Search

Category: Ankle; Trauma Introduction/Purpose: The use of artificial intelligence (AI) is particularly salient to visually oriented medical professions, especially orthopedics. The most prominent use of AI in orthopedics comes in the form of medical imaging examinations. AI has a huge potential to help doctors make diagnoses by acting as a second pair of eyes. Some results suggested a very high level of agreement between AI models’ and the clinician’s assessments of radiographs. Furthermore, it was also suggested that the sensitivity and specificity of emergency medicine physicians for detecting some pathologies are significantly improved when aided by an AI tool. In view of these observations, our study aimed to create an AI-based ankle fracture detection tool that can be used on smart devices for X-ray interpretation. Methods: We examined 2,193 patients’ charts from 2 academic and 1 community hospital in Boston. We retrieved the anteroposterior (AP), oblique, and lateral ankle X-rays of each patient. Patients with ankle fractures and adults older than 18 years met our inclusion criteria. We excluded patients younger than 18 years old and those with any artifact, such as a cast, screws, or other artifacts in their X-rays. The study comprised 352 healthy controls and a total of 579 ankle fracture patients. Other than the digital images obtained from Electronic patient records (EPR), we used two different smart devices, a cellphone, and a tablet, to capture images from the monitor screen. Using Machine Learning models, we developed a fracture detection model using all three types of imaging and named it the “combination model”. We subsequently tested the combination model on digital X-rays, smart devices, and on both datasets together (Table 1). Results: We extracted the X-rays of a total of 931 patients in this study. Following the development and testing of our AI models, we noticed that all performed well with AUCs and accuracies above 0.85 and 0.86, respectively (Table1). The best performance was found when the combination model was tested on images taken from the camera of our smart devices, with an AUC of 0.88, a sensitivity of 0.86, and an accuracy of 0.89. This performance was closely followed by that of our model tested on a mix of both smart devices and original digital images with an AUC of 0.88, a sensitivity of 0.86, and an accuracy of 0.88. Conclusion: Our AI-based tool showed promising performance in the detection of ankle fractures using smart devices and images obtained from the monitor screen. We were able to reach an accuracy of diagnosis on smart device-captured images that were comparable with the original digital X-rays. The outcome of this study can be used to help providers who lack sufficient experience in detecting fractures. It can also be used for educational purposes for trainees in this field.

Category: Ankle; Trauma Introduction/Purpose: Chaput tubercle fractures, which are thought to represent tibial-sided avulsions of the anterior inferior tibiofibular ligament (AITFL), are prevalent in up to 30% of trimalleolar ankle fractures. The optimal treatment of small Chaput avulsions is debatable; direct fixation with suture anchor devices and indirect (syndesmotic) fixation are considered as viable options, with proponents on both sides. Moreover, recent literature highlights the potential anatomical alterations to the incisura tibialis resulting from malreduction of large Chaput fragments, furthering the case for direct fixation. Hence, we performed a CT- based three-dimensional fracture mapping study to identify the morphological characteristics of these fractures, and to determine whether they consistently involve the AITFL, tibial plafond and incisura tibialis. Methods: This study included adult patients who had an ankle fracture with a Chaput component; the scheme described by Rammelt et al. was used to classify these fractures. CT scans were obtained, and 3D models were generated. The models were superimposed over a statistical shape model of the right tibia which served as a template and fracture lines were marked. The footprints of proximal and main bands of the AITFL and Basset’s ligament were also marked on the template tibia. The tibial template along with the fracture lines was then imported into MATLAB, and an automated script was used to determine the fragment size (length, breadth, and height), fracture surface area, involvement of the tibial plafond, tibial incisura, and the anterior inferior tibiofibular ligament (AITFL) and Basset’s ligament. Fracture maps and heat maps were generated. Agglomerative cluster analysis using Ward’s linkage was used to identify discrete fracture categories. Results: 76 patients, 21 males and 55 females were included in this study. Cluster analysis identified two distinct groups of fractures, each with two unique subgroups. We present this as a modification of the existing classification system. The first group, corresponded to Rammelt Type 1 fractures (sub centimetric extra-articular avulsion fractures, n=47). Of these, 19% (n=9) did not involve the AITFL, which we termed as Type 1a, and 91% (n=48) involved the AITFL, which we termed as Type 1b. The second group consisted of large intra-articular fractures that corresponded to Rammelt Type 2 injuries. Of these 23% (n=6) involved only the incisura, which we termed as Type 2a; 77 % involved both the incisura and the tibial plafond and were termed as Type 2b. Conclusion: We propose a modification of the existing classification of Chaput fractures on the basis of quantitative fracture mapping. This study provides new insights into the morphological characteristics of Chaput fractures. 19% of small Chaput fractures do not involve the AITFL and may not require direct fixation. Conversely, all large-sized fragments involve the incisura and necessitate anatomical reduction to achieve accurate syndesmotic reduction. Our proposed modification can aid in surgical decision-making, particularly in choosing between direct and indirect syndesmotic fixation.

Category: Ankle; Trauma Introduction/Purpose: The term ‘quadrimalleolar fracture’ (QMF) describes a trimalleolar (TM) ankle fracture with an associated Chaput or Wagstaffe fracture. Optimal fixation of these injuries not only allows for bone-to-bone healing but also facilitates syndesmotic reduction. However, this is a relatively new concept, and the literature on QMFs is limited to a few case series. Hence, we conducted this study to determine the epidemiology, patterns, and clinicoradiologic correlations of QMFs. The primary objective of this study was to describe the prevalence and patterns of QMFs. The secondary objective was to determine if certain clinical or radiological parameters were associated with different types and patterns of QMFs. Methods: A retrospective analysis of ankle fractures presenting to three tertiary referral hospitals was undertaken. Adult patients (≥18 years) presenting with an acute, traumatic TM ankle fracture over 3 years (July 2018 to July 2021) were included. Isolated medial and lateral malleolar fractures, bimalleolar ankle fractures, pediatric patients, pilon fractures, pathological fractures, and those with delayed presentation or (≥3 weeks of injury) were excluded. Clinical demographic variables were obtained from the patients’ medical records. AP, lateral, and mortise ankle radiographs, and CT scans, (axial, coronal, and sagittal sections along with 3D volume reconstructed models) whenever available, were evaluated in detail. Demographics and radiological parameters were compared between TM and QMF, as well as between different types of QMFs. Odds ratios with 95% confidence intervals were determined to test the strength of association. Results: A total of 876 adult patients with ankle fractures were included after the screening, of which 323 had a TM ankle fracture for which a CT scan had been performed. A total of 159 AITFL avulsions were identified, yielding a prevalence of 18.2% amongst all ankle fractures and 26.1% in ankle fractures where a CT scan had been performed. TMFs had the significantly highest proportion of CT-confirmed AITFL avulsions (44.4%) in comparison to unimalleolar (4.4%) and bimalleolar fractures (6.4%) (P < 0.0001). Age and osteoporosis were significant associations of QMFs. Avulsion of the medial malleolus, Weber B fibular fracture, and supination external rotation mechanism were significantly associated with Wagstaffe fractures. Size of the Chaput fracture was inversely related to that of the posterior malleolar fracture. Conclusion: Quadrimalleolar ankle fractures account for a significant proportion of ankle fractures. The strengths of this study include a large sample size, which was derived from three different hospitals, consecutive inclusion (or exclusion) of cases, strict adherence to the STROBE guidelines, and the fact that only CT-confirmed cases were used to determine clinic-radiological associations. Wagstaffe and Chaput fractures have distinct clinical and radiological correlations. However, further research is needed to determine the optimal fixation protocols for these injuries.

Category: Midfoot/Forefoot; Other Introduction/Purpose: Up to one-third of Lisfranc injuries are missed on initial evaluation, resulting in significant morbidity. Prompt diagnosis of Lisfranc injuries is, therefore, critical in optimizing outcomes, and yet there remains a lack of consensus on diagnostic criteria to identify Lisfranc instability using anatomic measurements. Prior studies have underscored the utility of weight-bearing computed tomography (WBCT) in diagnosing Lisfranc instability, which allows for bilateral three-dimensional (3D) evaluation under physiologic load. The aim of this study was to quantify appropriate cutoff values using 3D anatomic measurements of Lisfranc instability under physiologic load and as compared to the uninjured extremity. Methods: A total of 234 adult patients with bilateral WBCT scans of the foot were included: 43 patients with Lisfranc instability & 191 patients without a history of midfoot injuries. Lisfranc instability was confirmed intraoperatively or by clinical evaluation by the surgeon. Patients with prior midfoot surgery, Charcot arthropathy, severe midfoot arthropathy, or significantly displaced fracture of the cuneiforms or 1st, 2nd, or 3rd metatarsal bases were excluded. Lisfranc joint measurements were conducted bilaterally, including C1-M2 diastasis, C1-M2 area and 3D WBCT volume. Patient demographics were collected. Descriptive statistics were calculated for quantitative variables. Percent difference as compared to the contralateral side was calculated for volume, area, and diastasis. Mann Whitney U tests were utilized to determine differences in measurements between patients in the Lisfranc cohort and patients in the uninjured cohort (p-value < 0.05). Diagnostic cutoffs for Lisfranc instability were determined with minimum distance to the corner analysis on ROC curves. Results: The distribution of patients with Lisfranc instability was 58% female with an average age of 38.5±17.6 years. Of the Lisfranc injuries, 36% were purely ligamentous, 18% demonstrated an avulsion injury, and 43% involved a metatarsal base fracture. In the Lisfranc cohort, the median percent difference between injured and uninjured feet was 28.2% (IQR: 20.9%) for volume, 36.3% (IQR: 31.1%) for C1-M2 area, and 40.0% (IQR: 23.3%) for C1-M2 diastasis. Mann-Whitney testing was significant for percent difference between left and right feet for Lisfranc patients versus patients with an uninjured midfoot (p < 0.001). The area under the curve and associated diagnostic cutoffs for Lisfranc instability, were 0.81 and 18% for volume, 0.84 and 20% for area, and 0.91 and 28% for diastasis (Figure 1). Conclusion: Diastasis (1D), area (2D), and volume (3D) measurements are effective measurements to diagnose Lisfranc instability on WBCT. With current diagnostic algorithms, however, 1D measurements afford superior diagnostic sensitivity as compared to 2D and 3D measurements when using WBCT, suggesting Lisfranc instability may be best appreciated in the axial plane. Additional studies are necessary to expand the sample population to assess for corroboration with the current results, especially for subtle Lisfranc instability.

Category: Midfoot/Forefoot; Other Introduction/Purpose: Lisfranc instability is often a challenging injury to diagnose, with up to one-third being missed on initial evaluation. Lisfranc instability is assessed by widening of the space between the second metatarsal base and the medial cuneiform. This space is visualized on X-Ray; however, in subtle Lisfranc instability cases computed tomography (CT) imaging is often obtained. Given the 1D nature of diastasis measurement, X-ray should be an adequate means of evaluating this widening, yet clinical practice suggests weightbearing CT (WBCT) is more sensitive. This suggests the 3D location of the diastasis measurement is crucial. This study aimed to first compare weight-bearing X-ray and WBCT diastasis and area measurements of the Lisfranc complex, and second to compare WBCT diastasis measurements at various locations in the coronal plane. Methods: A total of 90 patients with both weight-bearing foot X-ray and bilateral foot WBCT were included: 37 patients had confirmed Lisfranc instability, and 53 patients had no history of midfoot injury. Lisfranc instability was confirmed intraoperatively or by clinical examination performed by the surgeon. For all 90 patients, the interosseous diastasis and area between the medial cortex of the second metatarsal and the lateral cortex of the medial cuneiform were measured on both weightbearing X-ray and axial slices of WBCT. For a subset of patients (12 in each cohort) the diastasis between second metatarsal and medial cuneiform was measured at 4 distinct axial locations using coronal slices of the WBCT (Figure 1). Results: A Wilcoxon test comparison of diastasis and area measurements on weightbearing X-ray and axial slices of WBCT revealed a significant difference in the weightbearing X-ray and WBCT measurements (p < 0.001) for both the control and Lisfranc cohorts. A comparison of weightbearing X-ray diastasis to dorsal, interosseous, and plantar diastasis measurements on coronal slices using Kruskal Willis analysis for the control cohort revealed a significant difference in all three measurements (Figure 1, p-values listed). Kruskal Willis analysis of the Lisfranc cohort demonstrated a significant difference between the X-ray diastasis and the plantar WBCT diastasis (p=0.01), but no difference from interosseous (p=0.08) or dorsal (p=0.33) diastases. A comparison of the axial WBCT diastasis measurements compared to the three coronal diastasis measurements revealed no difference in either cohort. Conclusion: Our results showed that for assessing the Lisfranc joint in subtle cases and healthy individuals, WBCT remains the most accurate imaging tool. However, in patients with confirmed Lisfranc instability, X-ray measurements are more reliable if the instability is at the dorsal and interosseous levels and not the plantar level. Hence, to assess the Lisfranc joint at different levels, WBCT has superiority over X-ray. Additionally, the axial diastasis of the Lisfranc joint on WBCT seems similar to coronal diastasis rendering both views reliable for measurements.

Category: Ankle; Other Introduction/Purpose: Geospatial access to healthcare is defined as the ability of patients to obtain healthcare services based on their locations. Therefore, we aimed to investigate patients’ proximity to healthcare and its correlation with the complications of surgically treated Achilles tendon rupture (ATR), including venous thromboembolism (VTE), re-rupture, and wound dehiscence. The outcome of this study can help providers and policy makers to determine the equity of healthcare for these patients in the area under their coverage. Methods: We included 426 patients who lived in the United States (US) tri-state area who received surgical treatment for ATR. We used patient and hospital addresses and zip codes to calculate the distances to healthcare centers. A geographic information system (GIS) platform analysis in the tri-state was completed using Maptive, powered by Google Maps, to create heat maps displaying the geographical distribution of patients treated for ATR with and without complications (Figure 1). The Shapiro–Wilk test was used to determine normal distribution. Mann–Whitney U test was used to compare the groups with and without complication. The point biserial correlation test was used to determine any correlations between driving distance and the incidence of complications (p < 0.05 was considered significant). Results: The average driving distance to the patient’s specific healthcare center was 62.16 ± 76.54 km. There was no significant difference between the distances for patients with and without overall complications (p = 0.65), with and without VTE (p = 0.70), with and without re-rupture (p = 0.84), and with and without wound problems (p = 0.36). No correlation between complications and the distance to healthcare centers was found (p = 0.65; Table 1). Conclusion: Geospatial information is important within the context of healthcare accessibility and can provide crucial guidance to healthcare planning for patients and healthcare policymakers. Although this study showed that distance to healthcare facilities did not lead to significantly higher complication rates amongst ATR patients, it does not resolve the need for further studies looking at a larger population and broader geographical segments. These studies can be used to assess the equity of healthcare in specific areas and make modifications if needed.

Category: Midfoot/Forefoot; Trauma Introduction/Purpose: Untreated Lisfranc injuries can lead to chronic pain, midfoot arthritis, and functional disability 1,2,5. Approximately 20% of Lisfranc injuries are misdiagnosed or completely missed on initial evaluation, which can be attributed to a lack of parameters of what is within normal limits for subjects without injury in the Lisfranc complex3,4. The purpose of this study is to identify anatomic variations of unidimensional, bidimensional, and tridimensional (1D, 2D, 3D) measurements of the Lisfranc complex of normal individuals using weight-bearing computed tomography (WBCT) imaging. Our hypothesis is that there is some variation among individuals with a difference between right and left of the same individual of less than 10%. Methods: A total of 191 subjects with bilateral WBCT scans of the foot were collected from three tertiary medical centers from 2019-2022. Exclusion criteria included: history of Lisfranc injury, first to fourth metatarsal base fractures, Charcot arthropathy, midfoot arthritis, cuneiform fractures, and forefoot surgery proximal to the metatarsal neck region. The following Lisfranc joint measurements were collected6 bilaterally: joint volume (3D), area of the joint on a consistent axial slice (2D), and distance between the second metatarsal and medial cuneiform (1D). Foot alignment was assessed using lateral talar-first metatarsal angle (Meary’s Angle). Patient demographics (age, sex, weight, height, BMI) were also collected. Descriptive statistics were calculated for quantitative variables. Percent difference was used to assess volume, area, and distance variation between sides.7 Correlation between demographic data and Lisfranc measurements was assessed using Pearson Correlation tests. A T-test was used for categorical variables. A p-value < 0.05 was considered statistically significant. Results: The cohort included 61% female and 39% male with an average age of 45.2±17.3 years. The mean volume, area, and distance measurements of the Lisfranc joint for both left and right sides are shown in Figure 1. Overall, the mean percent difference between left and right sides were 11.9%, 14.9% and 13.9% for volume, area and diastasis, respectively (Figure 1). No association was found between age, BMI, or weight and the volume, area or distance measurements. A correlation (r=0.48, p= < 0.001) was found between height and left Lisfranc joint volume, but no other measurements. Significant difference between sex in volume (p < 0.001) and area (p < 0.001) measurements were found. Foot alignment did not have a correlation with a variation of the Lisfranc complex measurements. Conclusion: This study provides an objective characterizing of the anatomic variations of the Lisfranc joint amongst healthy individuals. Differences in the measurements and the large standard deviation in absolute measurements of volume, area, and distance suggests that percent difference with the contralateral side may be a better metric to use to diagnose Lisfranc instability. Our results showed that the difference between bilateral sides in a patient is less than 15% for volume, area, and diastasis. The results of this study can set the foundation for future studies to determine the cut-off value for the diagnosis of Lisfranc instability using WBCT imaging.