Failure Of Segmentation Research Articles

Notice of Removal: IAT: A Full-Scale IVUS Analysis Toolbox Based on Deep Learning Algorithms for Clinical Diagnosis

Intravenous ultrasound (IVUS) produces important imaging information for the clinical medical detection of intravascular lesions. However, most information hidden in the IVUS image must be manually annotated by doctors and obtained by professional parameter calculation tools, which is time-consuming and labor-intensive. The existing IVUS segmentation algorithms generally ignore the positive effect of expert prior knowledge on the segmentation results. We, thus, develop a full-scale IVUS Analysis Toolbox (IAT) that can automatically extract more image information through the image segmentation model of intravascular ultrasound (ISM-IVUS), allowing experts to modify suboptimal segmentation results and support the calculation and acquisition of instantaneous parameters. The ISM-IVUS integrates the expert prior knowledge extraction module. By fusing edge operators based on intensity changes to obtain prior matrices and constraining the morphological features of the segmentation results, the failure of complex IVUS image segmentation can be effectively avoided. The model uses the U-Net framework and attention mechanism, combines deep and shallow features, and reduces the loss of important features in the segmentation process. To ensure the generalization and clinical applicability of the toolbox, we used a large-scale professional dataset provided by the hospital and a public dataset to conduct comparison experiments. The mIoU, the dice index, and the Hausdorff distance of the segmentation results in IAT reached 0.8616, 0.9343, and 6.0799, respectively, which verified the effectiveness of the model. In the comparison experiments of the number of parameters, FLOPs, and robustness, ISM-IVUS also has certain advantages. The correlation and p-value of the calculated parameters also achieved good results. A series of experiments have shown that IAT is in line with clinical needs and provides an application basis for the clinical practice of subsequent IVUS research.

An Active Contour Model Based on Local Pre-Piecewise Fitting Bias Corrections for Fast and Accurate Segmentation

The lack of grasp of the image information and the unstable fluctuation of the model energy may cause segmentation failure of the active contour model (ACM). Minimizing the impact of these two factors is critical. A local pre-piecewise fitting (LPPF) bias correction model for fast and accurate segmentation is proposed in this paper. It defines a pre-fitting function of local regions and an energy function. The grayscale information of small areas in the image is fully extracted, so that the contour accurately locates the target. Then, the optimal solution to the estimated value of the bias field is obtained. The real image information is described by the bias field, and the energy function of the model is constructed. The optimized distance regularized term and neighborhood average filtering method are utilized to achieve level set function regularization and contour smoothing. This optimization process reduces the amount of calculation and improves the robustness of LPPF model. Experiments are performed to verify that LPPF model has strong robustness to initial contours and has ability to segment blurry images while satisfactory segmentation efficiency and accuracy are obtained.

Flexural failure of the unrestrained segment for sleeved compression struts with pinned ends

Confining with an outer sleeve is an effective means of enhancing the load-bearing and deformation capacity of compression struts. The outer sleeve and compression strut, i.e., inner core, constitute a sleeved core-loaded member, where the latter ideally bears all the axial compression under the lateral restraint of the former. Additionally, an extension segment of the inner core is typically required as the allowance for contraction. However, the unrestrained extension segment is prone to excess flexural deformation, dramatically decreasing the overall load carrying capacity of the sleeved member. In this paper, first, an analytical investigation of the behaviour of the sleeved member with pinned ends is presented. Second, a numerical study was conducted to determine the limit state of the localized flexural failure and to propose the corresponding design criterion combining the analytical model. Last, the design criterion was further validated by compressive tests, indicating that the proposed method accurately predicted the localized​ flexural failure of both seamless-sleeved members and members with assembled outer sleeves.

Multiple Meningocele in Klippel-Feil Syndrome: Case Report

Klippel–Feil syndrome (KFS) is a rare congenital skeletal malformation characterized by the fusion of various cervical and occasionally thoracic and lumbar vertebrae. It may be seen as the fusion between two vertebras of all cervical vertebras, Congenital cervical fusion occurs as a result of failure in normal segmentation of cervical summits. Cause of this failure is unknown and occurs at second-eight weeks of gestation. 25 years old, patient presented to our counseling center with a 4-month history of progressive dysphagia. Ent exam showed a soft retropharyngeal mass displacing the left tonsil measuring a little about 6 cm with no sign of ulceration or local inflammation, a short neck was noted with a limitation of flexion and extension movement and low-lying hairline with a psychomotor retardation; which raises the suspicion of Klippel Feil syndrome undiagnosed. Cervical magnetic resonance imaging (MRI) was done which showed a cervical cystic formation compressive oval well limited with regular contours and fluid content with dark aspect in T1 weighted sequences and bright in T2 weighted sequences with regular wall enhanced after injection of gadolinium measuring 8.2*4.3 cm spread over 8.9 cm high This formation appears to continue with the subarachnoid spaces of the posterior cerebral fossa through the hypoglossal nerve canal with doubt about involvement of the jugular foramen. Thoracic CT scan demonstrated a lateral cystic mass exiting out of the intervertebral foramen of T6-T7 showing the same characteristics as the cervical formation above. Klippel-Feil syndrome is a rare congenital skeletal malformation characterized by fusion of several cervical and sometimes thoracic and lumbar vertebrae, leading to a low hairline and a short neck. The pathological feature is congenital fusion of two or more cervical vertebrae. The fusion is thought to be associated with reduced expression of the Pax gene in sclerotomes during chondrification of the vertebra. Klippel-Feil syndrome is usually diagnosed in childhood. However, associated abnormalities may delay diagnosis. The challenge for the specialist is to recognize concomitant abnormalities and perform the appropriate work-up for the diagnosis of Klippel-Feil syndrome.

The Early Effect of Anterior Cervical Discectomy and Fusion on Adjacent Segment Degeneration in Cases of Cervical Degenerative Disease: A Clinical and Retrospective Study

BACKGROUND: The cervical spine is the most mobile part of the spine in the sagittal plane. It is important for surgeons to have reliable, simple, and reproducible parameters to analyze the cervical spine pre-operative and post-operative. Despite anterior cervical discectomy and fusion (ACDF) is a common procedure, adjacent segment failure after surgery is an ongoing clinical concern; adjacent segment disease (ASD) occurs in approximately 3% of patients per year, with an expected incidence of 25% within the first 10 years following fusion. It has been suggested that the increased stress placed on adjacent segments after successful ACDF may increase the rate of symptomatic disc disease at those segments, so our study focuses on how to evaluate ASD using magnetic resonance imaging (MRI) and X-ray. METHODS: The retrospective study conducted on 72 patients in the department of neurosurgery at Kasr Al-Ainy hospitals to evaluate ASD after ACDF for degenerative cervical spondylosis using pre-operative and post-operative cervical spine X-ray lateral view as well as MRI for cervical spine after 2 years from the date of surgery. RESULTS: Among the 72 patients in our study, the follow-up MRI after 2 years showed new pathology in 14 patients (19.4%) while 58 patients (80.6%) showed no new pathology; also 20 patients (27.8%) presented with new complaints while 52 patients (72.2%) had no new complaints and only 3 patients (4.16%) were scheduled for surgery. CONCLUSION: ASD is an ongoing process of degeneration that could be accelerated by ACDF procedures, but not every adjacent segment degeneration or symptomatic disease requires additional surgery. Cervical mal-alignment greatly affects the neck pain rather than development of other new neurological manifestations and plays a role in adjacent segment failure which should be considered during cervical fusion surgeries. ASD must be differentiated from adjacent segment degeneration using clinical examination and imaging to help in patients’ follow-up and decision-making of further intervention.

Comparison of short-term effectiveness of Prodisc-C Vivo artificial disc replacement and Zero-P fusion for treatment of single-segment cervical spondylosis

To compare the short-term effectiveness and the impact on cervical segmental range of motion using Prodisc-C Vivo artificial disc replacement and Zero-P fusion for the treatment of single-segment cervical spondylosis. The clinical data of 56 patients with single-segment cervical spondylosis who met the selection criteria between January 2015 and December 2018 were retrospectively analyzed, and they were divided into study group (27 cases, using Prodisc-C Vivo artificial disc replacement) and control group (29 cases, using Zero-P fusion) according to different surgical methods. There was no significant difference between the two groups in terms of gender, age, type of cervical spondylosis, disease duration, involved segments and preoperative pain visual analogue scale (VAS) score, Japanese Orthopaedic Association (JOA) score, neck disability index (NDI), surgical segments range of motion, upper and lower adjacent segments range of motion, overall cervical spine range of motion, and cervical curvature (P>0.05). The operation time, intraoperative blood loss, postoperative hospitalization stay, time of returning to work, clinical effectiveness indicators (VAS score, JOA score, NDI, and improvement rate of each score), and imaging indicators (surgical segments range of motion, upper and lower adjacent segments range of motion, overall cervical spine range of motion, and cervical curvature, prosthesis position, bone absorption, heterotopic ossification, etc.) were recorded and compared between the two groups. There was no significant difference in operation time and intraoperative blood loss between the two groups (P>0.05); the postoperative hospitalization stay and time of returning to work in the study group were significantly shorter than those in the control group (P<0.05). Both groups were followed up 12-64 months, with an average of 26 months. There was no complication such as limb or organ damage, implant failure, and severe degeneration of adjacent segments requiring reoperation. The VAS score, JOA score, and NDI of the two groups at each time point after operation significantly improved when compared with those before operation (P<0.05); there was no significant difference in the above scores at each time point after operation between the two groups (P>0.05); there was no significant difference in the improvement rate of each score between the two groups at last follow-up (P>0.05). The surgical segments range of motion in the study group maintained to varying degrees after operation, while it in the control group basically disappeared after operation, showing significant differences between the two groups (P<0.05). At last follow-up, there was no significant difference in the upper and lower adjacent segments range of motion in the study group when compared with preoperative ones (P>0.05), while the upper adjacent segments range of motion in the control group increased significantly (P<0.05). The overall cervical spine range of motion and cervical curvature of the two groups decreased at 3 months after operation, and increased to varying degrees at last follow-up, but there was no significant difference between groups and within groups (P>0.05). At last follow-up, X-ray films and CT examinations showed that no prosthesis loosening, subsidence, or displacement was found in all patients; there were 2 cases (7.4%) of periprosthetic bone resorption and 3 cases (11.1%) of heterotopic ossification which did not affect the surgical segments range of motion. Both the Prodisc-C Vivo artificial disc replacement and Zero-P fusion have satisfactory short-term effectiveness in treatment of single-segment cervical spondylosis. Prodisc-C Vivo artificial disc replacement can also maintain the cervical spine range of motion to a certain extent, while reducing the occurrence of excessive motion of adjacent segments after fusion.

Survival and failure rate of lower lingual bonded retainers: Aretrospective cohort evaluation.

The objectives of the study were to evaluate the survival rates for lower lingual retainers (LLRs) and to establish a correlation between patients' treatment-related factors (age, sex, malocclusion, appliance used for treatment, teeth bonded, retention protocols) to the survival of LLRs. A total of 765 subjects [474 females and 291 males: mean age=24.29 ± 10.67 years] between 2013 and 2022 were included. A customized data collection form was utilized to gather the data from the electronic health record (EHR) of subjects. Patient-related factors, crowding or spacing, deep bite or open bite and duration of treatment were extracted from patients' files. Kaplan-Meier estimator was used for the survival function, whereas Cox proportional hazards regression models were used to associate risk factors with retainer survival. 328 (42.9%) subjects had their LLRs failed, and the survival period was on average 17.37 ± 22.85 months. On the other hand, the follow-up period for the retainers that did not fail was on average 47.19 ± 23.66 months. 192 (28.3%) subjects had segment failure (retainer detached from 3 teeth or less), while 51 (7.5%) subjects had failures in more than 3 teeth (complete). None of the evaluated clinical variables were significantly associated with LLRs failure except for the bite category (P=.013) and the appliance used for treatment (P < .001). Success rate for LLRs was 57.1% over 47.19 months, failure rate was 42.9% over 17.37 months. The presence of deep bite and treatment with aligners were significantly associated with increased failure rate.

Ensemble learning and test-time augmentation for the segmentation of mineralized cartilage versus bone in high-resolution microCT images

High-resolution non-destructive 3D microCT imaging allows the visualization and structural characterization of mineralized cartilage and bone. Deriving statistically relevant quantitative structural information about these tissues, however, requires automated segmentation procedures, mainly because manual contouring is user-biased and time-consuming. Despite the increased spatial resolution in microCT 3D volumes, automatic segmentation of mineralized cartilage versus bone remains non-trivial since they have similar grayscale values. Our work investigates how reliable 2D segmentation masks can be predicted automatically based on a (set of) convolutional neural network(s) trained with a limited number of manually annotated samples. To do that, we compared different strategies to select the 2D samples to annotate and considered ensemble learning and test-time augmentation (TTA) to mitigate the limited accuracy and robustness resulting from the small number of annotated training samples. We show that, for a fixed amount of annotated image samples, 2D microCT slices to annotate should preferably be selected in distinct 3D volumes, at regular intervals, rather than being grouped in adjacent slices of a same 3D volume. Two main lessons are drawn regarding the use of ensembles or TTA instead of a single model. First, ensemble learning is shown to improve segmentation accuracy and to reduce the mean and standard deviation of the absolute errors in cartilage characteristics obtained with different initializations of the neural network training process. In contrast, TTA appears to be unable to improve the model’s robustness to unlucky initializations. Second, both TTA and ensembling improved the model’s confidence in its predictions and segmentation failure detection.

A hybrid active contour model based on pre-fitting energy and adaptive functions for fast image segmentation

• An adaptive edge indicator function is proposed to make evolution process more stable and robust to initial contours. • An adaptive sign function is proposed to enable the evolution curve to adjust evolution direction and speed. • New pre-fitting functions are computed before the iteration without online update to reduce computation cost. • A single-well potential function is provided to prevent segmentation failure and increase segmentation speed. In this study, a hybrid active contour model driven by pre-fitting energy with an adaptive edge indicator function and an adaptive sign function is proposed. The key idea of employing the pre-fitting energy is to define two pre-fitting functions to calculate mean intensities of two sub-regions separated from the selected local region based on pre-calculated median intensity of the selected local region before the curve evolves, which saves a huge amount of computation cost. In addition, a brand-new single-well potential function and its associated evolution speed function are put forward to enable evolution process to converge faster as well as more robust. Experimental outcomes indicate that this model is competent to obtain motion boundaries of different targets effectively and efficiently. Compared with traditional models and recently developed models, this model not only reduces the CPU running time and iteration number significantly as well as improves segmentation accuracies (Dice similarity coefficient (DSC) and Jaccard similarity coefficient (JSC)), but also exhibits insensitivity to initialization.

Recent Seismic Sequences and Activation of Normal Fault Systems in the Mugello Basin and Surrounding Areas (Northern Apennines, Italy)

The Mugello Basin (North-Eastern Tuscany) is an intermontane basin of the Northern Apennines (Italy) with a well-documented record of seismicity; the two major historical earthquakes occurred in 1542 (Mw = 6.0) and in 1919 (Mw = 6.4). In this study, we integrate different seismic catalogs spanning the 2005–2019 time interval, and complement these data with phase arrival times from a temporary network that specifically operated in the area throughout the 2019–2021 time interval. The subsequent relocation of this data set with a double-difference algorithm allows for accurate analyses of the most relevant seismic sequences which affected the study area in 2008, 2009, 2015–2017, and 2019. These sequences are associated with the activation of adjacent segments of larger NW-striking fault systems, one of which bounds the NE margin of the Mugello Basin (Ronta Fault System). For each seismic sequence, best-fit fault surfaces are derived from orthonormal regression of relocated hypocenters, yielding consistent results with that derived from fault plane solutions. The four sequences mark a significant increase in the seismicity rate with respect to what was recorded in the previous decades. This suggests that, following the 2008 renewal of seismicity, static or dynamic stress changes, or both depending on the case, played a role in advancing the time of failure of the fault segments activated subsequently.

Towards reliable cardiac image segmentation: Assessing image-level and pixel-level segmentation quality via self-reflective references.

Cardiac image segmentation is a fundamental step in cardiovascular disease diagnosis, where many deep learning models have achieved promising performance. However, when deploying these well-trained models for real clinical usage, the network will inevitably produce inferior results due to domain shifts, motion artifacts, etc. How to avoid the potential poor-quality segmentations involved in clinical decision making is crucial for reliable computer-aided cardiac disease diagnosis. To this end, we develop a quality control method to identify failure segmentations by measuring their qualities, and report them to physicians for professional opinions. In specific, we propose a reference-based framework to assess the image-level quality (i.e.per-class Dice) for overall evaluation and pixel-level quality (i.e.pixel-wise correct map) to locate mis-segmented regions. Following previous works, we create informative references first, and investigate their relative relationships (e.g.differences) to the inputs to expose segmentation failures. However, we generate and leverage the references in different ways. We instantiate the references by recovering input images from segmentations by a self-reflective reference generator. If the segmentation is of good quality, the reference (i.e.the reconstructed image) will be close to the input image, and the inconsistency between them would be a good indicator to deduce the qualities. To effectively explore these inconsistency, we employ a difference investigator equipped with semantic class-aware compactness constraint to force the correctly-segmented features more separable to the wrongly-segmented ones. The experiments on ACDC and MSCMR datasets demonstrated our method could effectively capture segmentation failures, and the results on low-quality (Dice∈[0,0.6)), medium-quality (Dice∈[0.6,0.8)) and high-quality (Dice∈[0.8,1.0)) segmentations showed satisfying robustness of our method.

Role of visceral pleural invasion and tumor sizing at CT of resected NSCLC in clinical-radiological and pathological T agreement.

To describe in non-small cell lung cancer (NSCLC) the impact of visceral pleural invasion (VPI) and of tumor sizing assessed at computed tomography (CT) on the agreement between clinical-radiological and pathological T staging and its prognostic value. Patients affected by NSCLC treated by surgery in the period from January 2017 to September 2020 were retrospectively evaluated. Exclusion criteria were: (1) baseline CT not performed in our hospital; (2) failure of software segmentation at CT of the primary lesion. Clinical-radiological T (cT) was assessed at baseline CT, evaluating in particular T size by semi-automatic tool and VPI (cVPI) visually. Pathological T (pT) and VPI (pVPI) were recorded by pathological report and obtained after formalin-fixation and eventual elastic stain on surgical specimen. The agreement between cT and pT was evaluated by calculating the weighted kappa by Cohen (κw); the association between progression free survival (PFS) with both cT and pT was assessed by the Cox regression analysis. The study included 84 NSCLC in 82 patients (median age 71 years, IQR 63-76 years; females 22/82, 27%). The agreement between cT and pT was poor (κw 0.302, 95%CI 0.158-0.447). The main causes of disagreement were CT oversizing (21%) and false positive cVPI (29%). A significant association was found between PFS and pT2-T3 (HR 2.75, 95%CI 1.21-6.25, p=0.015) but not with cT2-T3 (not retained in the model). False positive cVPI and oversizing at CT are causes of disagreement between cT and pT in around one-third of resected NSCLC. PFS was significantly associated with pT but not with cT.

OCT-Angiography Face Mask-Associated Artifacts During the COVID-19 Pandemic.

Précis:Face mask wearing has no significant effects on artifacts or vessel density measurements in optic nerve head (ONH) and macular optical coherence tomography-angiography (OCT-A) scans.Purpose:The aim was to assess the difference in area of artifacts observed in optical OCT-A scans with and without face mask wear and to verify if mask wear interferes with OCT-A vessel density measurements.Subjects and Controls:A total of 64 eyes of 10 healthy subjects, 4 ocular hypertensive, 8 glaucoma suspects, and 17 glaucoma patients were included.Materials and Methods:High-density ONH and macula OCT-A scans were obtained in patients with and without surgical masks. Seven different artifacts (motion, decentration, defocus, shadow, segmentation failure, blink, and Z-offset) were quantitatively evaluated by 2 trained graders. The changes in the area (% of scan area) of artifacts, without and with mask wearing, and differences of vessel density were evaluated.Results:Trends of increasing motion artifact area for the ONH scans [4.23 (−0.52, 8.98) %, P=0.08] and defocus artifact area for the macular scans [1.06 (−0.14, 2.26) %, P=0.08] were found with face mask wear. However, there were no significant differences in the mean % area of any artifacts (P>0.05 for all). Further, the estimated mean difference in vessel density in images acquired without and with masks was not significant for any type of artifact.Conclusion:Face mask wearing had no significant effect on area of artifacts or vessel density measurements. OCT-A vessel density measurements can be acquired reliably with face mask wear during the pandemic.

Risk factors of postoperative pulmonary complications after primary posterior fusion and hemivertebra resection in congenital scoliosis patients younger than 10 years old: a retrospective study

BackgroundPostoperative pulmonary complications are common and associated with morbidity and mortality. Congenital scoliosis is a failure of vertebral formation and/or segmentation arising from abnormal vertebral development. Posterior fusion and osteotomy are necessary for these patients to prevent deterioration of spine deformity. The incidence of postoperative pulmonary complications in this specific group of patients, especially young children were unknown.MethodsA retrospective study was conducted and electronic medical records of early-onset scoliosis patients who had primary posterior fusion and hemivertebra resection at our institution from January 2014 to September 2019 were reviewed. The demographic characteristics, the intraoperative and postoperative parameters were collected to identify the predictors of postoperative pulmonary complications.ResultsA total of 174 patients (57.5% boys) with a median age of 3 years old were included for analysis. Eighteen patients (10.3%) developed perioperative pulmonary complications and pneumonia (n=13) was the most common. History of recent upper respiratory infection was not related to postoperative pulmonary complications. Multifactorial regression analysis showed thoracoplasty was the only predictive risk factor of postoperative pulmonary complications.ConclusionsFor congenital scoliosis patients younger than 10 years old, thoracoplasty determine the occurrence of postoperative pulmonary complications. Both surgeons and anesthesiologists should pay attention to patients undergoing thoracoplasty and preventive measures are necessary.

PeMNet for Pectoral Muscle Segmentation.

Simple SummaryDeep learning has become a popular technique in modern computer-aided (CAD) systems. In breast cancer CAD systems, breast pectoral segmentation is an important procedure to remove unwanted pectoral muscle in the images. In recent decades, there are numerous studies aiming at developing efficient and accurate methods for pectoral muscle segmentation. However, some methods heavily rely on manually crafted features that can easily lead to segmentation failure. Moreover, deep learning-based methods are still suffering from poor performance at high computational costs. Therefore, we propose a novel deep learning segmentation framework to provide fast and accurate pectoral muscle segmentation result. In the proposed framework, the novel network architecture enables more useful information to be used and therefore improve the segmentation results. The experimental results using two public datasets validated the effectiveness of the proposed network.As an important imaging modality, mammography is considered to be the global gold standard for early detection of breast cancer. Computer-Aided (CAD) systems have played a crucial role in facilitating quicker diagnostic procedures, which otherwise could take weeks if only radiologists were involved. In some of these CAD systems, breast pectoral segmentation is required for breast region partition from breast pectoral muscle for specific analysis tasks. Therefore, accurate and efficient breast pectoral muscle segmentation frameworks are in high demand. Here, we proposed a novel deep learning framework, which we code-named PeMNet, for breast pectoral muscle segmentation in mammography images. In the proposed PeMNet, we integrated a novel attention module called the Global Channel Attention Module (GCAM), which can effectively improve the segmentation performance of Deeplabv3+ using minimal parameter overheads. In GCAM, channel attention maps (CAMs) are first extracted by concatenating feature maps after paralleled global average pooling and global maximum pooling operation. CAMs are then refined and scaled up by multi-layer perceptron (MLP) for elementwise multiplication with CAMs in next feature level. By iteratively repeating this procedure, the global CAMs (GCAMs) are then formed and multiplied elementwise with final feature maps to lead to final segmentation. By doing so, CAMs in early stages of a deep convolution network can be effectively passed on to later stages of the network and therefore leads to better information usage. The experiments on a merged dataset derived from two datasets, INbreast and OPTIMAM, showed that PeMNet greatly outperformed state-of-the-art methods by achieving an IoU of , global pixel accuracy of , Dice similarity coefficient of , and Jaccard of , respectively.

Long Term Patient Reported Outcomes of Tarsal Coalition Resection in Pediatric Patients

Category:Hindfoot; Midfoot/ForefootIntroduction/Purpose:Tarsal Coalition has been known to be associated with pes planovalgus, unrelenting pain, recurrent sprains, and arthrosis due to an abnormal connection between two or more bones of the feet following failure of segmentation during development. Three entities of coalition exist fibrous, cartilaginous and osseous. Most commonly patients present with pain during activity. On exam range of motion is usually limited and there may be valgus or varus malalignment of the hindfoot. Patients are often offered surgical resection following failure of conservative management. Few large cohort studies exist analyzing complications and patient reported outcomes of pediatric patients undergoing operative treatment.Methods:After IRB approval, patients under the age of 18 undergoing excision of tarsal coalition at a single academic center from 2010-2019 were identified. Medical records were reviewed for variables such as demographics, surgical technique, coalition characteristics, postoperative complications, and coalition recurrence. Patients were contacted via telephone to complete a PROMIS Physical Function and Pain Interference in addition to the Foot Function Index. Contact was attempted at least 3 times before a patient was considered non-respondent. A total of 44 patients and 54 feet were included in the analysis with 17 patients (22 feet) completing patients reported outcomes.Results:The median age at the time of surgery was 12 (IQR 3). Thirty-five feet (64.8%) were from male patients, while 19 feet (35.2%) were from female patients. Pain was the presenting symptom in 49 (90.7%) of feet. Pes planovalgus was present in 25 feet (46.3%). Forty feet (74.1%) had calcaneonavicular coalition compared to 12 feet (22.2%) with talocalcaneal coalition and 2 feet (3.7%) with both. Fibrous coalition was found in 19 feet (35.2%) and the most common. Interposition material was used in 49/54 (90.7%) feet. Wound complications or temporary neuropraxias were seen in 7/54 feet (13.0%), and coalition recurrence occurred in 3/54 feet (5.6%). At a median of 6 years (6.0 IQR) postoperatively, the median postoperative outcomes scores were as follows: PROMIS Physical Function 54.7 (17.9 (IQR)), PROMIS Pain Interference 38.7 (10.0), FFI Pain 14.0 (30.0), FFI Disability 1.0 (9.0), FFI Activity Limitation 3.0 (8.0), and FFI Total 6.0 (12.0).Conclusion:Our study investigated complications and long-term patient outcomes following coalition resection. At follow-up, pediatric patients who underwent excision of tarsal coalition had higher physical function and less pain than the average U.S. population (PROMIS population mean is score of 50). Our recurrence rate of 5.6% is lower than reported in literature. Infection rate was similar to reports in literature at 3.1%. Overall, Coalition resection in pediatrics is a well-tolerated procedure that improves both physical and emotional health.

Assessing the role of cross‐linked sodium hyaluronate in improving the quality of circumpapillary OCT scans

AbstractPurposeAssessment of cross‐linked sodium hyaluronate (XLHA) in improving circumpapillary OCT scan quality and segmentation accuracyMethodsCircumpapillary OCT scans (Heidelberg Spectralis) of prospectively recruited glaucoma clinic patients were assessed. Patients were included if suboptimal image quality was suspected, based on live preview images. Exclusion criteria included severe loss of media clarity due to corneal opacification or dense cataract, adverse reaction to any constituents of the investigative eyedrop, contact lens use, eyedrop use in the preceding two hours, significant imaging artefact, or scanning and segmentation failure. Prior to repeat scanning, within‐patient eyes were randomised to receive topical application of XLHA (VisuXL) or blinking alone as a control group. The objective primary outcome measure was the mean proprietary image quality indicator (QI). Subjectively, masked clinician assessment of image preference was assessed. Global circumpapillary RNFL thickness and the improvement of segmentation errors were secondary outcomes.ResultsSixty‐seven patients (134 eyes) were eligible for analysis with a mean age (±SD) of 69.7 ± 14.4 years, of whom 57% were male and 43% female. Mean baseline QI in the XLHA group was 23.0 ± 4.95, and 23.3 ± 5.04 in the control eyes (p = 0.76). The mean change in QI was +0.28 ± 3.93 in the XLHA eyes, and +0.34 ± 4.01 in the control eyes (p = 0.93). No significant effect on RNFL thickness (p = 0.82), subjective image preference (p = 0.73) or the extent of segmentation errors (p = 0.46) was seen. Post‐hoc analysis revealed baseline QI as a predictor of repeat scan improvement across all groups, with a baseline QI less than 24.5 associated with improvement on repeat scanning (OR 4.49, 95% CI 2.06 – 9.59).ConclusionsRepeat scanning was beneficial with baseline QI below 24.5; however, no evidence was found to support the use of XLHA over blinking to improve repeat OCT scan quality.

Single-phase Scan Using Trapezoidal Cross-injection Method in 3D-CT before Pulmonary Segmentectomy

We performed three-dimensional computed tomography (3D-CT) with pulmonary artery (PA) and pulmonary vein (PV) segmentation for image support in patients undergoing pulmonary segmentectomy. Automatic segmentation of PA and PV is possible with single-phase scan by the use of SYNAPSE VINCENT (Fujifilm) image integration module. However, there are limitations to this automatic segmentation with many cases of segmentation errors and failures. Therefore, we examined the CT value and automatic extraction of PA and PV using a contrast-enhanced protocol in which the chest is included in the imaging range. As a result of a retrospective study, a new protocol using the trapezoidal cross-injection method in which the CT value of PA and PV is 200 Hounsfield unit (HU) or more and the CT value difference of PA and PV is 50 HU or more is desirable. The amount of contrast used was decreased to about 60% with the new protocol. CT value and difference in density were enough for imaging the PA and PV. The automatic image integration with SYNAPSE VINCENT had tendency for improved success rate with the new protocol, suggesting the usefulness of the new protocol.

Penerapan Convolutional Neural Networks untuk Mesin Penerjemah Bahasa Daerah Minangkabau Berbasis Gambar

One of efforts by the Indonesian people to defend the country is to preserve and to maintain the regional languages. The current era of modernity makes the regional language image become old-fashioned, so that most them are no longer spoken. If it is ignored, then there will be a cultural identity crisis that causes regional languages to be vulnerable to extinction. Technological developments can be used as a way to preserve regional languages. Digital image-based artificial intelligence technology using machine learning methods such as machine translation can be used to answer the problems. This research will use Deep Learning method, namely Convolutional Neural Networks (CNN). Data of this research were 1300 alphabetic images, 5000 text images and 200 vocabularies of Minangkabau regional language. Alphabetic image data is used for the formation of the CNN classification model. This model is used for text image recognition, the results of which will be translated into regional languages. The accuracy of the CNN model is 98.97%, while the accuracy for text image recognition (OCR) is 50.72%. This low accuracy is due to the failure of segmentation on the letters i and j. However, the translation accuracy increases after the implementation of the Leveinstan Distance algorithm which can correct text classification errors, with an accuracy value of 75.78%. Therefore, this research has succeeded in implementing the Convolutional Neural Networks (CNN) method in identifying text in text images and the Leveinstan Distance method in translating Indonesian text into regional language texts.&#x0D;

Study on Mechanical Characteristics of Segmental Joints of a Large-Diameter Shield Tunnel under Ultrahigh Water Pressure.

At present, there is no clear design standard for segmental joints of large-diameter shield tunnels under high water pressure. In this paper, a theoretical calculation model for the bending stiffness of segmental joints under high water pressure is proposed. The numerical simulation method is used to investigate the failure and crack formation processes of single-layer and double-layer lining segments under large axial forces. The effects of axial force, bolt strength, and concrete strength on the bending stiffness of joints are then studied using a theoretical calculation model of segmental joints. The results show that under extremely high water pressure, the influence of double lining on joint stiffness is limited. It is more rational and safe to compute the bending stiffness of segmental joints using this theoretical model rather than the numerical simulation method. The parameter analysis reveals that increasing the bolt strength has a minor impact on bending stiffness and deformation, whereas increasing the concrete strength has the opposite effect. The influence of ultimate bearing capacity and deformation decreases non-linearly as the axial force increases.