensity across subsequent scans and independent investigations. The purpose of this study was to create an artificial intelligence (AI)-based index (Thessaly Graft Index [TGI]) for the evaluation of graft integrity following ACL reconstruction. Methods: The cohort study included 24 patients with an isolated ACL injury that had been treated with a hamstring tendon autograft and followed for 1 year. MRI was performed preoperatively and 1 year postoperatively. The clinical and functional evaluations were performed with use of the KT-1000 and with the following patient-reported outcome measures (PROMs): the Knee Injury and Osteoarthritis Outcome Score (KOOS), the International Knee Documentation Committee Subjective Knee Function form (IKDC), the Lysholm score, and the Tegner Activity Scale (TAS). An AI model, based on the YOLOv5 Nano version, was designed to compute the probability of accurately detecting, in the sagittal plane, a healthy ACL (on a percentage scale) and was trained on healthy and injured knees from the KneeMRI dataset. The model was used to assess the integrity of ACL grafts, with a maximum score of 100. The results were compared with the MRI assessment from an independent radiologist and were correlated with PROMs and KT-1000 laxity. Results: The mean preoperative and postoperative TGI scores were 64.21 ± 8.96 and 82.37 ± 3.53, respectively. A mean increase of 15% in the TGI scores was observed between preoperative and postoperative images. The minimum threshold for TGI to categorize a graft as healthy on the postoperative MRI was 79.21%. Twenty-two grafts were characterized as intact and 2 as reruptured, with postoperative TGI scores of 71% and 42%. The radiologist’s assessment was in total agreement with the TGI scores. The correlation of the TGI ranged from moderate to good with the TAS (0.668), IKDC (0.516), Lysholm (0.521), KOOS total (0.594), and KT-1000 (0.561). Conclusions: The TGI is an AI tool that is able to accurately recognize an ACL graft rupture. Moreover, the TGI correlated with the KT-1000 postoperative values and PROM scores. Level of Evidence: Diagnostic Level IV . See Instructions for Authors for a complete description of levels of evidence.

Background: Football has a high injury risk due to speed and contact, and injury patterns may vary by playing position. Positional roles affect physical and physiological demands and may influence injury characteristics. Although this has been examined in professionals, data from amateur players are scarce. This study examined the incidence, type, and severity of injuries among amateur footballers in Greece with respect to playing position. Methods: A retrospective epidemiological study analyzed musculoskeletal injuries in 222 amateur male football players during the 2022–2023 season. Data were collected via a CHERRIES-compliant online survey (SurveyMonkey®) from May to July 2023. Eligible participants were active male athletes aged ≥18 years competing in amateur Greek leagues. Injuries were defined according to the FIFA–UEFA consensus and expressed as incidence rates per 1000 h of exposure. Statistical analyses used SPSS v25 with significance at p < 0.05. Results: Among players (mean age: 25.3 ± 5.7 years), injury prevalence ranged from 65.1% (DFs) to 79.3% (GKs) with no significant association between playing position and injury risk (p = 0.379). Injury incidence ranged from 4.5 to 5.7 per 1000 h, highest among MFs. Incidence rates ranged between 1.33 and 2.74 injuries/1000 h in matches versus 1.33 to 2.09/1000 h in training, with DFs, FWs, and MFs more prone to match injuries, whereas GKs had slightly higher training rates; however, the number of injuries did not significantly differ between games and training across positions (χ2 = 5.21, p = 0.517). Muscle strains and lower-limb injuries predominated. Injury severity differed significantly by position (p = 0.001), but injury type and mechanism did not. Conclusions: GKs and MFs showed the highest prevalence and incidence, but position was not linked to overall risk. Severity differences highlight the need for position-specific prevention strategies.

AbstractThe field of metrology, which focuses on the scientific study of measurement, is grappling with a significant challenge: predicting the measurement accuracy of sophisticated 3D scanning devices. These devices, though transformative for industries like manufacturing, construction, and archeology, often generate complex point cloud data that traditional machine learning models struggle to manage effectively. To address this problem, we proposed a PointNet-based model, designed inherently to navigate point cloud data complexities, thereby improving the accuracy of prediction for scanning devices’ measurement accuracy. Our model not only achieved superior performance in terms of mean absolute error (MAE) across all three axes (X, Y, Z) but also provided a visually intuitive means to understand errors through 3D deviation maps. These maps quantify and visualize the predicted and actual deviations, which enhance the model’s explainability as well. This level of explainability offers a transparent tool to stakeholders, assisting them in understanding the model’s decision-making process and ensuring its trustworthy deployment. Therefore, our proposed model offers significant value by elevating the level of precision, reliability, and explainability in any field that utilizes 3D scanning technology. It promises to mitigate costly measurement errors, enhance manufacturing precision, improve architectural designs, and preserve archeological artifacts with greater accuracy.

Purpose: Meniscal allograft transplantation (MAT) is a valuable option for postmeniscectomy syndrome in young and active patients, which can successfully improve symptoms, function and quality of life. This study aimed to report outcomes and complications in patients treated with isolated MAT or MAT combined with revision anterior cruciate ligament reconstruction (ACLR) and lateral extra-articular tenodesis (LET). Methods: In total, 18 patients who were treated with MAT using a soft tissue fixation technique were retrospectively analysed. Patients were allocated into two groups, the isolated MAT group and the MAT plus revision ACLR + LET group (MAT+ group). Both groups were assessed using the following Patient-Reported Outcome Measures (PROMs): Knee Injury and Osteoarthritis Outcome Score (KOOS), International Knee Documentation Committee (IKDC), Tegner Activity Scale, Lysholm score, EuroQol 5-Dimension 5-level (EQ-5D-5L) questionnaire and a patient acceptable symptom state (PASS) statement. Results: The mean follow-up was 3.2 years and the mean age of patients was 29 (+/- 7.6) years. Ten patients were treated with isolated MAT and eight patients were treated with MAT combined with revision ACLR and LET. No complications and no reoperations were reported. In both groups, all PROMs significantly improved postoperatively (p < 0.05), except for the Tegner score in the MAT+ group, which did not reach statistical significance (not significant [n.s.]). No significant difference was found postoperatively in PROMs between the MAT and the MAT+ group; however, differences in the KOOS Symptoms, Sports and Quality-of-life subscales reached the minimal clinically important threshold. Mean values were: IKDC 61.2 (+/- 13.9) versus 64.7 (+/- 23.2), KOOS Total 72.8 (+/- 14.9) versus 68.0 (+/- 16.7), Lysholm score 83.2 (+/- 11) versus 84.2 (+/- 8.7) and EQ-5D-5L 71.8 (+/- 14.9) versus 72.2 (+/- 22.4). Median values for the Tegner Activity Scale were 4 versus 3. PASS statement was negative in 37.5% of the MAT+ group in comparison to 20% of the MAT group (n.s.). Conclusion: In terms of patient-reported outcomes, MAT is an effective and safe procedure, even when combined with revision ACLR + LET. Compared to the preoperative assessment, a significant improvement of the postoperative PROMs can be expected either when MAT is performed isolated or in combination with ACLR + LET. However, clinically important differences between the two groups can be noted in favour of the isolated MAT procedure, especially regarding symptoms and sporting activity.

Due to their detection capabilities and low cost, unmanned aerial vehicles (UAVs) are commonly used in search and rescue (SAR) operations. In a SAR mission, the UAV's height may change, causing objects to shrink or increase thus affecting generalization. This paper proposes a hybrid object detection (OD) technique that combines altitude-dependent local deep learning (DL) models, each one designed for a given flight altitude range. Seven cutting-edge OD models, including YOLOv4 and v5, EfficientDet, Detectron2, MobileNet, and Faster R-CNN, were trained locally with YOLOv5 and Scaled YOLOv4 being the best performers in low and high-altitude images, respectively. The suggested hybrid strategy, which uses the best OD performers, outperformed well-known DL algorithms with 86.2% mAP on a public dataset. Computing efficiency and accuracy with images of varying resolutions were also explored. Dividing the fundamental detection issue into local subproblems that are treated separately by powerful OD networks might increase SAR capabilities.

Injuries in football affect team performance, players' quality of life, and careers. While research mainly targets professionals, limited data exists for amateur players, especially in Greece. Methodological differences and variations in preparation and technique make applying professional injury data to amateurs uncertain. This study aimed to investigate the injury profile, sports history, and training routines of amateur male football players in Greece over one competitive season. It was conducted as a retrospective cross-sectional study using an electronic questionnaire to survey 222 amateur players aged 18-44 years. Among them, 70.7% reported at least one injury. A total of 220 injuries were recorded, with an incidence rate of 5.3 injuries per 1000 hours of play. Most injuries occurred during matches, predominantly affecting the lower limbs, with the posterior and inner thighs being the most common injury sites. Players were found to be 6.6 times more likely to sustain injuries in matches (12.1 injuries/1000 hours) than in training sessions (1.8 injuries/1000 hours). Strains were identified as the most common type of injury, often resulting from running/sprinting. These findings underscore the importance of implementing specific injury prevention measures in amateur football leagues in Greece to minimise the risks and severity of injuries among players.

This study presents an innovative hybrid deep learning (DL) framework that reformulates the sagittal MRI-based anterior cruciate ligament (ACL) tear classification task as a novelty detection problem to tackle class imbalance. We introduce a highly discriminative novelty score, which leverages the aleatoric semantic uncertainty as this is modeled in the class scores outputted by the YOLOv5-nano object detection (OD) model. To account for tissue continuity, we propose using the global scores (probability vector) when the model is applied to the entire sagittal sequence. The second module of the proposed pipeline constitutes the MINIROCKET timeseries classification model for determining whether a knee has an ACL tear. To better evaluate the generalization capabilities of our approach, we also carry out cross-database testing involving two public databases (KneeMRI and MRNet) and a validation-only database from University General Hospital of Larissa, Greece. Our method consistently outperformed (p-value<0.05) the state-of-the-art (SOTA) approaches on the KneeMRI dataset and achieved better accuracy and sensitivity on the MRNet dataset. It also generalized remarkably good, especially when the model had been trained on KneeMRI. The presented framework generated at least 2.1 times less carbon emissions and consumed at least 2.6 times less energy, when compared with SOTA. The integration of aleatoric semantic uncertainty-based scores into a novelty detection framework, when combined with the use of lightweight OD and timeseries classification models, have the potential to revolutionize the MRI-based injury detection by setting a new precedent in diagnostic precision, speed and environmental sustainability. Our resource-efficient framework offers potential for widespread application.

Modern lifestyles require new tools for determining a person's ability to return to daily activities after knee surgery. These quantitative instruments must feature high discrimination, be non-invasive, and be inexpensive. Machine learning is a revolutionary approach that has the potential to satisfy the aforementioned requirements and bridge the knowledge gap. The scope of this study is to summarize the results of a systematic literature review on the identification of gait-related changes and the determination of the functional recovery status of patients after knee surgery using advanced machine learning algorithms. The current systematic review was conducted using multiple databases in accordance with the PRISMA guidelines, including Scopus, PubMed, and Semantic Scholar. Six out of the 405 articles met our inclusion criteria and were directly related to the quantification of the recovery status using machine learning and gait data. The results were interpreted using appropriate metrics. The results demonstrated a recent increase in the use of sophisticated machine learning techniques that can provide robust decision-making support during personalized post-treatment interventions for knee-surgery patients.

Given the increased interest in smart assistive technologies and autonomous robot vehicles, path planning has emerged as one of the most researched and challenging topics in navigation. Moving to partially known or unknown environment, an assistive navigation system should be able to extract spatiotemporal information and dynamically identify objects and adjust the route. Current approaches typically rely on external services to perform high demanding computations and employ a plethora of overlapping sensors to accurately scan the surrounding environment. This increases their energy demands, size and weight, while incommodes their use in real time applications making their application to wearable assistive systems, such as smart glasses, a challenge. Aiming to provide a comfortable and computationally efficient wearable solution that can be used by human or robotic assistive systems, in this study we propose a novel two-level hierarchical architecture combining global and local path planning. The macroscale navigation involves the construction of the initial global path while the microscale navigation includes the local path planning with obstacle detection and avoidance. The methodology consists of: (i) a novel chaotic ant colony optimization algorithm with fuzzy logic (CACOF) for path construction; (ii) powerful and light weight deep convolutional neural networks for obstacle detection; and (iii) a Bug-like algorithm enhanced with fuzzy rules for obstacle avoidance in case of static objects. A vast experimental evaluation was conducted to test the proposed methodologies in a simulation environment based on the topology of real area. The results proved the computational efficiency and ability of the proposed path planning algorithms to address effectively multi-objective global and path planning problems which make them suitable for real time applications.

Background:Long-term studies of patients after anterior cruciate ligament (ACL) reconstruction with or without concomitant meniscal tear treatment are limited. Purposes:To (1) report postoperative outcomes after anatomic ACL reconstruction with a hamstring autograft, (2) investigate how concomitant treatment of meniscal injury could affect these outcomes, and (3) evaluate the association between quality of life and activity levels at a minimum 10-year follow-up. Study Design:Cohort study; Level of evidence, 3. Methods:Patients treated with a unilateral, anatomic ACL reconstruction between 2005 and 2011 were investigated. The following patient-reported outcome measures (PROMs) were reported for the overall sample as well as a subsample of patients with meniscal injury: International Knee Documentation Committee Subjective Knee Form (IKDC-SKF), Knee injury and Osteoarthritis Outcome Score (KOOS), Lysholm knee score, Tegner activity scale, 5-level EQ-5D (EQ-5D-5L), and patient satisfaction. Sex, age, body mass index (BMI), and meniscal injury treatment (meniscectomy vs meniscal repair) were examined as patient-specific risk factors regarding long-term activity and quality of life. Results:Overall, 106 patients, 90 men (85%) and 16 women (15%), were enrolled in the study, with a mean follow-up of 13.2 years. The ACL retear rate was 2.8%. The mean scores were 80.6 ± 16.7 (IKDC-SKF), 87.4 ± 15.0 (KOOS), 90.5 ± 11.5 (Lysholm), 5.6 ± 1.9 (Tegner), and 91.8 ± 14.5 (EQ-5D-5L). The majority (90.6%) of patients considered their knee state satisfactory during follow-up. When compared with patients who underwent meniscal repair, patients who underwent meniscectomy had statistically significantly lower scores on all PROMs except for the Tegner and EQ-5D-5L (P < .05 for all). The mean difference between the 2 groups was ≥7 points on all PROM scores. Patient sex, age, and BMI did not affect PROM scores. There was a statistically significant, strong positive correlation between quality of life and activity. Conclusion:Patients had few or no symptoms and considered their knee state satisfactory 13.2 years after anatomic ACL reconstruction. Patients with concomitant meniscal tears having undergone meniscal repair had improved PROMs compared with those treated with meniscectomy. Finally, participation in activities of daily living and sports was interrelated with quality of life and was not affected by patient age, sex, or BMI.

Meniscal allograft transplantation (MAT) is an option for young and active patients when most of the native meniscus has been removed. Although the effectiveness of the procedure has been already established, it is unclear whether MAT leads to better clinical outcomes when performed alone or in combination with ACL reconstruction. In total, 20 patients that were treated with a MAT procedure using a soft tissue fixation technique from 2015 to 2020 were retrospectively analyzed. Patients were allocated to 2 groups, the isolated MAT group (group 1) or the MAT plus ACL reconstruction group (group 2). Both groups were assessed clinically and also with the following patient-reported outcome measures (PROMs): the KOOS, the IKDC, the Tegner Activity Scale (difference before the injury and in the latest follow-up–TASd), and the Lysholm Score. The health-related quality of life of the patients was evaluated with the EQ-5D-5L questionnaire. A comparison between the clinical outcomes of medial versus lateral MAT was performed. Mean follow-up was 3.2 years. Mean age of patients at surgery was 29 years. Twelve patients were treated with isolated MAT and 8 patients were treated with MAT and additional ACL reconstruction. All patients had previously undergone a total or subtotal meniscectomy. Twelve medial and eight lateral MAT procedures were performed. Graft survival was 100% and re-operation rate was 0%. For group 1 and 2 mean scores in PROMs were respectively 73% (±13) and 68% (±17) for KOOS, 61,62% (±12) and 65% (±23) for IKDC, 83% (±9) and 84% (±9) for Lysholm score, 2 (±1,04) and 2,99 (±0,43) for TASd and finally 72,4% (±13) versus 73% (±22) for EQ-5D-5L. During statistical analysis, no significantly important difference was found between the two groups regarding PROMs. Regarding the comparison between patients that underwent medial (n = 12) or lateral (n = 8) MAT, a statistically significant difference was found only for KOOS (p = 0.017) with mean scores 65% (±14) for medial meniscus and 79% (±11) for lateral meniscus. MAT is an effective procedure either as an isolated procedure or in combination with ACL reconstruction. In cases of combined MAT and ACL reconstruction patients can benefit from a one-stage operation and avoid multiple surgical procedures that can lead to extended rehabilitation and prolonged time to return to normal activities and sports. Our study was tied in with current literature regarding the effectiveness of MAT procedure between medial and lateral meniscus with better KOOS score for lateral MAT.

Background Machine learning (ML) is increasingly being applied in Cardiology to predict outcomes and assist in clinical decision-making. We sought to develop and validate an ML model for the prediction of mortality after heart transplantation (HT) in adults with congenital heart disease (ACHD). Methods The United Network for Organ Sharing (UNOS) database was queried from 2000 to 2020 for ACHD patients who underwent isolated HT. The study cohort was randomly split into derivation (70%) and validation (30%) datasets that were used to train and test a CatBoost ML model. Feature selection was performed using SHapley Additive exPlanations (SHAP). Recipient, donor, procedural, and post-transplant characteristics were tested for their ability to predict mortality. We additionally used SHAP for explainability analysis, as well as individualized mortality risk assessment. Results The study cohort included 1033 recipients (median age 34 years, 61% male). At 1 year after HT, there were 205 deaths (19.9%). Out of a total of 49 variables, 10 were selected as highly predictive of 1-year mortality and were used to train the ML model. Area under the curve (AUC) and predictive accuracy for the 1-year ML model were .80 and 75.2%, respectively, and .69 and 74.2% for the 3-year model, respectively. Based on SHAP analysis, hemodialysis of the recipient post-HT had overall the strongest relative impact on 1-year mortality after H tau, followed by recipient-estimated glomerular filtration rate, age and ischemic time. Conclusions ML models showed satisfactory predictive accuracy of mortality after HT in ACHD and allowed for individualized mortality risk assessment.

CrossFit (CF) is a popular and rapidly expanding training program in Greece and worldwide. However, there is a lack of scientific evidence on the risk of musculoskeletal injuries related to CF in the Greek population. A self-administered survey of 1224 Greek CF practitioners aged 18 to 59 was conducted and analyzed using the Statistical Package for Social Sciences (SPSS) software. The highest percentage of the participants (34%) practiced 5 days per week for 60 min (42.2%) and had 2 days per week of rest (41.7%). A total of 273 individuals (23%) participated in CF competitions and 948 (77%) did not. The results showed that the most common injuries were muscle injuries (51.3%), followed by tendinopathies (49.6%) and joint injuries (26.6%). The shoulders (56.6%; n = 303), knees (31.8%; n = 170), and lumbar spine (33.1%; n = 177) were the most commonly injured locations. The logistic regression model showed that participation in competitions (p = 0.001), rest per week (p = 0.01), duration of training per session (p = 0.001), and frequency of training per week (p = 0.03) were statistically significant factors for injury. Training level was not a statistically significant factor for injury (p = 0.43). As CF continues to gain popularity on a global scale and the number of athletes gradually increases, it is important to monitor the safety of practitioners. Clinicians should consider participation in competitions, rest, training duration, and frequency in order to make CF safer.

The purpose of this study is to evaluate the clinical and radiological mid-term outcomes of large osteochondral lesions treated with impaction bone grafting in combination with autologous matrix-induced chondrogenesis (BG-AMIC) and identify a possible correlation between the two. A retrospective analysis of 25 patients treated with the BG-AMIC technique for knee osteochondral lesions was performed. Patients were assessed using the following PROMs: the IKDC, the KOOS and the Lysholm score, the Tegner activity scale, and a patient-acceptable symptom state (PASS). The EQ-5D-5L score was used to assess health-related quality of life. Radiological assessment was performed using the MOCART 2.0 score on a 3 T MRI. At a mean of 3.8 years follow-up, all functional scores increased significantly (p < 0.005) when compared to the preoperative baseline. IKDC increased from 44.5 to 81.4, KOOS from 41.5 to 91.6, and Lysholm from 54.4 to 95.2 (p < 0.005). The EQ-5D-5L score also revealed a significant improvement (59.9 to 93.4, p < 0.005). Mean Tegner score reached pre-injury levels. The PASS was positive in 100% of patients. The minimum clinically important difference was reached in all PROMs except for the KOOS Sports subscale. The reoperation rate was 0%. Morphological evaluation of the repair tissue using the MOCART 2.0 score revealed a mean total score of 52.8 (SD). A statistically significant, positive correlation was found between the MOCART 2.0 score and the IKDC score, the KOOS ADL subscale, and the EQ-5D-5L (p < 0.05). BG-AMIC is a safe and reliable option for treating deep, knee osteochondral lesions providing satisfactory patient-reported outcomes. No severe complications were noticed after the procedure. The correlation between MOCART 2.0 and clinical outcomes should be further investigated.

Osteochondral knee defects usually affect young, active patients and may alter knee biomechanics and progressively lead to joint degeneration. Various treatment options exist with autologous, impaction bone grafting in combination with autologous matrix-induced chondrogenesis (BG-AMIC) being a less-expensive, one-step, promising option. The purpose of this study is to evaluate the clinical and radiological mid-term outcomes of large osteochondral lesions treated with BG-AMIC, identify a possible correlation between the two and report postoperative complications and reoperation rate. A retrospective analysis of 25 patients treated with the BG-AMIC technique for knee osteochondral lesions was performed. Patients were assessed using the following PROMs: the IKDC, the KOOS and the Lysholm score, the Tegner activity scale and a patient acceptable symptom state (PASS). The EQ-5D-5L score was used to assess health-related quality of life. Radiological assessment was performed using the MOCART 2.0 score on a 3 T MRI. At a mean of 3.8 (± 0.8)-year follow-up, all functional scores increased significantly (p < 0.005) when compared to the preoperative baseline. IKDC increased from 44.5 (± 15.9) to 81.4 (± 14.7), KOOS from 41.5 (± 16.1) to 91.6 (± 11.6) and Lysholm from 54.4 (± 23) to 95.2 (± 5.5) (p < 0.005). The EQ-5D-5L score also revealed a significant improvement [59.9 (± 25) to 93.4 (± 10.2), p < 0.005]. Mean Tegner score reached pre-injury levels. The PASS was positive in 100

Aim: To test the feasibility and accuracy of a new attention-based deep learning (DL) method for right ventricular (RV) quantification using 2D echocardiography (2DE) with cardiac magnetic resonance imaging (CMR) as reference. Methods and results: We retrospectively analyzed images from 50 adult patients (median age 51, interquartile range 32-62 42% women) who had undergone CMR within 1 month of 2DE. RV planimetry of the myocardial border was performed in end-diastole (ED) and end-systole (ES) for 8 standardized 2DE RV views with calculation of areas. The DL model comprised a Feature Tokenizer module and a stack of Transformer layers. Age, gender and calculated areas were used as inputs, and the output was RV volume in ED/ES. The dataset was randomly split into training, validation and testing subsets (35, 5 and 10 patients respectively). Mean RVEDV, RVESV and RV ejection fraction (EF) were 163±70ml, 82±42ml and 51±8% respectively without differences among the subsets. The proposed method achieved good prediction of RV volumes (R =0.953, absolute percentage error [APE]=9.75±6.23%) and RVEF (APE=7.24±4.55%). Per CMR, there was 1 patient with RV dilatation and 3 with RV dysfunction in the testing dataset. The DL model detected RV dilatation in 1/1 case and RV dysfunction in 4/3 cases. Conclusions: An attention-based DL method for 2DE RV quantification showed feasibility and promising accuracy. The method requires validation in larger cohorts with wider range of RV size and function. Further research will focus on the reduction of the number of required 2DE to make the method clinically applicable.

Anterior cruciate ligament (ACL) tears are very common among athletes. The success of enhanced ACL injury therapy hinges on accurate and cost-effective detection. Deep learning-based techniques have recently dominated ACL injury detection in MRI research. The goal of this study is to develop a robust and lightweight deep learning pipeline for identifying ACL in 3D MRI data of healthy knees. Specifically, we aim at finding the slices in the sagittal plane where the ACL is present. This could be utilized by clinicians for further evaluation. To this end, we build and test an advanced pipeline that relies on the newest object detection network, YOLOv5-Nano. We go on to compare our model to other pipelines that rely on YOLOv5-xlarge, YOLOX-small and YOLOX-nano. YOLOv5-nano is shown to be the best performer, obtaining the highest overall mAP@0.5 performance (0.9727) on augmented data, while at the same time having the smallest model size (3.7 MB). Conclusive object detection is a key step in identifying damage. YOLOv5-nano offers a great solution towards achieving robust object detection healthcare systems that will permit local processing by devices with limited computational resources.

The improved treatment of knee injuries critically relies on having an accurate and cost-effective detection. In recent years, deep-learning-based approaches have monopolized knee injury detection in MRI studies. The aim of this paper is to present the findings of a systematic literature review of knee (anterior cruciate ligament, meniscus, and cartilage) injury detection papers using deep learning. The systematic review was carried out following the PRISMA guidelines on several databases, including PubMed, Cochrane Library, EMBASE, and Google Scholar. Appropriate metrics were chosen to interpret the results. The prediction accuracy of the deep-learning models for the identification of knee injuries ranged from 72.5–100%. Deep learning has the potential to act at par with human-level performance in decision-making tasks related to the MRI-based diagnosis of knee injuries. The limitations of the present deep-learning approaches include data imbalance, model generalizability across different centers, verification bias, lack of related classification studies with more than two classes, and ground-truth subjectivity. There are several possible avenues of further exploration of deep learning for improving MRI-based knee injury diagnosis. Explainability and lightweightness of the deployed deep-learning systems are expected to become crucial enablers for their widespread use in clinical practice.