Incomplete Image Data Research Articles

Postoperative Effusion-Synovitis Is Associated With Poor Clinical Outcomes in Patients With Femoroacetabular Impingement Syndrome

Purpose(1) to investigate the consecutive changes in effusion-synovitis following primary arthroscopic treatment for patients with femoroacetabular impingement syndrome (FAIS), and (2) to determine the effect of postoperative effusion-synovitis on clinical outcomes. MethodsData between March 2021 and January 2022 was reviewed. Patients diagnosed with FAIS and undergoing hip arthroscopic treatment were included. Exclusion criteria were incomplete magnetic resonance imaging (MRI) data, prior history of hip surgery, labral reconstruction, and concomitant hip conditions. MRI (non-contrast 3.0 T) was performed preoperatively and 3, 6, 12-month postoperatively, and the measurement of the largest femoral neck fluid thickness (FTM) and cross-sectional area (CSA) of the effusion-synovitis were collected. Preoperative and a minimum of 2-year postoperative patient-reported outcome (PRO) scores including Visual Analog pain Scale (VAS), modified Harris Hip Score (mHHS), and international Hip Outcome Tool, 12-component form (iHOT-12) were collected and compared. Postoperative Tegner activity scale was also collected. The PROs and achievements of minimal clinically important difference (MCID) and patient acceptable symptom state (PASS) were compared between patients with and without postoperative effusion-synovitis. Multivariate linear regression analysis was performed to determine the effect of the effusion-synovitis size on PROs. ResultsA total of 61 patients (61 hips) were included in the study. The 3-month postoperative FTM, CSA and grade of effusion-synovitis presented a significant increase compared to the preoperative values (all with P < .05). No significant differences were observed in the 6-month postoperative measurements compared to the preoperative values (all with P > .05). At the 12-month follow-up, although there was a significant decrease in all measurements compared to the preoperative values (all with P < .001), 39 patients (63.9%) still presented effusion-synovitis. Compared to the other 22 patients (36.1%) without effusion-synovitis, these patients presented inferior mHHS, iHOT-12 (all with P < .05), as well as lower achievement of PASS of mHHS (82.1% vs 100%, P = .035) and iHOT-12 (38.5% vs 81.8%, P = .001). The achievement of MCID of mHHS (79.5% vs 77.3%, P = .839) and iHOT-12 (89.7% vs 95.5%, P = .839) were comparable between patients with and without effusion-synovitis. The postoperative sagittal CSA (Beta = -.302, P = .039) were negatively related to mHHS in the regression analysis. ConclusionAfter arthroscopic treatment for FAIS, the level of effusion-synovitis presented an initial increase, then followed by a subsequent decrease. Effusion-synovitis was significantly alleviated at 12 months compared to the preoperative level. Patients with postoperative effusion-synovitis had inferior clinical outcomes and lower achievement of PASS compared to those without. Level of EvidenceLevel IV; retrospective case series.

Three-Dimensional Reconstruction of Railway Bridges Based on Unmanned Aerial Vehicle–Terrestrial Laser Scanner Point Cloud Fusion

To address the incomplete image data collection of close-to-ground structures, such as bridge piers and local features like the suspension cables in bridges, obtained from single unmanned aerial vehicle (UAV) oblique photography and the difficulty in acquiring point cloud data for the top structures of bridges using single terrestrial laser scanners (TLSs), as well as the lack of textural information in TLS point clouds, this study aims to establish a high-precision, complete, and realistic bridge model by integrating UAV image data and TLS point cloud data. Using a particular large-scale dual-track bridge as a case study, the methodology involves aerial surveys using a DJI Phantom 4 RTK for comprehensive image capture. We obtain 564 images circling the bridge arches, 508 images for orthorectification, and 491 images of close-range side views. Subsequently, all images, POS data, and ground control point information are imported into Context Capture 2023 software for aerial triangulation and multi-view image dense matching to generate dense point clouds of the bridge. Additionally, ground LiDAR scanning, involving the placement of six scanning stations both on and beneath the bridge, was conducted and the point cloud data from each station are registered in Trimble Business Center 5.5.2 software based on identical feature points. Noise point clouds are then removed using statistical filtering techniques. The integration of UAV image point clouds with TLS point clouds is achieved using the iterative closest point (ICP) algorithm, followed by the creation of a TIN model and texture mapping using Context Capture 2023 software. The effectiveness of the integrated modeling is verified by comparing the geometric accuracy and completeness of the images with those obtained from a single UAV image-based model. The integrated model is used to generate cross-sectional profiles of the dual-track bridge, with detailed annotations of boundary dimensions. Structural inspections reveal honeycomb surfaces and seepage in the bridge piers, as well as painted rust and cracks in the arch ribs. The geometric accuracy of the integrated model in the X, Y, and Z directions is 1.2 cm, 0.8 cm, and 0.9 cm, respectively, while the overall 3D model accuracy is 1.70 cm. This method provides technical reference for the reconstruction of three-dimensional point cloud bridge models. Through 3D reconstruction, railway operators can better monitor and assess the condition of bridge structures, promptly identifying potential defects and damages, thus enabling the adoption of necessary maintenance and repair measures to ensure the structural safety of the bridges.

3D wind field profiles from hyperspectral sounders: revisiting optic-flow from a meteorological perspective

In this work, we present an efficient optic flow algorithm for the extraction of vertically resolved 3D atmospheric motion vector (AMV) fields from incomplete hyperspectral image data measures by infrared sounders. The model at the heart of the energy to be minimized is consistent with atmospheric dynamics, incorporating ingredients of thermodynamics, hydrostatic equilibrium and statistical turbulence. Modern optimization techniques are deployed to design a low-complexity solver for the energy minimization problem, which is non-convex, non-differentiable, high-dimensional and subject to physical constraints. In particular, taking advantage of the alternate direction of multipliers methods (ADMM), we show how to split the original high-dimensional problem into a recursion involving a set of standard and tractable optic-flow sub-problems. By comparing with the ground truth provided by the operational numerical simulation of the European Centre for Medium-Range Weather Forecasts (ECMWF), we show that the performance of the proposed method is superior to state-of-the-art optical flow algorithms in the context of real infrared atmospheric sounding interferometer (IASI) observations.

Brain tumour segmentation with incomplete imaging data.

Progress in neuro-oncology is increasingly recognized to be obstructed by the marked heterogeneity-genetic, pathological, and clinical-of brain tumours. If the treatment susceptibilities and outcomes of individual patients differ widely, determined by the interactions of many multimodal characteristics, then large-scale, fully-inclusive, richly phenotyped data-including imaging-will be needed to predict them at the individual level. Such data can realistically be acquired only in the routine clinical stream, where its quality is inevitably degraded by the constraints of real-world clinical care. Although contemporary machine learning could theoretically provide a solution to this task, especially in the domain of imaging, its ability to cope with realistic, incomplete, low-quality data is yet to be determined. In the largest and most comprehensive study of its kind, applying state-of-the-art brain tumour segmentation models to large scale, multi-site MRI data of 1251 individuals, here we quantify the comparative fidelity of automated segmentation models drawn from MR data replicating the various levels of completeness observed in real life. We demonstrate that models trained on incomplete data can segment lesions very well, often equivalently to those trained on the full completement of images, exhibiting Dice coefficients of 0.907 (single sequence) to 0.945 (complete set) for whole tumours and 0.701 (single sequence) to 0.891 (complete set) for component tissue types. This finding opens the door both to the application of segmentation models to large-scale historical data, for the purpose of building treatment and outcome predictive models, and their application to real-world clinical care. We further ascertain that segmentation models can accurately detect enhancing tumour in the absence of contrast-enhancing imaging, quantifying the burden of enhancing tumour with an R 2 > 0.97, varying negligibly with lesion morphology. Such models can quantify enhancing tumour without the administration of intravenous contrast, inviting a revision of the notion of tumour enhancement if the same information can be extracted without contrast-enhanced imaging. Our analysis includes validation on a heterogeneous, real-world 50 patient sample of brain tumour imaging acquired over the last 15 years at our tertiary centre, demonstrating maintained accuracy even on non-isotropic MRI acquisitions, or even on complex post-operative imaging with tumour recurrence. This work substantially extends the translational opportunity for quantitative analysis to clinical situations where the full complement of sequences is not available and potentially enables the characterization of contrast-enhanced regions where contrast administration is infeasible or undesirable.

Altered cerebral blood flow and white matter during wakeful rest in patients with obstructive sleep apnea: a population-based retrospective study.

To explore changes in cerebral blood flow (CBF) and white matter during wakeful rest in patients with obstructive sleep apnea (OSA). The subjects comprised OSA patients and age- and sex-matched non-sleep apnea (NSA) subjects from December 2020 to December 2021. All subjects underwent structural and arterial spin labeling MRI examinations using a 3.0 T MRI scanner. Intergroup differences in regional and global CBF and white matter hyperintensities (WMHs) were analyzed. In this study, 100 (74 males) of 750 (439 males) subjects were diagnosed with OSA, so the prevalence of OSA in the general population was 13.3% (100/750), with 16.9% (74/439) in males and 8.4% (26/311) in females. Excluding four patients with incomplete imaging data, 96 OSA patients and 103 age- and sex-matched NSA subjects were included. At global level, OSA patients showed significantly decreased CBF values in gray matter and whole brain compared to NSA subjects (gray matter: p = 0.010; whole brain: p = 0.021). No significant difference in CBF values was found in WM between the two groups (p = 0.250). At regional level, compared with NSA subjects, patients with OSA exhibited significantly decreased regional CBF values mainly in right parietal lobe and right temporal lobe. Moreover, OSA patients had significantly higher WMHs burden than NSA subjects (p = 0.017). OSA patients exhibit decreased global and regional CBF values and increased WMHs burden. These findings provide a basis for exploring neuropathological changes of OSA and for early and appropriate treatment.

Deep Learning for Tumour Segmentation with Missing Data

Abstract AIMS Brain tumour segmentation remains a challenging task, complicated by the marked heterogeneity of imaging appearances and their distribution across multiple modalities: FLAIR, T1-weighted, T2-weighted, and contrast-enhanced T1-weighted sequences (T1CE). But the use of all four imaging sequences is not always possible. The causes for this are legion, with common examples including corruption by image artefacts and acquisition constraints, such as those imposed in pre-operative stealth studies. We therefore aimed to quantify how well tumour segmentation models perform with incomplete imaging data. METHOD We developed a collection of 30 state-of-the-art deep learning tumour segmentation models, nnU-Net-derived, and deployed them across all possible combinations of imaging modalities, trained, and tested with five-fold cross-validation on the 2021 BraTS-RSNA glioma population of 1251 patients, with additional out-of-sample comparison to neuroradiologist hand-labelled lesions from our own centre. RESULTS Regardless of imaging available, models largely performed well. The best models with varying degrees of missingness were as follows: single sequence available - FLAIR (Dice 0.938); two sequences available - FLAIR + T1CE (Dice 0.943) and three sequences available – FLAIR + T1CE + T2 (Dice 0.945). In comparison, a model with complete data (FLAIR + T1 + T1CE + T2) achieved a similar Dice coefficient of 0.945. CONCLUSION Tumour segmentation models with missing sequences – common in clinical practice - still delineate lesions well, often with comparable performance to when all data is available. This provides opportunity for quantitative imaging in patients and clinical situations wherein full MRI acquisitions are not possible.

Poster 235: Surgical Management of Patellar Instability in Adolescents with High Grade Trochlear Dysplasia

Objectives:High-grade trochlear dysplasia (HGTD) has previously been identified as a risk factor for failure of surgical management of patellofemoral instability (PFI). However, limited data exists on the influence of the type of surgical procedure performed on postoperative outcomes in these patients. The purpose of this study was to identify the rates of failure of primary PFI surgery in patients with HGTD compared to those without HGTD and to assess the effect of the type of surgery on recurrent instability and return to OR.Methods:We retrospectively reviewed patients aged ≤19 years who underwent primary surgery for PFI at our institution between 2008 and 2017. Patients underwent procedures including medial retinacular plication (MRP), medial patellofemoral ligament reconstruction (MPFLR), tibial tubercle osteotomy (TTO), or combined procedures. We excluded patients with incomplete imaging data, BMI >30kg/m2, those that underwent surgery for fixation of a chondral fracture >1cm, and patients with a diagnosis of syndromic disorder. Trochlear dysplasia was evaluated based on pre-operative MRI. Patients were divided between those that had HGTD and those without HGTD. Comparisons were made between the two groups and logistic regression analysis was utilized to determine if surgery category was associated with rate of recurrence in the HGTD patients.Results:Three-hundred three patients, average age 15.1 years, underwent primary surgical stabilization for PFI with median follow up time of 3.0 years. Forty-nine percent (149/303) of the cohort had HGTD. Demographics were comparable between HGTD groups (Table 1). When compared to the patients without HGTD, patients with HGTD showed similar rates of recurrent instability, return to the OR for instability, and return to the OR for other reasons. Looking at patients with just HGTD, there was a significant association between surgery type and outcomes (Table 2). HGTD patients who underwent MPFLR had a 63% decrease in the odds of experiencing recurrent instability compared to those undergoing MRP (OR=0.37; p=0.03). Similarly, the TTO surgery group had an 80% decrease in the odds of experiencing recurrent instability compared to patients undergoing MRP (OR=0.20; p=0.002).Conclusions:While adolescents with PFI and HGTD overall have similar results after surgical stabilization, there is a significant effect of the type of surgery performed. Patients with HGTD that underwent MRP had significantly more instability and return to the OR for instability surgeries than patients that underwent other procedures. Based on this data, adolescents with PFI and HGTD should be considered for more powerful surgeries to limit the risk of recurrence or return to the OR.Table 1.Patient Demographics and Outcomes Compared Between Trochlear Dysplasia Groups (n=303)Table 2.Outcomes of Patients with HGTD Compared by Type of Surgery (N=149).

Poster 234: Radiographic Predictors of Failure of Patellar Instability Surgery in Adolescents

Objectives:Adolescents often benefit from surgical management of recurrent patellofemoral instability (PFI). However, the debate continues as to which preoperative radiographic factors can be used to predict the risk of failure of patellar stabilization surgery. The purpose of this study was to determine the rate of failure, defined as recurrence of instability, in a large cohort of adolescent athletes undergoing primary surgery for PFI, and to identify preoperative radiographic risk factors for recurrence.Methods:We retrospectively reviewed patients aged 19 years and younger who underwent primary surgery for PFI at our institution between 2008 and 2017. Patients underwent surgical procedures including medial retinacular plication (MRP), medial patellofemoral ligament reconstruction (MPFLR), tibial tubercle osteotomy (TTO), or a combination thereof. Preoperative magnetic resonance imaging (MRI) studies were reviewed and radiographic parameters were recorded. We excluded patients with incomplete imaging data, BMI >30kg/m2, those that underwent surgery for fixation of a chondral fracture >1cm, and patients with a diagnosis of syndromic disorder. Demographic and radiographic variables were compared between patients who experienced recurrence of instability and those that did not. Logistic regression was utilized to determine if any imaging characteristics were associated with recurrent instability while controlling for age and sex.Results:Three-hundred three patients, average age 15.1 years, underwent primary surgical stabilization for PFI with median follow up time of 3.0 years. At final follow-up, 25% of patients (76/303) experienced recurrent instability with 41 patients (14%) requiring additional surgery because of this instability. Patients who had recurrent instability were younger (14.3vs15.4 years; p<0.001) and were more likely to have undergone MRP surgery alone (66% vs 35%; p<0.001) (Table 1). They were also more likely to have an open physeal status (62% vs 49%; p<0.001), a larger sulcus angle (163.9 vs 159.2; p<0.001), a larger patellar tilt angle (26.3 vs 23.6; p=0.04), and larger Caton-Deschamps ratio (1.3 vs 1.2; p=0.03) (Table 2). Each additional degree of sulcus angle increased the odds of recurrent instability by 5% (OR=1.05; p=0.002). Notably, each additional year of age at the time of surgery, decreased the odds of failure by 23% (OR=0.77; p<0.001).Conclusions:In conclusion, 25% of adolescents experienced recurrent instability after primary patellar stabilization surgery. These patients who failed initial surgical treatment were younger and had more MRP procedures. Radiographically, they had more open physes, flatter trochleae, more patellar tilt, and larger patellar height. This information can be used to guide treatment decisions and inform the risk of failure of surgical stabilization procedures in adolescents with PFI.Table 1.Patient Demographics by Presence or Absence of Recurrent Instabillity (N=303).Table 2.Imaging Measurements by Presence or Absence of Recurrent Instability (N=303).

Structure-constrained combination-based nonlinear association analysis between incomplete multimodal imaging and genetic data for biomarker detection of neurodegenerative diseases.

Multimodal imaging data are widely applied in imaging genetic studies to identify associations between imaging and genetic data for the biomarker detection of neurodegenerative diseases (NDs). However, the incomplete multimodal imaging data and complex relationships among imaging and genetic data make it difficult to effectively analyze associations between imaging and genetic data and accurately detect disease-related biomarkers. This study proposed a novel structure-constrained combination-based nonlinear association analysis method to exploit associations between incomplete multimodal imaging and genetic data for potential biomarker detection of NDs. Two types of structure constraints were used in imaging and genetic data. First, a parallel concatenated projection method with multiple constraints was adopted to handle missing data. Modality-shared and modality-specific information could be well captured to obtain latent imaging representations. A locality preserving constraint was applied to the imaging data for retaining structure information before and after projection. A connectivity penalty was also included to capture structure associations among latent imaging representations. Second, a group-induced graph self-expression constraint was incorporated into our method to exploit strong structure correlations among inter- and intra-group of genetic data. Finally, a nonlinear kernel-based method was used to explore the complex associations between latent imaging representations and genetic data for biomarker detection. A set of simulation data and two sets of real ND data, which were obtained from Alzheimer's disease neuroimaging initiative and Parkinson's progression markers initiative databases, were applied to assess the effectiveness of our method. High accuracy of biomarker detection was achieved. Moreover, the identification of disease-related biomarkers was confirmed in previous studies. Therefore, our method may provide a novel way to gain insights into the pathological mechanism of NDs and early prediction of these diseases.

External Validation and Comparison of Two Nomograms Predicting the Probability of Lymph Node Involvement in Patients subjected to Robot-Assisted Radical Prostatectomy and Concomitant Lymph Node Dissection: A Single Tertiary Center Experience in the MRI-Era.

IntroductionTo externally validate and directly compare the performance of the Briganti 2012 and Briganti 2019 nomograms as predictors of lymph node invasion (LNI) in a cohort of patients treated with robot-assisted radical prostatectomy (RARP) and extended pelvic lymph node dissection (ePLND).Materials and MethodsAfter the exclusion of patients with incomplete biopsy, imaging, or clinical data, 752 patients who underwent RARP and ePLND between December 2014 to August 2021 at our center, were included. Among these patients, 327 (43.5%) had undergone multi-parametric MRI (mpMRI) and mpMRI-targeted biopsy. The preoperative risk of LNI was calculated for all patients using the Briganti 2012 nomogram, while the Briganti 2019 nomogram was used only in patients who had performed mpMRI with the combination of targeted and systematic biopsy. The performances of Briganti 2012 and 2019 models were evaluated using the area under the receiver-operating characteristics curve analysis, calibrations plot, and decision curve analysis.ResultsA median of 13 (IQR 9–18) nodes per patient was removed, and 78 (10.4%) patients had LNI at final pathology. The area under the curves (AUCs) for Briganti 2012 and 2019 were 0.84 and 0.82, respectively. The calibration plots showed a good correlation between the predicted probabilities and the observed proportion of LNI for both models, with a slight tendency to underestimation. The decision curve analysis (DCA) of the two models was similar, with a slightly higher net benefit for Briganti 2012 nomogram. In patients receiving both systematic- and targeted-biopsy, the Briganti 2012 accuracy was 0.85, and no significant difference was found between the AUCs of 2012 and 2019 nomograms (p = 0.296). In the sub-cohort of 518 (68.9%) intermediate-risk PCa patients, the Briganti 2012 nomogram outperforms the 2019 model in terms of accuracy (0.82 vs. 0.77), calibration curve, and net benefit at DCA.ConclusionThe direct comparison of the two nomograms showed that the most updated nomogram, which included MRI and MRI-targeted biopsy data, was not significantly more accurate than the 2012 model in the prediction of LNI, suggesting a negligible role of mpMRI in the current population.

Preoperative T and N Restaging of Rectal Cancer After Neoadjuvant Chemoradiotherapy: An Accuracy Comparison Between MSCT and MRI.

BackgroundIt is well established that magnetic resonance imaging (MRI) is better than multi-slice computed tomography (MSCT) for the accurate diagnosis of pretreatment tumor (T) and node (N) staging of rectal cancer. However, the diagnostic value of MRI and MSCT in local restaging of rectal cancer after neoadjuvant chemoradiotherapy (NCRT) is controversial. The aim of this study is to investigate the performance of the two imaging exams in restaging of patients with rectal cancer.MethodsPatients with rectal cancer from April 2015 to April 2021 were analyzed retrospectively. The inclusion criteria are as follows: 1) diagnosis of rectal cancer through pathology; 2) NCRT had been performed; 3) all patients had undergone both MSCT and MRI examination before the surgery. Exclusion criteria are as follows: 1) incomplete clinical and imaging data; 2) previous history of pelvic surgery. Two radiologists performed T and N staging of patient images. Diagnostic accuracy, consistency analysis, and error restaging distribution of the two imaging exams for T and N restaging of rectal cancer were assessed using postoperative pathological staging as the gold standard.ResultsA total of 62 patients (49 men; mean age: 59 years; age range 29–83 years) were included in the study. The diagnostic accuracy of MSCT and MRI for T restaging was 51.6% (95% CI 39.3%–63.9%) and 41.9% (95% CI 29.6%–54.2%), respectively, and no statistical difference was found between them (p > 0.05). The diagnostic accuracy of MSCT and MRI for N restaging was 56.5% (95% CI 44.2%–68.8%) and 53.2% (95% CI 40.8%–65.6%), respectively, and no statistical difference was found between them (p > 0.05). The consistency analysis showed that T restaging (κ = 0.583, p < 0.001) and N restaging (κ = 0.644, p < 0.001) were similar between MSCT and MRI. There was no significant difference in the distribution of over, accurate, or low staging in T restaging (p > 0.05) and N restaging (p > 0.05) between MSCT and MRI.ConclusionsMSCT and MRI have similarly poor performance in the diagnosis of preoperative T and N restaging of rectal cancer after NCRT. Neither of them cannot effectively stage the ypT0-1 of rectal cancer. These findings may be of clinical relevance for planning less imaging exam.

Transmission line surface defect detection method based on uav autonomous inspection

The existing transmission line surface defect detection methods have the problem of incomplete image data set, resulting in a low recognition success rate. A transmission line surface defect detection method based on uav autonomous inspection is designed. The safety of power grid operation is evaluated, the local linearization process is transformed into linear equation expression, the image data set is obtained by uav autonomous inspection, the transmission line state is judged, the corresponding constraint conditions are set, the type of transmission line surface defects are identified, the number of image poles and towers is matched, and the detection mode is optimized by edge detection algorithm. Experimental results: The average recognition success rate of the transmission line surface defect detection method in this paper and the other two detection methods is 59.89%, 51.89% and 52.03%, proving that the transmission line surface defect detection method integrating UAV technology inspection has a wider application space.

Deep-gated recurrent unit and diet network-based genome-wide association analysis for detecting the biomarkers of Alzheimer's disease

Genome-wide association analysis (GWAS) is a commonly used method to detect the potential biomarkers of Alzheimer's disease (AD). Most existing GWAS methods entail a high computational cost, disregard correlations among imaging data and correlations among genetic data, and ignore various associations between longitudinal imaging and genetic data. A novel GWAS method was proposed to identify potential AD biomarkers and address these problems. A network based on a gated recurrent unit was applied without imputing incomplete longitudinal imaging data to integrate the longitudinal data of variable lengths and extract an image representation. In this study, a modified diet network that can considerably reduce the number of parameters in the genetic network was proposed to perform GWAS between image representation and genetic data. Genetic representation can be extracted in this way. A link between genetic representation and AD was established to detect potential AD biomarkers. The proposed method was tested on a set of simulated data and a real AD dataset. Results of the simulated data showed that the proposed method can accurately detect relevant biomarkers. Moreover, the results of real AD dataset showed that the proposed method can detect some new risk-related genes of AD. Based on previous reports, no research has incorporated a deep-learning model into a GWAS framework to investigate the potential information on super-high-dimensional genetic data and longitudinal imaging data and create a link between imaging genetics and AD for detecting potential AD biomarkers. Therefore, the proposed method may provide new insights into the underlying pathological mechanism of AD.

Evaluation of the MANTA Vascular Closure Device in Transfemoral TAVI.

The MANTA vascular closure device (VCD) is a novel collagen plug-based VCD for large bore arteriotomies. The current literature regarding complication rates of this device is quite variable and mostly limited to relatively small case series. This study is retrospective analysis of the MANTA VCD-related main access site complications according to Valve Academic Research Consortium-2 (VARC-2) criteria during the hospital stay. Particular attention was paid to the detailed analysis of multislice computed tomography with regard to the anatomy of the access vessel and the puncture site itself. A total of 524 patients underwent transfemoral transcatheter aortic valve implantation (TF TAVI) including the use of the MANTA device (18F) for percutaneous vascular closure. A group of 22 patients was excluded from the study due to incomplete imaging data. During the study period, we observed 28 major (5.6%) and five minor (1.0%) MANTA device-related vascular complications. There was no patient death related to these adverse events. Female gender, vessel angulation at the puncture site, and at least moderate calcification of the dorsal vessel segment were identified as independent predictors for major complications. The MANTA device is a feasible option for vascular closure of large bore arteriotomies in patients undergoing TF TAVI or other percutaneous transfemoral interventions. Furthermore, we have identified novel predictors for device failure/complications that should be taken into account for selection of the appropriate closure device. To our knowledge, this report is one of the largest case series analyzing the use of the MANTA VCD.

Multi-Resemblance Multi-Target Low-Rank Coding for Prediction of Cognitive Decline With Longitudinal Brain Images.

An effective presymptomatic diagnosis and treatment of Alzheimer's disease (AD) would have enormous public health benefits. Sparse coding (SC) has shown strong potential for longitudinal brain image analysis in preclinical AD research. However, the traditional SC computation is time-consuming and does not explore the feature correlations that are consistent over the time. In addition, longitudinal brain image cohorts usually contain incomplete image data and clinical labels. To address these challenges, we propose a novel two-stage Multi-Resemblance Multi-Target Low-Rank Coding (MMLC) method, which encourages that sparse codes of neighboring longitudinal time points are resemblant to each other, favors sparse code low-rankness to reduce the computational cost and is resilient to both source and target data incompleteness. In stage one, we propose an online multi-resemblant low-rank SC method to utilize the common and task-specific dictionaries in different time points to immune to incomplete source data and capture the longitudinal correlation. In stage two, supported by a rigorous theoretical analysis, we develop a multi-target learning method to address the missing clinical label issue. To solve such a multi-task low-rank sparse optimization problem, we propose multi-task stochastic coordinate coding with a sequence of closed-form update steps which reduces the computational costs guaranteed by a theoretical convergence proof. We apply MMLC on a publicly available neuroimaging cohort to predict two clinical measures and compare it with six other methods. Our experimental results show our proposed method achieves superior results on both computational efficiency and predictive accuracy and has great potential to assist the AD prevention.

A novel transfer learning model for predictive analytics using incomplete multimodality data

Multimodality datasets are becoming increasingly common in various domains to provide complementary information for predictive analytics. One significant challenge in fusing multimodality data is that the multiple modalities are not universally available for all samples due to cost and accessibility constraints. This situation results in a unique data structure called an Incomplete Multimodality Dataset. We propose a novel Incomplete-Multimodality Transfer Learning (IMTL) model that builds a predictive model for each sub-cohort of samples with the same missing modality pattern, and meanwhile couples the model estimation processes for different sub-cohorts to allow for transfer learning. We develop an Expectation-Maximization (EM) algorithm to estimate the parameters of IMTL and further extend it to a collaborative learning paradigm that is specifically valuable for patient privacy preservation in health care applications. We prove two advantageous properties of IMTL: the ability for out-of-sample prediction and a theoretical guarantee for a larger Fisher information compared with models without transfer learning. IMTL is applied to diagnosis and prognosis of Alzheimer’s disease at an early stage called Mild Cognitive Impairment using incomplete multimodality imaging data. IMTL achieves higher accuracy than competing methods without transfer learning.

A Unified Tensor Framework for Clustering and Simultaneous Reconstruction of Incomplete Imaging Data

Incomplete observations in the data are always troublesome to data clustering algorithms. In fact, most of the well-received techniques are not designed to encounter such imperative scenarios. Hence, clustering of images under incomplete samples is an inquisitive yet unaddressed area of research. Therefore, the aim of this article is to design a single-stage optimization procedure for clustering as well as simultaneous reconstruction of images without breaking the intrinsic spatial structure. The method employs the self-expressiveness property of submodules, and images are stacked as the lateral slices of a three-dimensional tensor. The proposed optimization method is designed to extract a sparse t -linear combination tensor with low multirank constraint, consisting of a unique set of linear coefficients in the form of mode-3 fibers and the spectral clustering is performed on these fibers. Simultaneously, the recovery of lost samples is accomplished by twisting the entire lateral slices of the data tensor and applying a low-rank approximation on each slice. The prominence of the proposed method lies in the simultaneous execution of data clustering and reconstruction of incomplete observations in a single step. Experimental results reveal the excellence of the proposed method over state-of-the-art clustering algorithms in the context of incomplete imaging data.

Diffusion-weighted imaging for identifying patients at high risk of tumor recurrence following liver transplantation

BackgroundTumor recurrence is the major risk factor affecting post-transplant survival. In this retrospective study, we evaluate the prognostic values of magnetic resonance (MR) diffusion-weighted imaging (DWI) in liver transplantation for hepatocellular carcinoma (HCC).MethodsFrom April 2014 to September 2016, 106 HCC patients receiving living donor liver transplantation (LDLT) were enrolled. Nine patients were excluded due to postoperative death within 3 months and incomplete imaging data. The association between tumor recurrence, explant pathologic findings, and DWI parameters was analyzed (tumor-to-liver diffusion weighted imaging ratio, DWIT/L; apparent diffusion coefficients, ADC). The survival probability was calculated using the Kaplan–Meier method.ResultsSixteen of 97 patients (16%) developed tumor recurrence during the follow-up period (median of 40.9 months; range 5.2–56.5). In those with no viable tumor (n = 65) on pretransplant imaging, recurrence occurred only in 5 (7.6%) patients. Low minimum ADC values (p = 0.001), unfavorable tumor histopathology (p < 0.001) and the presence of microvascular invasion (p < 0.001) were risk factors for tumor recurrence, while ADCmean (p = 0.111) and DWIT/L (p = 0.093) showed no significant difference between the groups. An ADCmin ≤ 0.88 × 10− 3 mm2/s was an independent factor associated with worse three-year recurrence-free survival (94.4% vs. 23.8%) and overall survival rates (100% vs. 38.6%).ConclusionsQuantitative measurement of ADCmin is a promising prognostic indicator for predicting tumor recurrence after liver transplantation.

Tumor growth rate as an early indicator of the efficacy of anti-PD-1 immunotherapy in advanced melanoma.

e21050 Background: As immunotherapeutics take longer time to show clinical efficacy compared with chemotherapy and targeted therapy, it is critical to select patients with low tumor growth rate (TGR) prior immunotherapy as to get sufficient time for immunotherapeutics to play functions. However, which threshold of TGR before immunotherapy may be associated with good outcome remains unknown. Methods: Medical records from patients with advanced refractory melanoma prospectively treated in clinical trials (NCT02836795 and NCT03013101) of anti-PD-1 antibody toripalimab were analyzed. TGR was computed during the pretreatment period (reference) and the treatment period (treatment). Associations between TGR and objective response at the first evaluation, progression free survival (PFS), overall survival (OS) were computed. Results: We analyzed a total of 142 patients enrolled in these two clinical trials. Excluded for no measurable lesions in pretreatment period or incomplete imaging data through the pretreatment/treatment periods, a total of 90 patients could be explored for the relationship of TGR and the efficacy of anti-PD-1 antibody. TGR and hyperprogression were defined as Champiat S. used to make. The distribution of TGR is as follows: median 63.7 (range: -51~720). A total of 15 (16.7%), 41 (45.6%), 34 (37.8%) and 0 patients exhibited PR, SD, PD, CR, respectively. An association between lower TGR (TGR ≤55) and objective response was observed (P≤0.001). Among evaluable patients at week 8, 83.9% (13+34/56) and 26.5% (2+7/34) showed PR/SD from baseline tumor measurements for the group of TGR≤55 and TGR &gt; 55, respectively. Median PFS was 5.5 0.9m in the group of TGR≤55 compared 1.8 0.4m in the group of TGR &gt; 55 (P≤0.001). Median OS was not reached in the group of TGR≤55 group and was 15.9 1.8m in the group of TGR &gt; 55 (p = 0·02). Two patients were confirmed pseudoprogression in the follow-up. The TGR of these two patients was lower than 55. Five patients experienced hyperprogression. The TGR of each patient was 9, 12, 54, 56, 78, respectively. Conclusions: Melanoma patients with TGR≤55 prior anti-PD-1 antibody therapy are more likely to benefit from this regimen. However, it could not predict patients who may develop hyperprogression after anti-PD-1 antibody therapy. Clinical trial information: NCT02836795 and NCT03013101.

Abstract P6-17-26: Care 001: multi-center randomized open-label phase II trial of neoadjuvant trastuzumab emtansine (T-DM1) in combination with lapatinib and nab-paclitaxel compared with paclitaxel, trastuzumab and pertuzumab in HER2-neu over-expressed breast cancer patients (TEAL study)

Abstract Background: We conducted a multicenter, randomized open-label phase II neoadjuvant study of trastuzumab-emtansine (T-DM1), Lapatinib (L) and Nab Paclitaxel (Nab-P) compared to standard of care (SOC) Paclitaxel (Pac), Trastuzumab (T), and Pertuzumab (P) in patients with HER2 over-expressed breast cancer. Methods: Patients in the experimental arm received a biologic window of targeted therapies alone for 6 weeks (T-DM1 and L) followed by T-DM1 3.0 mg/kg Q3W, L 750mg oral daily and Nab-P 80 mg/m2 weekly (QW) X 12 weeks. Patients in SOC arm received targeted therapies alone for 6 weeks (T and P) followed by Pac 80mg/m2QW, T 2mg/kg QW, and P 420mg Q3W X 12 weeks. The primary objective was to evaluate the proportion of patients with residual cancer burden (RCB) 0 or 1. Key secondary objectives included correlative assessments of PIK3CA mutations, PTEN expression, and HER2 subtypes which are being reported. Results: Thirty of the 33 enrolled patients were evaluable. Patient demographics were well balanced. HER2 subtypes and altered PIK3CA (low PTEN or PIK3CA mutations) pathway were not statistically different between both arms. We have previously reported that all patients achieved RCB 0 &amp; I in the T-DM1, L and Nab-P arm, compared to SOC (100% vs. 62.5%, p 0.0035). In the SOC arm, the 6 week change in tumor size on breast MRI during targeted biologic window treatment is significantly different between the responders and non-responders based on two-sided Wilcoxon rank-sum test (p =0.0065). Consistent with literature, among ER positive patients treated with SOC, PTEN low expressers were less likely to respond (0%, 0 of 2) than PTEN high expressers (67%, 2 of 3). In the experimental arm, all patients responded regardless of PTEN. There was only 1 PIK3CA mutation on the experimental arm where all responded. Table 1:Breast MRI Tumor Size Standard of Care ArmResponseNMeanStandard Deviation95% CL MeanMinimumMaximumNo6-0.13330.4457-0.60110.3344-1.00.3Yes52.58001.88330.24154.91850.24.9Sixteen patients total were present in standard of care arm but 5 had incomplete imaging data. Conclusions: TDM1 plus L and Nab-P therapy was well tolerated with noteworthy responses in all patients, including in PTEN low expressers. Change in tumor size at 6 weeks of biologic therapies was significant between responders and non-responders and can be evaluated as a surrogate for future studies. Citation Format: Creamer SL, Patel TA, Ensor JE, Rodriguez AA, Niravath PA, Darcourt JG, Kaklamani VG, Meisel JL, Li X, Zhao J, Kuhn JG, Rosato RR, Qian W, Belcheva A, Boone T, Chang J. Care 001: multi-center randomized open-label phase II trial of neoadjuvant trastuzumab emtansine (T-DM1) in combination with lapatinib and nab-paclitaxel compared with paclitaxel, trastuzumab and pertuzumab in HER2-neu over-expressed breast cancer patients (TEAL study) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-17-26.