Deep Learning-Based Automated Magnetic Resonance Image Segmentation of the Lumbar Structure and Its Adjacent Structures at the L4/5 Level

(1) Background: This study aims to develop a deep learning model based on a 3D Deeplab V3+ network to automatically segment multiple structures from magnetic resonance (MR) images at the L4/5 level. (2) Methods: After data preprocessing, the modified 3D Deeplab V3+ network of the deep learning model was used for the automatic segmentation of multiple structures from MR images at the L4/5 level. We performed five-fold cross-validation to evaluate the performance of the deep learning model. Subsequently, the Dice Similarity Coefficient (DSC), precision, and recall were also used to assess the deep learning model's performance. Pearson's correlation coefficient analysis and the Wilcoxon signed-rank test were employed to compare the morphometric measurements of 3D reconstruction models generated by manual and automatic segmentation. (3) Results: The deep learning model obtained an overall average DSC of 0.886, an average precision of 0.899, and an average recall of 0.881 on the test sets. Furthermore, all morphometry-related measurements of 3D reconstruction models revealed no significant difference between ground truth and automatic segmentation. Strong linear relationships and correlations were also obtained in the morphometry-related measurements of 3D reconstruction models between ground truth and automated segmentation. (4) Conclusions: We found it feasible to perform automated segmentation of multiple structures from MR images, which would facilitate lumbar surgical evaluation by establishing 3D reconstruction models at the L4/5 level.

The role of 3-dimensional preoperative planning for primary total hip arthroplasty based on artificial intelligence technology to different surgeons: A retrospective cohort study

Preoperative planning with computed tomography (CT)-based 3-dimensiona (3D) templating has been achieved precise placement of hip components. This study investigated the role of the software (3-dimensional preoperative planning for primary total hip arthroplasty [THA] based on artificial intelligence technology, artificial intelligence hip [AIHIP]) for surgeons with different experience levels in primary THA. In this retrospective cohort study, we included patients, who had undergone THA with the help of the AIHIP, and matched to patients, who had undergone THA without the help of the AIHIP, by age and the doctor who operated on them. The subjects were divided into 4 groups, senior surgeon (Chief of Surgery) with AIHIP group, senior surgeon without AIHIP group, junior surgeon (Associate Chief of Surgery) with AIHIP group and junior surgeon without AIHIP group. The general data, imaging index, clinical outcomes and accuracy of stem size prediction and cup size prediction were retrospectively documented for all patients. There was a significant difference in discrepancy in leg length (P = .010), neck-shaft angle (P = .025) and femoral offset (P = .031) between the healthy side and the affected side, operation duration (P < .001), decrease in hemoglobin (Hb) per 24 hours (P = .046), intraoperative radiation exposure frequency (P < .050) and postoperative complications (overall P = .035) among the patients in junior surgeon group. No significant differences were found between senior surgeon groups with respect to discrepancy in leg length (P = .793), neck-shaft angle (P = .088)and femoral offset (P = .946) between the healthy side and the affected side, operation duration (P = .085), decrease in Hb per 24 hours (P = .952), intraoperative radiation exposure frequency (P = .094) and postoperative complications (overall P = .378). The stem sizes of 95% were accurately estimated to be within 1 stem size, and 97% of the cup size estimates were accurate to within 1 cup size in senior surgeon group with AIHIP. A total of 87% stem sizes were accurately estimated to be within 1 stem size, and 85% cup sizes were accurate to within 1 cup size in junior surgeon group with AIHIP. In conclusion, our study suggests that an AI-based preoperative 3D planning system for THA is a valuable adjunctive tool for junior doctor and should routinely be performed preoperatively.

3D Reconstruction of Wrist Bones from C-Arm Fluoroscopy Using Planar Markers

In orthopedic surgeries, such as osteotomy and osteosynthesis, an intraoperative 3D reconstruction of the bone would enable surgeons to quickly assess the fracture reduction procedure with preoperative planning. Scanners equipped with such functionality are often more expensive than a conventional C-arm fluoroscopy device. Moreover, a C-arm fluoroscopy device is commonly available in many orthopedic facilities. Based on the widespread use of such equipment, this paper proposes a method to reconstruct the 3D structure of bone with a conventional C-arm fluoroscopy device. We focus on wrist bones as the target of reconstruction in this research as this will facilitate a flexible imaging scheme. Planar markers are attached to the target object and are tracked in the fluoroscopic image for C-arm pose estimation. The initial calibration of the device is conducted using a checkerboard pattern. In general, reconstruction algorithms are sensitive to geometric calibration errors. To assess the practicality of the method for reconstruction, a simulation study demonstrating the effect of checkerboard thickness and spherical marker size on reconstruction quality was conducted.

Are 3D Printing Templates an Advantage in Upper Thoracic Pedicle Screw Fixation?

Background This study aims to compare the clinical results of patients with upper thoracic vertebral fractures treated with pedicle screw and posterior spinal fusion with preoperative surgical planning and 3-dimensional (3D) modeling and patients treated with freehand screws. Methods Fifty patients who underwent pedicle screw placement with a diagnosis of upper thoracic fracture between June 2018 and October 2020 were included in our study. Pedicle screws were used in 25 patients (group 1) after the planning was completed with the help of 3D preoperative printing and modeling. Pedicle screws were applied in 25 patients in the control group (group 2) using the freehand technique. Intraoperative bleeding amount, operation time, and correct screw placement data in both groups were recorded. Results The operation time was 134 ± 22 minutes for group 1 and 152 ± 38 minutes for group 2. The difference in operation times was found to be statistically significant (p < 0.05). Based on axial and sagittal reconstruction images, the accuracy rate of pedicle screw placement (grades 0 and 1) in group I was 96.6% compared to 83.6% in group II. The minor perforation rate (grade 1, <2 mm) was 5.8% in group I compared to 11.8% in group II. The moderate perforation rate (grade 2, 2-4 mm) was 3.4% in group I compared to 14% in group II. The severe perforation rate (grade 3, >4 mm) was 2.3% in group II; however, misplaced screws were not associated with neurological deficits. The difference in overall accuracy rates between the two groups was significant (p < 0.05). Conclusions For 3D models of upper thoracic pedicle screw insertion, guide plates can be produced inexpensively and individually. It provides a new method for the accurate placement of upper thoracic pedicle screws with high accuracy and secure use in screw insertion.

[Research and application of artificial intelligence based three-dimensional preoperative planning system for total hip arthroplasty]

OBJECTIVE

To develop an artificial intelligence based three-dimensional (3D) preoperative planning system (AIHIP) for total hip arthroplasty (THA) and verify its accuracy by preliminary clinical application.

METHODS

The CT image database consisting of manually segmented CT image series was built up to train the independently developed deep learning neural network. The deep learning neural network and preoperative planning module were assembled within a visual interactive interface-AIHIP. After that, 60 patients (60 hips) with unilateral primary THA between March 2017 and May 2020 were enrolled and divided into two groups. The AIHIP system was applied in the trial group ( n=30) and the traditional acetate templating was applied in the control group ( n=30). There was no significant difference in age, gender, operative side, and Association Research Circulation Osseous (ARCO) grading between the two groups ( P>0.05). The coincidence rate, preoperative and postoperative leg length discrepancy, the difference of bilateral femoral offsets, the difference of bilateral combined offsets of two groups were compared to evaluate the accuracy and efficiency of the AIHIP system.

RESULTS

The preoperative plan by the AIHIP system was completely realized in 27 patients (90.0%) of the trial group and the acetate templating was completely realized in 17 patients (56.7%) of the control group for the cup, showing significant difference ( P<0.05). The preoperative plan by the AIHIP system was completely realized in 25 patients (83.3%) of the trial group and the acetate templating was completely realized in 16 patients (53.3%) of the control group for the stem, showing significant difference ( P<0.05). There was no significant difference in the difference of bilateral femoral offsets, the difference of bilateral combined offsets, and the leg length discrepancy between the two groups before operation ( P>0.05). The difference of bilateral combined offsets at immediate after operation was significantly less in the trial group than in the control group ( t=-2.070, P=0.044); but there was no significant difference in the difference of bilateral femoral offsets and the leg length discrepancy between the two groups ( P>0.05).

CONCLUSION

Compared with the traditional 2D preoperative plan, the 3D preoperative plan by the AIHIP system is more accurate and detailed, especially in demonstrating the actual anatomical structures. In this study, the working flow of this artificial intelligent preoperative system was illustrated for the first time and preliminarily applied in THA. However, its potential clinical value needs to be discovered by advanced research.

Application of optimized digital surgical guides in mandibular resection and reconstruction with vascularized fibula flaps: Two case reports

RATIONALE

Currently, digital surgical techniques have been widely used in the precise treatment of mandibular resection and reconstruction with fibula flaps. Utilizing these innovative techniques in surgical planning and hardware fabrication before surgery has shown to provide great help. However, it is difficult for even experienced surgeons to place the preformed reconstruction plate in the same position as its preoperative design, causing surgical results to differ from preoperative planning. This study aims to solve these acknowledged challenges by creating newly designed equipment.

PATIENT CONCERNS

Two patients suffering from long-term expansion of the mandible were admitted to our department. Case I was a 39-year-old female patient who was concerned about the disease in the middle of the mandible, Case II was a 45-year-old female patient who was concerned about the disease at the left mandibular angle and ramus region.

DIAGNOSES

Two patients were diagnosed with the mandibular ameloblastoma based on computed tomography (CT) scan and pathological results.

INTERVENTIONS

Personalized 3-dimensional (3D) surgical guides were applied to 2 patients with mandibular ameloblastoma who underwent mandibular resection and reconstruction with vascularized fibula flaps using a specially optimized and designed reconstruction guide plate.

OUTCOMES

We achieved precise mandibular repair with such a guide in full accordance with the preoperative plan and ensured the restoration of patient facial symmetry.

LESSONS

Optimized reconstruction guide template could accurately locate the preformed reconstruction plate. This component had the ability to ensure that the location of the actual reconstruction plates were highly consistent with preoperative designed models.

Computer navigation re-creates planned glenoid placement and reduces correction variability in total shoulder arthroplasty: an in vivo case-control study

BACKGROUND

Accurate glenoid component placement is important to prevent glenoid component failure in total shoulder arthroplasty (TSA). Navigation may reduce the variability of glenoid component version and inclination; therefore, the aims of this study were to determine, in patients undergoing TSA, whether computer navigation improved the ability to achieve neutral postoperative version and inclination, as well as achieve the individualized preoperative plan.

METHODS

Patients undergoing TSA using navigation (computer-assisted surgery [CAS], n = 33) or the conventional technique (n = 27) from January 2014 to July 2017 were recruited and compared. Preoperative and postoperative version and inclination, as well as postoperative inferior overhang, were measured using computed tomography scans.

RESULTS

The CAS group had more than twice as many augmented glenoid components as the conventional group (45.5% vs. 19.2%). CAS significantly reduced the between-patient variability in postoperative version and led to a greater proportion of components positioned in "neutral" alignment for both inclination and version (P < .015). The incidence of neutral inclination or version postoperatively was significantly higher in the CAS group, and the glenoid was implanted within 5° of the surgical plan in more than 70% of cases, with more than 40% displaying no detectable difference.

CONCLUSION

An integrated system of 3-dimensional surgical planning, augmented glenoid components, and intraoperative navigation may reduce the risk of glenoid placement outside of a neutral position in patients undergoing TSA compared with conventional methods. This study demonstrated the capacity for CAS to replicate the surgical plan in a majority of cases.