Medical Imaging Applications Developed Using Artificial Intelligence Demonstrate High Internal Validity Yet Are Limited in Scope and Lack External Validation.

PURPOSE: To (1) review definitions and concepts necessary to interpret applications of deep learning (DL; a domain of artificial intelligence that leverages neural networks to make predictions on media inputs such as images) and (2) identify knowledge and translational gaps in the literature to provide insight into specific areas for improvement as adoption of this technology continues. METHODS: A comprehensive search of the literature was performed in December 2023 for articles regarding the use of DL in sports medicine. For each study, information regarding the joint of focus, specific anatomic structure/pathology to which DL was applied, imaging modality utilized, source of images used for model training and testing, data set size, model performance, and whether the DL model was externally validated was recorded. A numerical scale was used to rate each DL model's clinical impact, with 1 corresponding to proof-of-concept studies with little to no direct clinical impact and 5 corresponding to practice-changing clinical impact and ready for clinical deployment. RESULTS: Fifty-five studies were identified, all of which were published within the past 5 years, while 82% were published within the past 3 years. Of the DL models identified, 84% were developed for classification tasks, 9% for automated measurements, and 7% for segmentation. A total of 62% of studies utilized magnetic resonance imaging as the imaging modality, 25% radiographs, and 7% ultrasound, while 1 study each used computed tomography, arthroscopic images, or arthroscopic video. Sixty-five percent of studies focused on the detection of tears (anterior cruciate ligament [ACL], rotator cuff [RC], and meniscus). The diagnostic performance of ACL tears, as determined by the area under the receiver operator curve (AUROC), ranged from 0.81 to 0.99 for ACL tears (excellent to near perfect), 0.83 to 0.94 for RC tears (excellent), and from 0.75 to 0.96 for meniscus tears (acceptable to excellent). In addition, 3 studies focused on detection of cartilage lesions had AUROC ranging from 0.90 to 0.92 (excellent performance). However, only 4 (7%) studies externally validated their models, suggesting that they may not be generalizable or may not perform well when applied to populations other than that used to develop the model. Finally, the mean clinical impact score was 2 (range, 1-3) on scale of 1 to 5, corresponding to limited clinical applicability. CONCLUSIONS: DL models in orthopaedic sports medicine show generally excellent performance (high internal validity) but require external validation to facilitate clinical deployment. In addition, current models have low clinical applicability and fail to advance the field due to a focus on routine tasks and a narrow conceptual framework. LEVEL OF EVIDENCE: Level IV, scoping review of Level I to IV studies.

Hip Arthroscopy Simulator Training With Immersive Virtual Reality Has Similar Effectiveness to Nonimmersive Virtual Reality.

PURPOSE: To (1) compare the efficacy of immersive virtual reality (iVR) to nonimmersive virtual reality (non-iVR) training in hip arthroscopy on procedural and knowledge-based skills acquisition and (2) evaluate the relative cost of each platform. METHODS: Fourteen orthopaedic surgery residents were randomized to simulation training utilizing an iVR Hip Arthroscopy Simulator (n = 7; PrecisionOS) or non-iVR simulator (n = 7; ArthroS Hip VR; VirtaMed). After training, performance was assessed on a cadaver by 4 expert hip arthroscopists through arthroscopic video review of a diagnostic hip arthroscopy. Performance was assessed using the Objective Structured Assessment of Technical Skills (OSATS) and Arthroscopic Surgery Skill Evaluation Tool (ASSET) scores. A cost analysis was performed using the transfer effectiveness ratio (TER) and a direct cost comparison of iVR to non-iVR. RESULTS: Demographic characteristics did not differ between treatment arms or by training level, hip arthroscopy experience, or prior simulator use. No significant differences were observed in OSATS and ASSET scores between iVR and non-iVR cohorts (OSATS: iVR 19.6 ± 4.4, non-iVR 21.0 ± 4.1, P = .55; ASSET: iVR 23.7 ± 4.5, non-iVR 25.8 ± 4.8, P = .43). The absolute TER was 0.06 and there was a 132-fold cost difference of iVR to non-iVR. CONCLUSIONS: Hip arthroscopy simulator training with iVR had similar performance results to non-iVR for technical skill and procedural knowledge acquisition after expert arthroscopic video assessment. The iVR platform had similar effectiveness in transfer of skill compared to non-iVR with a 132 times cost differential. CLINICAL RELEVANCE: Due to the accessibility, effectiveness, and relative affordability, iVR training may be beneficial in the future of safe arthroscopic hip training.

A practical guide to the development and deployment of deep learning models for the orthopaedic surgeon: Part III, focus on registry creation, diagnosis, and data privacy.

Deep learning is a subset of artificial intelligence (AI) with enormous potential to transform orthopaedic surgery. As has already become evident with the deployment of Large Language Models (LLMs) like ChatGPT (OpenAI Inc.), deep learning can rapidly enter clinical and surgical practices. As such, it is imperative that orthopaedic surgeons acquire a deeper understanding of the technical terminology, capabilities and limitations associated with deep learning models. The focus of this series thus far has been providing surgeons with an overview of the steps needed to implement a deep learning-based pipeline, emphasizing some of the important technical details for surgeons to understand as they encounter, evaluate or lead deep learning projects. However, this series would be remiss without providing practical examples of how deep learning models have begun to be deployed and highlighting the areas where the authors feel deep learning may have the most profound potential. While computer vision applications of deep learning were the focus of Parts I and II, due to the enormous impact that natural language processing (NLP) has had in recent months, NLP-based deep learning models are also discussed in this final part of the series. In this review, three applications that the authors believe can be impacted the most by deep learning but with which many surgeons may not be familiar are discussed: (1) registry construction, (2) diagnostic AI and (3) data privacy. Deep learning-based registry construction will be essential for the development of more impactful clinical applications, with diagnostic AI being one of those applications likely to augment clinical decision-making in the near future. As the applications of deep learning continue to grow, the protection of patient information will become increasingly essential; as such, applications of deep learning to enhance data privacy are likely to become more important than ever before. Level of Evidence: Level IV.

High Rate of Concomitant Lumbo-Sacral Spine Pathology in the Setting of Hip Abductor Tendon Repair.

BACKGROUND: Hip abductor deficiency is a common cause of lateral hip pain in middle-aged patients. Identifying upstream muscle denervation originating in the lumbo-sacral spine could potentially impact the management of patients who have abductor deficiency. The purpose of this study was to estimate the prevalence of lumbo-sacral pathology (L4 to S1) in patients undergoing hip abductor tendon repair. METHODS: All cases of primary hip abductor repair performed at a tertiary care center between January 2010 and December 2021 were reviewed. Patients were classified into the following groups: A) confirmed L4 to S1 disease based on preoperative or perioperative L4 to S1 interventions (ie, surgery, epidural injections, and/or positive electromyography findings); B) radiographic evidence on lumbar spine magnetic resonance imaging demonstrating nerve compression at L4 to S1; and C) no evidence of L4 to S1 disease. RESULTS: There were 131 cases of primary hip abductor repair that were included. Over 80% of patients were women, who had a mean age of 64 years (range, 20 to 85). There were thirteen patients (9.9%) who underwent concomitant total hip arthroplasty (THA). Of the included patients, 29% (n = 38) were categorized into group A, 12% (n = 16) into group B, and 59% (n = 77) into group C. Patients who had L4 to S1 pathology were older than patients who did not have L4 to S1 pathology (67 versus 61 years, P = .004). Of the patients undergoing concomitant THA and hip abductor repair, 54% demonstrated evidence of lumbo-sacral spine pathology. CONCLUSIONS: Over 40% of patients undergoing isolated hip abductor tendon repair and >50% of patients undergoing concomitant hip abductor tendon repair and THA demonstrated evidence of L4 to S1 disease perioperatively. Patients demonstrating symptomatic hip abductor deficiency should be screened for concomitant lower lumbo-sacral spine pathology.

Modern Hip Arthroscopy for FAIS May Delay the Natural History of Osteoarthritis in 25% of Patients: A 12-Year Follow-up Analysis.

BACKGROUND: Little is known about the effect of modern hip arthroscopy on the natural history of femoroacetabular impingement syndrome (FAIS) with respect to joint preservation. PURPOSE: To (1) characterize the natural history of FAIS and (2) understand the effect of modern hip arthroscopy by radiographically comparing the hips of patients who underwent only unilateral primary hip arthroscopy with a minimum follow-up of 10 years. STUDY DESIGN: Cohort study; Level of evidence, 3. METHODS: Between 2010 and 2012, 619 consecutive patients were reviewed from the practice of a single fellowship-trained hip arthroscopic surgeon. Inclusion criteria were FAIS, bilateral radiographic findings of femoroacetabular impingement, primary unilateral hip arthroscopy (labral repair, femoroplasty, or capsular closure), and minimum 10-year follow-up. The preoperative and minimum 10-year postoperative radiographs of patients were evaluated at each time point. Both operative and nonoperative hips were graded using the Tönnis classification or the presence of hip arthroplasty by 2 independent reviewers. Subgroup analyses were performed. RESULTS: A total of 200 hips from 100 patients were evaluated at a mean follow-up of 12.0 years. Preoperatively, 98% and 99% of operative and nonoperative hips were evaluated as Tönnis grades 0 and 1, respectively; 5% of nonoperative hips had worse Tönnis grades than operative hips. The nonoperative hip advanced to a worse Tönnis grade in 48% (48/100) of cases compared with 28% (28/100) among operative hips. At follow-up, Tönnis grades between hips were equal in 70% (70/100) of the cases, the operative hip had a better grade 25% (25/100) of the time, and the nonoperative hip had a better grade 5% (5/100) of the time. Modern hip arthroscopy was associated with a relative risk reduction of 42% in osteoarthritis progression. Impingement with borderline dysplasia, age, preoperative Tönnis grade, and alpha angle >65° were key risk factors in the radiographic progression of osteoarthritis. CONCLUSION: Although the majority of patients (70%) undergoing hip arthroscopy for FAIS did not experience differences between operative and nonoperative hips in terms of the radiographic progression of osteoarthritis, the natural history may be favorably altered for 25% of patients whose Tönnis grade was better after undergoing arthroscopic correction. Modern hip arthroscopy indications and techniques represent a valid joint-preservation procedure conferring a relative risk reduction of 42% in the progression of osteoarthritis. Arthroscopy for mixed patterns of impingement and instability were the fastest to degenerate.

Clinical Results of Primary Repair Versus Reconstruction of the Anterior Cruciate Ligament: A Systematic Review and Meta-analysis of Contemporary Trials.

Background: Primary anterior cruciate ligament (ACL) repair has gained renewed interest in select centers for patients with proximal or midsubstance ACL tears. Therefore, it is important to reassess contemporary clinical outcomes of ACL repair to determine whether a clinical benefit exists over the gold standard of ACL reconstruction (ACLR). Purpose: To (1) perform a meta-analysis of comparative trials to determine whether differences in clinical outcomes and adverse events exist between ACL repair versus ACLR and (2) synthesize the midterm outcomes of available trials. Study Design: Systematic review; Level of evidence, 3. Methods: The PubMed, OVID/Medline, and Cochrane databases were queried in August 2023 for prospective and retrospective clinical trials comparing ACL repair and ACLR. Data pertaining to tear location, surgical technique, adverse events, and clinical outcome measures were recorded. DerSimonian-Laird random-effects models were constructed to quantitatively evaluate the association between ACL repair/ACLR, adverse events, and clinical outcomes. A subanalysis of minimum 5-year outcomes was performed. Results: Twelve studies (893 patients; 464 ACLR and 429 ACL repair) were included. Random-effects models demonstrated a higher relative risk (RR) of recurrent instability/clinical failure (RR = 1.64; 95% confidence interval [CI], 1.04-2.57; P = .032), revision ACLR (RR = 1.63; 95% CI, 1.03-2.59; P = .039), and hardware removal (RR = 4.94; 95% CI, 2.10-11.61; P = .0003) in patients who underwent primary ACL repair versus ACLR. The RR of reoperations and complications (knee-related) were not significantly different between groups. No significant differences were observed when comparing patient-reported outcome scores. In studies with minimum 5-year outcomes, no significant differences in adverse events or Lysholm scores were observed. Conclusion: In contemporary comparative trials of ACL repair versus ACLR, the RR of clinical failure, revision surgery due to ACL rerupture, and hardware removal was greater for primary ACL repair compared with ACLR. There were no observed differences in patient-reported outcome scores, reoperations, or knee-related complications between approaches. In the limited literature reporting on minimum 5-year outcomes, significant differences in adverse events or the International Knee Documentation Committee score were not observed.

Maternal and Fetal Outcomes in Pulmonary Hypertension During Pregnancy: A Contemporary Nationwide Analysis.

Pulmonary hypertension (PH) disproportionately affects women, presenting challenges during pregnancy. Historically, patients with PH are advised to avoid pregnancy; however, recent reports have indicated that the incidence of adverse events in pregnant patients with PH may be lower than previously reported. We conducted a retrospective cohort study in pregnant patients with PH using the National Readmission Database from January 1, 2016, to December 31, 2020. PH was categorized according to the World Health Organization classification. Primary end points include maternal mortality and 30-day nonelective readmission rate. Other adverse short-term maternal (cardiovascular and obstetric) and fetal outcomes were also analyzed. Of 9,922,142 pregnant women, 3,532 (0.04%) had PH, with Group 1 PH noted in 1,833 (51.9%), Group 2 PH in 676 (19.1%), Group 3 PH in 604 (17.1%), Group 4 PH in 23 (0.7%), Group 5 PH in 98 (2.8%), and multifactorial PH in 298 (8.4%). PH patients exhibited higher rates of adverse cardiovascular events (15.7% vs 0.3% without PH, p <0.001) and mortality (0.9% vs 0.01% without PH, p <0.001). Mixed PH and Group 2 PH had the highest prevalence of adverse cardiovascular events in the World Health Organization PH groups. Patients with PH had a significantly higher nonelective 30-day readmission rate (10.4% vs 2.3%) and maternal adverse obstetric events (24.2% vs 9.1%) compared with those without PH (p <0.001) (Figure 1). In conclusion, pregnant women with PH had significantly higher adverse event rates, including in-hospital maternal mortality (85-fold), compared with those without PH.

Segond Fractures Can Be Identified With Excellent Accuracy Utilizing Deep Learning on Anteroposterior Knee Radiographs.

Purpose: To develop a deep learning model for the detection of Segond fractures on anteroposterior (AP) knee radiographs and to compare model performance to that of trained human experts. Methods: AP knee radiographs were retrieved from the Hospital for Special Surgery ACL Registry, which enrolled patients between 2009 and 2013. All images corresponded to patients who underwent anterior cruciate ligament reconstruction by 1 of 23 surgeons included in the registry data. Images were categorized into 1 of 2 classes based on radiographic evidence of a Segond fracture and manually annotated. Seventy percent of the images were used to populate the training set, while 20% and 10% were reserved for the validation and test sets, respectively. Images from the test set were used to compare model performance to that of expert human observers, including an orthopaedic surgery sports medicine fellow and a fellowship-trained orthopaedic sports medicine surgeon with over 10 years of experience. Results: A total of 324 AP knee radiographs were retrieved, of which 34 (10.4%) images demonstrated evidence of a Segond fracture. The overall mean average precision (mAP) was 0.985, and this was maintained on the Segond fracture class (mAP = 0.978, precision = 0.844, recall = 1). The model demonstrated 100% accuracy with perfect sensitivity and specificity when applied to the independent testing set and the ability to meet or exceed human sensitivity and specificity in all cases. Compared to an orthopaedic surgery sports medicine fellow, the model required 0.3% of the total time needed to evaluate and classify images in the independent test set. Conclusions: A deep learning model was developed and internally validated for Segond fracture detection on AP radiographs and demonstrated perfect accuracy, sensitivity, and specificity on a small test set of radiographs with and without Segond fractures. The model demonstrated superior performance compared with expert human observers. Clinical Relevance: Deep learning can be used for automated Segond fracture identification on radiographs, leading to improved diagnosis of easily missed concomitant injuries, including lateral meniscus tears. Automated identification of Segond fractures can also enable large-scale studies on the incidence and clinical significance of these fractures, which may lead to improved management and outcomes for patients with knee injuries.

Novel Risk Stratification and Hemodynamic Profiling in Acute Pulmonary Embolism: A Proposed Classification Inspired by Society for Cardiovascular Angiography and Intervention Shock Staging.

Treatment options for patients with acute pulmonary embolism (PE) and right ventricular shock (RVS) have grown exponentially. Therapy options include anticoagulation, systemic thrombolysis, catheter-based thrombolysis/ thrombectomy, and may include short-term mechanical circulatory support. However, the incidence of short-term morbidity and mortality has not changed despite the emergence of several advanced therapies in acute PE. This is possibly due to the inclusion of heterogenous populations in research studies without differentiation based on the acuity/severity of presentation. We propose a novel classification for PE-RVS to allow for standardizing appropriate therapy escalation and better communication of the severity among cardiovascular critical care, and emergency health care professionals.