Objective Owing to the shortcomings of clinical examination and radiographs, injury to the syndesmotic ligaments is often misdiagnosed. When there is no indication requiring that the fractured ankle be operated on, the syndesmosis is not tested intra-operatively, and rupture of this ligamentous complex may be missed. Subsequently the patient is not treated properly leading to chronic complaints such as instability, pain, and swelling. We evaluated three fracture classification methods and radiographic measurements with respect to syndesmotic injury. Materials and methods Prospectively the radiographs of 51 consecutive ankle fractures were classified according to Weber, AO-Müller, and Lauge-Hansen. Both the fracture type and additional measurements of the tibiofibular clear space (TFCS), tibiofibular overlap (TFO), medial clear space (MCS), and superior clear space (SCS) were used to assess syndesmotic injury. MRI, as standard of reference, was performed to evaluate the integrity of the distal tibiofibular syndesmosis. The sensitivity and specificity for detection of syndesmotic injury with radiography were compared to MRI. Results The Weber and AO-Müller fracture classification system, in combination with additional measurements, detected syndesmotic injury with a sensitivity of 47% and a specificity of 100%, and Lauge-Hansen with both a sensitivity and a specificity of 92%. TFCS and TFO did not correlate with syndesmotic injury, and a widened MCS did not correlate with deltoid ligament injury. Conclusion Syndesmotic injury as predicted by the Lauge-Hansen fracture classification correlated well with MRI findings. With MRI the extent of syndesmotic injury and therefore fracture stage can be assessed more accurately compared to radiographs.
Early detection improves prognosis in pancreatic ductal adenocarcinoma (PDAC), but is challenging as lesions are often small and poorly defined on contrast-enhanced computed tomography scans (CE-CT). Deep learning can facilitate PDAC diagnosis; however, current models still fail to identify small (<2 cm) lesions. In this study, state-of-the-art deep learning models were used to develop an automatic framework for PDAC detection, focusing on small lesions. Additionally, the impact of integrating the surrounding anatomy was investigated. CE-CT scans from a cohort of 119 pathology-proven PDAC patients and a cohort of 123 patients without PDAC were used to train a nnUnet for automatic lesion detection and segmentation (nnUnet_T). Two additional nnUnets were trained to investigate the impact of anatomy integration: (1) segmenting the pancreas and tumor (nnUnet_TP), and (2) segmenting the pancreas, tumor, and multiple surrounding anatomical structures (nnUnet_MS). An external, publicly available test set was used to compare the performance of the three networks. The nnUnet_MS achieved the best performance, with an area under the receiver operating characteristic curve of 0.91 for the whole test set and 0.88 for tumors <2 cm, showing that state-of-the-art deep learning can detect small PDAC and benefits from anatomy information.
The median laparotomy is frequently used by abdominal surgeons to gain rapid and wide access to the abdominal cavity with minimal damage to nerves, vascular structures and muscles of the abdominal wall. However, incisional hernia remains the most common complication after median laparotomy, with reported incidences varying between 2-20%. Recent clinical and experimental data showed a continuous suture technique with many small tissue bites in the aponeurosis only, is possibly more effective in the prevention of incisional hernia when compared to the common used large bite technique or mass closure. The STITCH trial is a double-blinded multicenter randomized controlled trial designed to compare a standardized large bite technique with a standardized small bites technique. The main objective is to compare both suture techniques for incidence of incisional hernia after one year. Secondary outcomes will include postoperative complications, direct costs, indirect costs and quality of life. A total of 576 patients will be randomized between a standardized small bites or large bites technique. At least 10 departments of general surgery and two departments of oncological gynaecology will participate in this trial. Both techniques have a standardized amount of stitches per cm wound length and suture length wound length ratio's are calculated in each patient. Follow up will be at 1 month for wound infection and 1 year for incisional hernia. Ultrasound examinations will be performed at both time points to measure the distance between the rectus muscles (at 3 points) and to objectify presence or absence of incisional hernia. Patients, investigators and radiologists will be blinded during follow up, although the surgeon can not be blinded during the surgical procedure. The STITCH trial will provide level 1b evidence to support the preference for either a continuous suture technique with many small tissue bites in the aponeurosis only or for the commonly used large bites technique. Clinicaltrials.gov NCT01132209
Objective The optimal MRI scan planes of collateral ligaments of the ankle have been described extensively, with the exception of the syndesmotic ligaments. We assessed the optimal scan plane for depicting the distal tibiofibular syndesmosis. Materials and Methods In order to determine the optimal oblique caudal-cranial and lateral-medial MRI scan plane, two fresh frozen cadaveric ankles were used. The angle of the scan plane that demonstrated the anterior and posterior distal tibiofibular ligament uninterrupted in their full length was determined. In a prospective study this oblique scan plane was then used in addition to the axial and coronal planes, for MRI scans of both ankles in 21 healthy volunteers. Two observers independently evaluated the anterior tibiofibular ligament (ATIFL) and posterior tibiofibular ligament (PTIFL) regarding the continuity of the individual fascicles, thickness and wavy contour of the ligaments in both the axial and the oblique plane. Kappa was calculated to determine the interobserver agreement. McNemar’s test was used to statistically quantify the significance of the two scan planes. Results In the axial plane the ATIFL was in 31% (13/42) partly and in 69% (29/42) completely discontinuous; in the oblique plane the ATIFL was continuous in 88% (37/42) and partly discontinuous in 12% (5/42). Compared with the axial plane, the oblique plane demonstrated significantly less discontinuity (p < 0.001), but not significantly less thickening (p = 1.00) or less wavy contour (p = 0.06) of the ATIFL. In the axial scan plane the PTIFL was continuous in 76% (32/42), partially discontinuous in 19% (8/42) and completely discontinuous in 5% (2/42); in the oblique plane the PTIFL was continuous in 100% (42/42). Compared with the axial plane, the oblique plane demonstrated significantly less discontinuity (p = 0.002), but not significantly less thickening (p = 1.00) or less wavy contour (p = 0.50) of the PTIFL. The interobserver agreement score and kappa (κ) regarding the continuity for the ATIFL in the axial and oblique planes was 91% (κ = 0.79) and 91% (κ = 0.55) respectively; for the PTIFL it was 86% (κ = 0.65) and 100% (κ = not defined). Conclusion The ATIFL and PTIFL are routinuely scanned in the orthogonal planes. The advantage of MRI scanning in an oblique image plane of about 45 degrees permits a better evaluation of the ligaments compared with the axial plane, particularly a better interpretation of ligament continuity, thickening and wavy contour. This may lead to a reduction in false-positive results, especially regarding partial or complete ligament ruptures. This can be of considerable aid in therapeutic management.