The 100 most cited articles in artificial intelligence related to orthopedics.

Background: This bibliometric study aimed to identify and analyze the top 100 articles related to artificial intelligence in the field of orthopedics. Methods: The articles were assessed based on their number of citations, publication years, countries, journals, authors, affiliations, and funding agencies. Additionally, they were analyzed in terms of their themes and objectives. Keyword co-occurrence, co-citation of authors, and co-citation of references analyses were conducted using VOSviewer (version 1.6.19). Results: The number of citations of these articles ranged from 32 to 272, with six papers having more than 200 citations The years of 2019 (n: 37) and 2020 (n: 19) together constituted 56% of the list. The USA was the leading contributor country to this field (n: 61). The most frequently used keywords were "machine learning" (n: 26), "classification" (n: 18), "deep learning" (n: 16), "artificial intelligence" (n: 14), respectively. The most common themes were decision support (n: 25), fracture detection (n: 24), and osteoarthrtitis staging (n: 21). The majority of the studies were diagnostic in nature (n: 85), with only two articles focused on treatment. Conclusions: This study provides valuable insights and presents the historical perspective of scientific development on artificial intelligence in the field of orthopedics. The literature in this field is expanding rapidly. Currently, research is generally done for diagnostic purposes and predominantly focused on decision support systems, fracture detection, and osteoarthritis classification.

Mortality Factors in Crush Syndrome.

BACKGROUND: Crush Syndrome is a major cause of morbidity and mortality following large-scale catastrophic earthquakes. Since there are no randomized controlled studies on Crush Syndrome, knowledge on this subject is limited to expert experience. The primary objective is to analyze the epidemiological and demographic characteristics, clinical outcomes, and mortality factors of earthquake victims after the Pazarcik and Elbistan earthquakes on February 6, 2023. METHODS: This cross-sectional and observational retrospective study evaluated 610 earthquake victims who presented to our center between February 6 and April 30, 2023. Among these patients, 128 with Crush Syndrome were included in the study. Patient information was gathered from hospital records during their stay and from national registries upon referral. The primary outcome was to identify risk factors for mortality. Demographic and laboratory data were analyzed by acute kidney injury (AKI) stages; mortality-affecting factors were identified through regression analysis. RESULTS: Of the 128 Crush Syndrome patients (100 adults, 28 children), 64 were female. The AKI rate was 32.8%. Among patients with AKI, the frequency of hemodialysis requirement was 69%, and the mortality rate was 14.2%. The overall mortality rate for patients with Crush Syndrome was 4.6%, compared to 3.9% (19/482) in earthquake victims without Crush Syndrome (p=0.705). Notably, low systolic blood pressure at admission was the only factor significantly affecting mortality in Crush Syndrome patients (Hazard Ratio [HR]: 1.088, p=0.021, 95% Confidence Interval [CI]). CONCLUSION: Our study highlights low systolic blood pressure upon admission as a significant risk factor for increased mortality in Crush Syndrome patients. This finding may contribute to the literature by emphasizing the importance of monitoring blood pressure under rubble and administering more aggressive fluid therapy to patients with low systolic blood pressure.

Evaluation of musculoskeletal system injuries after the 2023 Kahramanmaras Turkey earthquakes: Our single-center experience.

INTRODUCTION: This study aimed to retrospectively analyze patients who presented to the orthopedic and traumatology clinic following the 2023 Kahramanmaras earthquakes. PATIENTS AND METHODS: Over a week after the earthquakes, two hundred and sixty patients were consulted at our clinic. Demographic data of the patients, duration of being under the rubble, fracture locations, types of surgeries performed, number of surgical sessions attended by individuals, and early mortality rate within one month were determined. RESULTS: The mean age of the patients was 40.2 ± 22.4 years. One hundred thirty-eight (53.1%) were female, and 122 (46.9%) were male. The average duration of being under the rubble was determined as 27.1 ± 28.0 hours. Sixteen patients died within one month after the earthquake. The one-month mortality rate among patients with orthopedic injuries was 6.15%. Forty-seven fasciotomies were performed in 35 patients, and 22 amputations were performed in 19 patients. The most injured region was the lower extremity (78 cases, 40%). The ratio of external and internal fixation in extremity fractures was 22%. CONCLUSIONS: The management of musculoskeletal injuries can be successful with proper triage and treatment plans. Decisions regarding fasciotomy and amputation in patients with crush syndrome following an earthquake should be individualized. Implant sets should be planned accordingly, especially considering the higher occurrence of lower extremity injuries.

Comparison of artificial intelligence algorithm for the diagnosis of hip fracture on plain radiography with decision-making physicians: a validation study.

OBJECTIVE: This study aimed to compare an algorithm developed for diagnosing hip fractures on plain radiographs with the physicians involved in diagnosing hip fractures. METHODS: Radiographs labeled as fractured (n=182) and non-fractured (n=542) by an expert on proximal femur fractures were included in the study. General practitioners in the emergency department (n=3), emergency medicine (n=3), radiologists (n=3), orthopedic residents (n=3), and orthopedic surgeons (n=3) were included in the study as the labelers, who labeled the presence of fractures on the right and left sides of the proximal femoral region on each anteroposterior (AP) plain pelvis radiograph as fractured or non-fractured. In addition, all the radiographs were evaluated using an artificial intelligence (AI) algorithm consisting of 3 AI models and a majority voting technique. Each AI model evaluated each graph separately, and majority voting determined the final decision as the majority of the outputs of the 3 AI models. The results of the AI algorithm and labelling physicians included in the study were compared with the reference evaluation. RESULTS: Based on F-1 scores, here are the average scores of the group: majority voting (0.942) > orthopedic surgeon (0.938) > AI models (0.917) > orthopedic resident (0.858) > emergency medicine (0.758) > general practitioner (0.689) > radiologist (0.677). CONCLUSION: The AI algorithm developed in our previous study may help recognize fractures in AP pelvis in plain radiography in the emergency department for non-orthopedist physicians. LEVEL OF EVIDENCE: Level IV, Diagnostic Study.

The ensemble artificial intelligence (AI) method: Detection of hip fractures in AP pelvis plain radiographs by majority voting using a multi-center dataset.

Introduction: This article was undertaken to explore the potential of AI in enhancing the diagnostic accuracy and efficiency in identifying hip fractures using X-ray radiographs. In the study, we trained three distinct deep learning models, and we utilized majority voting to evaluate their outcomes, aiming to yield the most reliable and precise diagnoses of hip fractures from X-ray radiographs. Methods: An initial study was conducted of 10,849 AP pelvis X-rays obtained from five hospitals affiliated with Baskent University. Two expert orthopedic surgeons initially labeled 2,291 radiographs as fractures and 8,558 as non-fractures. The algorithm was trained on 6,943 (64%) radiographs, validated on 1,736 (16%) radiographs, and tested on 2,170 (20%) radiographs, ensuring an even distribution of fracture presence, age, and gender. We employed three advanced deep learning architectures, Xception (Model A), EfficientNet (Model B), and NfNet (Model C), with a final decision aggregated through a majority voting technique (Model D). Results: For each model, we achieved the following metrics:For Model A: F1 Score 0.895, Accuracy 0.956, Specificity 0.973, Sensitivity 0.893.For Model B: F1 Score 0.900, Accuracy 0.960, Specificity 0.991, Sensitivity 0.845.For Model C: F1 Score 0.919, Accuracy 0.966, Specificity 0.984, Sensitivity 0.899.For Model D: F1 Score 0.929, Accuracy 0.971, Specificity 0.991, Sensitivity 0.897.We concluded that Model D (majority voting) achieved the best results in terms of the F1 score, accuracy, and specificity values. Conclusions: Our study demonstrates that the results obtained by aggregating the decisions of multiple models through voting, rather than relying solely on the decision of a single algorithm, are more consistent. The practical application of these algorithms will be difficult due to ethical, legal, and confidentiality issues, despite the theoretical success achieved. Developing successful algorithms and methodologies should not be viewed as the ultimate goal; it is important to understand how these algorithms will be used in real-life situations. In order to achieve more consistent results, feedback from clinical practice will be helpful.

Comparison of efficacy of Oral versus Intra-articular Corticosteroid Application in the treatment of Frozen Shoulder: An experimental study in rats.

OBJECTIVE: The aim of this study was to compare clinical and histopathological effects of oral versus intraarticular corticosteroid application in a rat model of frozen shoulder. METHODS: In this study, eighty adult Sprague-Dawley rats were used. The animals were divided into 5 equal groups. The frozen shoulder model was created by immobilizing animals' shoulders with internal fixation with sutures for 8 weeks. At the 8th week, sham (n: 16) and control (n: 16) groups were sacrificed to collect data for healthy and affected shoulders. Also, at the 8th week, 50 mg/ kg methylprednisolone was started for the oral treatment group, and a single dose of 0.5 mg/kg triamcinolone acetonide was injected for the intraarticular treatment group. The effect of additional steroid treatment was expected for 2 weeks, then all remaining treatment and natural course groups were sacrificed on the 10th week. RESULTS: After sacrification, specimens taken as "en bloc" scapulothoracic disarticulation were randomly divided into two groups for a range of motion measurement and histopathological examination. The control (frozen shoulder model) group's shoulder range of motion in all directions was lower than the sham (healthy) group (P < 0.01). Natural course and intraarticular steroid groups, compared to the frozen shoulder model showed a significant increase in the direction of abduction (P < 0.05). Also, it was found for treatment groups that in all directions the range of motion was not as good as the healthy values (P < 0.01). The intraarticular treatment group showed higher degrees of abduction compared to the natural course and oral steroid treatment groups (P < 0.01). Oral steroid treatment group's range of motion was not significantly better than the disease model and had no superiority to the natural course group (P > 0.05). Histopathologically, no statistically significant difference was observed between the groups for signs of frozen shoulder which was found in the immobilized group (P > 0.05). Histopathologically, immobilization was found to cause thickening of the capsule that cannot be resolved by treatment. (P < 0.05). CONCLUSION: In frozen shoulder disease, intraarticular steroid injection seems to be superior in increasing the range of motion than oral steroid treatment.

Necrotizing Fasciitis: Evaluation of 85 Cases and Usage of LRINEC Score.

Objective: The present study aims to define the characteristics of the necrotizing fasciitis (NF) cases followed at our hospital and to compare our results with the literature. Materials and Methods: In this study, NF cases followed and treated at our hospital from January 2005 to April 2019 were evaluated retrospectively. Results: A total of 85 cases of NF were included in the study. Of the cases, 33 (39%) were female and the median age was 59.8±13.1 years (range: 26-92 years). Diabetes mellitus (DM) (56%) was the most prevalent comorbid condition. Extremities were the most frequently involved field found in 41 (48%) of the cases followed by Fournier's gangrene found in 34 (40%) of the cases. All of the cases had undergone surgical intervention (debridement and/or amputation) and received broad-spectrum antibiotic therapy. Laboratory risk indicator for necrotizing fasciitis (LRINEC) score was calculated for 60 cases, and it was 6 or higher in 78% of them. Nineteen (22%) of 85 cases had died. Conclusion: Necrotizing fasciitis affects generally older male patients with DM. In NF cases to avoid the higher risk of mortality, the removal of necrotic tissue via surgical procedure together with antimicrobial therapy is required urgently; therefore, it is very important to differentiate NF from soft tissue infections as soon as possible. As the LRINEC score predicted NF among nearly 80% of our patients, this score could be used as an early diagnostic tool of NF. Level of Evidence: Level IV, case series.

A brief history of artificial intelligence and robotic surgery in orthopedics & traumatology and future expectations.

Recently, the rate of the production and renewal of information makes it almost impossible to be updated. It is quite difficult to process and interpret large amounts of data by human beings. Unlimited memory capacities, learning abilities, artificial intelligence (AI) applications, and robotic surgery techniques cause orthopedic surgeons to be concerned about losing their jobs. The idea of AI, which was first introduced in 1956, has evolved over time by revealing deep learning and evolutionary plexus that can mimic the human neuron cell. Image processing is the leading improvement in developed algorithms. Theoretically, these algorithms appear to be quite successful in interpreting medical images and orthopedic decision support systems for preoperative evaluation. Robotic surgeons have emerged as significant competitors in carrying out the taken decisions. The first robotic applications of orthopedic surgery started in 1992 with the ROBODOC system. Applications started with hip arthroplasty continued with knee arthroplasty. Publications indicate that problems such as blood loss and infection caused by the long operation time in the early stages have been overcome in time with the help of learning systems. Comparative studies conducted with humans indicate that robots are better than humans in providing limb lengthening, patient satisfaction, and cost. As in all new technologies, the developments in both AI applications and robotics surgery indicate that technology is in favor in terms of cost/benefit analyses. Although studies indicate that new technologies are more successful than humans, the replacement of technology with experience and long-term results with traditional methods will not be observed in the near future.

Femoral neck fracture detection in X-ray images using deep learning and genetic algorithm approaches.

OBJECTIVES: This study aims to detect frontal pelvic radiograph femoral neck fracture using deep learning techniques. PATIENTS AND METHODS: This retrospective study was conducted between January 2013 and January 2018. A total of 234 frontal pelvic X-ray images collected from 65 patients (32 males, 33 females; mean age 74.9 years; range, 33 to 89 years) were augmented to 2106 images to achieve a satisfactory dataset. A total of 1,341 images were fractured femoral necks while 765 were non-fractured ones. The proposed convolutional neural network (CNN) architecture contained five blocks, each containing a convolutional layer, batch normalization layer, rectified linear unit, and maximum pooling layer. After the last block, a dropout layer existed with a probability of 0.5. The last three layers of the architecture were a fully connected layer of two classes, a softmax layer and a classification layer that computes cross entropy loss. The training process was terminated after 50 epochs and an Adam Optimizer was used. Learning rate was dropped by a factor of 0.5 on every five epochs. To reduce overfitting, regularization term was added to the weights of the loss function. The training process was repeated for pixel sizes 50x50, 100x100, 200x200, and 400x400. The genetic algorithm (GA) approach was employed to optimize the hyperparameters of the CNN architecture and to minimize the error after testing the model created by the CNN architecture in the training phase. RESULTS: Performance in terms of sensitivity, specificity, accuracy, F1 score, and Cohen's kappa coefficient were evaluated using five- fold cross validation tests. Best performance was obtained when cropped images were rescaled to 50x50 pixels. The kappa metric showed more reliable classifier performance when 50x50 pixels image size was used to feed the CNN. The classifier performance was more reliable according to other image sizes. Sensitivity and specificity rates were computed to be 83% and 73%, respectively. With the inclusion of the GA, this rate increased by 1.6%. The detection rate of fractured bones was found to be 83%. A kappa coefficient of 55% was obtained, indicating an acceptable agreement. CONCLUSION: This experimental study utilized deep learning techniques in the detection of bone fractures in radiography. Although the dataset was unbalanced, the results can be considered promising. It was observed that use of smaller image size decreases computational cost and provides better results according to evaluation metrics.

The effect of a new universal laser aiming device in C-arm fluoroscopy on the technician's accuracy.

OBJECTIVES: This study aims to introduce a new low-cost universal laser aiming device (LAD) that can be used in existing C-arm fluoroscopy devices, independent of brand and model, and to determine whether this new universal LAD improves technician accuracy in locating the desired region at the midpoint of the fluoroscopic image. MATERIALS AND METHODS: A low-priced universal LAD that is compatible with existing 12-inch C-arm fluoroscopy devices was designed. Eight radiology technicians with varied levels of experience in C-arm fluoroscopy participated in the study. A 12 mm cortical screw with a diameter of 3.5 mm was placed on proximal, diaphyseal, and distal points of femur, tibia, and humerus bones in the anteroposterior plane on L3 vertebrae and the left pubis arm in the pelvis bone model. Technicians were asked to align each screw in the image center 10 times from a distance of 30 cm in the anterolateral plane, first without the LAD and then with the LAD. The distance of the screw head to the center point was measured from the 3,520 images with the help of medical viewer software based on the X- and Y-axis. RESULTS: Each fluoroscopic image was divided into 48 equal parts and the length of a part was taken as one unit for distance measurements. The compliance between technicians without the LAD was 0.347 (95% confidence interval [CI]: 0.208-0.47, p=0.001) and with the LAD was 0.687 (95% CI: 0.621-0.741, p=0.001). The distance between the screw head and the center of the image without the LAD was 19.0±9.8 for technicians with more than 10 years of experience and 28.0±12.9 for those with less than 10 years of experience. This difference was statistically significant (p=0.001). When the LAD was used, the difference between the less experienced (3.1±1.5) and more experienced (3.3±2.0) technicians was statistically reduced, along with the distance (p=0.033). CONCLUSION: The use of the LAD with C-arm fluoroscopy appears to be successful in helping technicians capture the desired point in the center of the fluoroscopic image. The use of the LAD reduces the experience gap between technicians.

Do diabetic patients who undergo transtibial amputation receive adequate treatment?

AIMS: To determine if patients who undergo below-knee amputation (BKA) for intractable wounds caused by diabetes complications receive adequate treatment before surgery. MATERIALS AND METHODS: The study included a total of 528 patients who underwent transtibial amputation for diabetic foot. All patients were assessed on the basis of duration of preoperative treatment, HbO therapy, negative wound pressure therapy (NPWT), peripheral vascular angioplasty (PVA) treatment, wound cultures, antibiotic medications, consultations with plastic and vascular surgeons, need for hemodialysis treatment, use of anticoagulant treatment as an inpatient, and assessment of blood sugar regulation by an endocrinologist. HbA1c, BUN, Creatinine, ESR, and CRP values attained for preoperative assessment were noted. RESULTS: Eighteen patients (3.5%) received HbO therapy, 35 (6.7%) NPWT therapy and 347 (65.7%) anticoagulant treatment. Wound cultures were taken in 317 patients (60.5%) and 390 (73.9%) received preoperative antibiotic treatment. 45 (8.6%) patients were assessed by plastic surgeon with 22 (4.2%) subsequently undergoing surgery by the plastic surgeon. Vascular surgeons assessed 163 patients (30.9%) and performed procedures on 45 patients (8.6%). Endocrinologists assessed 316 patients (59.8%) and implemented blood sugar regulation. PVA treatment was performed in 246 patients (46.6%). Patients who were managed medically for more than 7 days after the initial assessment received more HbO therapy (p=0.037), anticoagulant treatment (p=0.015), IV antibiotics (0.001), blood sugar regulation attempts (p=0.001), and PVA therapy (0.001) and had more cultures taken (p=0.001). These patients also received overall more diagnostic and treatment modalities than those that received definitive surgical intervention within 7 days. CONCLUSIONS: The duration of time patients with diabetes-related foot problems who see orthopedic surgeons for longer periods of time receive more treatment modalities and are referred more often to specialists before transtibial amputation surgery. We believe that delayed presentation is one of the main obstacles prohibiting adequate treatment for these patients. KEY WORDS: Diabetic foot, Transtibial amputation, Treatment.

Factors affecting lifespan following below-knee amputation in diabetic patients.

INTRODUCTION: Untreatable foot problems in diabetics may require lower extremity amputation, which has a high level of patient mortality. This high mortality rate is worse than most malignancies. The present study aimed to identify parameters that can be used to estimate survival in DM patients undergoing below-knee amputations for diabetic foot problems. MATERIALS AND METHODS: A total of 470 patients (299 males, 171 females) with a mean age of 64.32 years who underwent below-knee amputation for diabetic foot problems between 2004 and 2014 were enrolled in the study. The length of time from the operation to time of death was recorded in days. Patient details were obtained, including age during surgery, BMI, oral antidiabetic and insulin usage, dialysis therapy history, lower extremity endovascular intervention, previous amputation at the same extremity, the need for stump revision surgery during follow-up, and above-knee amputation at the same site. Biochemical test results of pre-operative HbA1c, ESR, and levels of CRP, BUN, and creatinine were also obtained. RESULTS: A total of 333 patients (70.9%) died and 137 (29.1%) survived post-surgery. Survival rates were 90% in the first 7 days, 84% in the first 30 days, and 64% after the first year. Patient median life expectancy post-surgery was 930 ± 106 days. Hemodialysis treatment (p = 0.001), endovascular intervention (p = 0.04), sex (p = 0.004), age (p = 0.001), BUN level (p = 0.001), and duration of insulin use (p = 0.003) were shown to be effective predictors of mortality. CONCLUSIONS: Life expectancy is low (<3 years) in DM patients requiring below-knee amputations for untreatable foot problems. Survival could be predicted by duration of insulin use, age, sex, and renal insufficiency. LEVEL OF EVIDENCE: Level IV, Therapeutic study.

Comparison of tunnel expansion and isometric muscle strength after ACL reconstruction with single- or dual-bundle hamstring allograft: a prospective, randomized study.

OBJECTIVE: The aim of this study was to compare tunnel expansion and isokinetic muscle strength after single- and dual-bundle reconstruction of the anterior cruciate ligament (ACL). METHODS: This study included 34 patients who underwent ACL reconstruction in our clinic between November 2007 and March 2008. Eighteen patients (average age: 27.3 years; range: 19 to 35 years) underwent single-bundle ACL reconstruction and sixteen patients (average age: 30.1 years; range: 20 to 40 years) underwent dual-bundle ACL reconstruction. Method selection was determined by draw. Isokinetic hamstring and quadriceps muscle strength was tested preoperatively using Biodex 3. Three-dimensional computed tomography of the knee joint was taken in the 2nd and 3rd postoperative month. The three-dimensional computed tomography and isokinetic muscle strength tests were repeated at the 6th month follow-up. Each tunnel was divided into six equal parts, and the tunnel width in the sagittal and coronal planes was measured and the same points in the axial plane were measured in the tunnel area. RESULTS: No significant difference was found between the single- and dual-bundle reconstructions in isokinetic muscle strength values. No statistically significant difference was detected between the tunnel expansions in 2nd, 3rd and 6th month tomographies following single- and dual-bundle ACL reconstruction. CONCLUSION: Single- and dual-bundle ACL reconstructions seem to provide similar results in terms of early tunnel enlargement and isokinetic muscle strength.

[Elastofibroma dorsi: an unusual cause of shoulder pain]. / Elastofibroma dorsi: Nadir görülen bir omuz agrisi nedeni.

OBJECTIVES: We evaluated patients who underwent surgical treatment for elastofibroma dorsi (ED). METHODS: The study included 13 patients (11 women, 2 men; mean age 54 years; range 43 to 74 years) who were treated surgically for ED that caused persistent symptoms. Involvement was on the right in six patients, on the left in five patients, and bilateral in two patients. All the patients presented with a mass lesion that became apparent at the lower corner of the scapula on shoulder flexion and adduction. The complaints were swelling and pain on the back in nine patients, and a snapping sound on shoulder movements together with pain in four patients. Diagnosis of ED was made by magnetic resonance imaging (n=10) and computed tomography (n=3), with no utilization of preoperative biopsy. Marginal tumor excision was performed in all cases. Evaluation for recurrence was made by ultrasonography. The mean follow-up period was 32 months (range 8 to 90 months). RESULTS: All the masses were located at the inferior corner of the scapula, with adherence to the thorax between the serratus anterior, rhomboid, and latissimus dorsi muscles. The mean size of the surgical specimens was 9 x 6 x 3 cm (range 5 x 3 x 1 to 14 x 8 x 3 cm). Clinical diagnosis was confirmed by histopathologic examination in all cases. All major complaints resolved after surgery. Hematoma occurred in four cases postoperatively, but resolved without the need for surgical intervention. No recurrence was observed. CONCLUSION: Even though ED is a rare clinic entity, it should be recalled while evaluating shoulder pathologies. Marginal excision is adequate for the treatment of patients with sustaining complaints.