COVID-19 Infection Percentage Estimation from Computed Tomography Scans: Results and Insights from the International Per-COVID-19 Challenge.

COVID-19 analysis from medical imaging is an important task that has been intensively studied in the last years due to the spread of the COVID-19 pandemic. In fact, medical imaging has often been used as a complementary or main tool to recognize the infected persons. On the other hand, medical imaging has the ability to provide more details about COVID-19 infection, including its severity and spread, which makes it possible to evaluate the infection and follow-up the patient's state. CT scans are the most informative tool for COVID-19 infection, where the evaluation of COVID-19 infection is usually performed through infection segmentation. However, segmentation is a tedious task that requires much effort and time from expert radiologists. To deal with this limitation, an efficient framework for estimating COVID-19 infection as a regression task is proposed. The goal of the Per-COVID-19 challenge is to test the efficiency of modern deep learning methods on COVID-19 infection percentage estimation (CIPE) from CT scans. Participants had to develop an efficient deep learning approach that can learn from noisy data. In addition, participants had to cope with many challenges, including those related to COVID-19 infection complexity and crossdataset scenarios. This paper provides an overview of the COVID-19 infection percentage estimation challenge (Per-COVID-19) held at MIA-COVID-2022. Details of the competition data, challenges, and evaluation metrics are presented. The best performing approaches and their results are described and discussed.

Integrating Frailty Assessment to Enhance Care in Cancer Patients with Borderline Eastern Cooperative Oncology Group Performance Status.

BACKGROUND: The (ECOG) performance status (PS) is commonly used to evaluate the functional ability of patients undergoing antitumor therapy. An ECOG PS of 2, indicating patients capable of self-care but restricted strenuous activity, can complicate treatment decisions owing to concerns regarding treatment-related toxicity. We investigated whether frailty assessment could help discriminate treatment tolerance and survival outcomes in patients with an ECOG PS of 2. METHODS: We prospectively included 45 consecutive patients, aged ≥65 years, with an ECOG PS of 2, and newly diagnosed solid cancer scheduled for chemotherapy. Frailty was assessed using an eight-indicator geriatric assessment. The primary outcome was overall survival (OS) based on frailty status; secondary outcomes included treatment tolerance and toxicity. RESULTS: The median patient age was 73 years (range 65-94), and 71% had stage IV disease. Predominant frailty-related deficits were functional decline (96%), malnutrition (78%), and polypharmacy (51%). The median OS was 12.6 months (95% confidence interval [CI]: 6.8-18.4). Patients with 4-6 deficits had significantly lower OS than those with 1-3 deficits (9.9 months vs. 20.0 months, adjusted hazard ratio 2.51, 95% CI: 1.16-5.44, P = .020). Frailty significantly correlated with reduced 12-week chemotherapy competence (52% vs. 85%, adjusted odds ratio [OR] .14, 95% CI: .03-.70, P = .016) and enhanced risk of unexpected hospitalization (60% vs. 20%, adjusted OR 6.80, 95% CI: 1.64-28.1, P = .008). CONCLUSION: Our findings highlight the multifaceted nature of patients with an ECOG PS of 2 and emphasize the importance of frailty assessment for treatment outcomes.

Impact of preoperative frailty on the surgical and survival outcomes in older patients with solid cancer after elective abdominal surgery.

BACKGROUND: Frailty is common in older patients with cancer; however, its clinical impact on the survival outcomes has seldom been examined in these patients. This study aimed to investigate the association of frailty with the survival outcomes and surgical complications in older patients with cancer after elective abdominal surgery in Taiwan. METHODS: We prospectively enrolled 345 consecutive patients aged ≥65 years with newly diagnosed cancer who underwent elective abdominal surgery between 2016 and 2018. They were allocated into the fit, pre-frail, and frail groups according to comprehensive geriatric assessment (CGA) findings. RESULTS: The fit, pre-frail, and frail groups comprised 62 (18.0%), 181 (52.5%), and 102 (29.5%) patients, respectively. After a median follow-up of 48 (interquartile range, 40-53) months, the mortality rates were 12.9%, 31.5%, and 43.1%, respectively. The adjusted hazard ratio was 1.57 (95% confidence interval [CI], 0.73-3.39; p = 0.25) and 2.87 (95% CI, 1.10-5.35; p = 0.028) when the pre-frail and frail groups were compared with the fit group, respectively. The frail group had a significantly increased risk for a prolonged hospital stay (adjusted odds ratio, 2.22; 95% CI, 1.05-4.69; p = 0.022) compared with the fit group. CONCLUSION: Pretreatment frailty was significantly associated with worse survival outcomes and more surgical complications, with prolonged hospital stay, in the older patients with cancer after elective abdominal surgery. Preoperative frailty assessment can assist physicians in identifying patients at a high risk for surgical complications and predicting the survival outcomes of older patients with cancer.

Frailty assessment by two screening instruments in non-elderly patients with head and neck cancer.

PURPOSE: Frailty assessment is often overlooked in non-elderly patients with cancer, possibly due to the lack of an effective frailty screening tool. This study aimed to evaluate the performance of two modern frailty screening tools, the Flemish version of the Triage Risk Screening Tool (fTRST) and the modified 5-Item Frailty Index (mFI-5), compared to the gold standard comprehensive geriatric assessment (GA) among non-elderly patients with head and neck cancer (HNC). METHODS: We prospectively included 354 consecutive patients aged < 65 years with newly diagnosed HNC scheduled for definitive concurrent chemoradiotherapy (CCRT) at three academic hospitals in Taiwan between January 2020 and December 2022. Frailty assessment using the GA, fTRST, and mFI-5 was performed in all patients to evaluate the relationship between frailty and treatment outcomes. RESULTS: The prevalence of frailty was 27.1%, 37.0%, and 42.4% based on GA, mFI-5, and fTRST, respectively. mFI-5 and fTRST demonstrated good predictive value in identifying frail patients compared to the GA. Patients with frailty, as defined by GA, mFI-5, and fTRST, exhibited higher risks of treatment-related complications, incomplete treatment, and poorer baseline quality of life (QoL). However, only GA showed significant prognostic value for overall survival. CONCLUSIONS: Frailty assessment using fTRST and mFI-5 is valuable for predicting treatment-related adverse events, treatment tolerance, and QoL in non-elderly patients with HNC. Incorporating frailty assessment into the management of non-elderly cancer patients can aid in the identification of high-risk individuals. However, the performance of these tools varies, highlighting the need for further validation and refinement.

Effect of prophylactic tube feeding in head and neck cancer patients with high Mallampati score undergoing definitive concurrent chemoradiotherapy.

PURPOSE: There is no consensus on the selection of appropriate prophylactic tube feeding in patients with head and neck squamous cell carcinoma (HNSCC) undergoing concurrent chemoradiotherapy (CCRT). This study aimed to evaluate the effect of prophylactic tube feeding in patients with HNSCC who presented with a high Mallampati score and underwent CCRT. METHODS: We prospectively enrolled 185 consecutive patients with stage II to IVa HNSCC and a pre-treatment Mallampati score of 3 or 4 who received CCRT between August 2017 and December 2018 with follow-up data collected retrospectively. Patients were divided to either with or without prophylactic tube feeding group for comparison of treatment tolerance, toxicities, and quality of life(QOL). Propensity score matching (PSM) was used to achieve balanced covariates across the two groups. RESULTS: Of the cohort, 52 (28.1%) and 133 (71.9%) patients were allocated to the prophylactic and non-prophylactic tube feeding groups, respectively. Before and after PSM, patients in the tube feeding group had a significantly lower incidence of incomplete radiotherapy, incompletion of chemotherapy, emergency room visits, and grade 3 or higher infection, and improved symptoms of quality of life after CCRT than those in the non-tube feeding group. CONCLUSION: Prophylactic tube feeding was associated with better treatment tolerance, safety profiles, and quality of life in patients with HNSCC and high Mallampati scores who underwent CCRT. Therefore, Mallampati score might serve as a clinical tool for proactive selection of patients receiving prophylactic tube feeding in HNSCC patients upon receiving CCRT.

Hyperspectral Tensor Completion Using Low-Rank Modeling and Convex Functional Analysis.

Hyperspectral tensor completion (HTC) for remote sensing, critical for advancing space exploration and other satellite imaging technologies, has drawn considerable attention from recent machine learning community. Hyperspectral image (HSI) contains a wide range of narrowly spaced spectral bands hence forming unique electrical magnetic signatures for distinct materials, and thus plays an irreplaceable role in remote material identification. Nevertheless, remotely acquired HSIs are of low data purity and quite often incompletely observed or corrupted during transmission. Therefore, completing the 3-D hyperspectral tensor, involving two spatial dimensions and one spectral dimension, is a crucial signal processing task for facilitating the subsequent applications. Benchmark HTC methods rely on either supervised learning or nonconvex optimization. As reported in recent machine learning literature, John ellipsoid (JE) in functional analysis is a fundamental topology for effective hyperspectral analysis. We therefore attempt to adopt this key topology in this work, but this induces a dilemma that the computation of JE requires the complete information of the entire HSI tensor that is, however, unavailable under the HTC problem setting. We resolve the dilemma, decouple HTC into convex subproblems ensuring computational efficiency, and show state-of-the-art HTC performances of our algorithm. We also demonstrate that our method has improved the subsequent land cover classification accuracy on the recovered hyperspectral tensor.

Jointly Defending DeepFake Manipulation and Adversarial Attack Using Decoy Mechanism.

Highly realistic imaging and video synthesis have become possible and relatively simple tasks with the rapid growth of generative adversarial networks (GANs). GAN-related applications, such as DeepFake image and video manipulation and adversarial attacks, have been used to disrupt and confound the truth in images and videos over social media. DeepFake technology aims to synthesize high visual quality image content that can mislead the human vision system, while the adversarial perturbation attempts to mislead the deep neural networks to a wrong prediction. Defense strategy becomes difficult when adversarial perturbation and DeepFake are combined. This study examined a novel deceptive mechanism based on statistical hypothesis testing against DeepFake manipulation and adversarial attacks. First, a deceptive model based on two isolated sub-networks was designed to generate two-dimensional random variables with a specific distribution for detecting the DeepFake image and video. This research proposes a maximum likelihood loss for training the deceptive model with two isolated sub-networks. Afterward, a novel hypothesis was proposed for a testing scheme to detect the DeepFake video and images with a well-trained deceptive model. The comprehensive experiments demonstrated that the proposed decoy mechanism could be generalized to compressed and unseen manipulation methods for both DeepFake and attack detection.

Multi-head self-attention mechanism enabled individualized hemoglobin prediction and treatment recommendation systems in anemia management for hemodialysis patients.

Anemia is a critical complication in hemodialysis patients, but the response to erythropoietin-stimulating agents (ESA) treatment varies from patient to patient and is not linear across different time points. The aim of this study was to develop deep learning algorithms for individualized anemia management. We retrospectively collected 36,677 data points from 623 hemodialysis patients, including clinical data, laboratory values, hemoglobin levels, and previous ESA doses. To reduce the computational complexity associated with recurrent neural networks (RNN) in processing time-series data, we developed neural networks based on multi-head self-attention mechanisms in an efficient and effective hemoglobin prediction model. Our proposed model achieved a more accurate hemoglobin prediction than the state-of-the-art RNN model, as shown by the smaller mean absolute error (MAE) of hemoglobin (0.451 vs. 0.593 g/dL, p = 0.014). In ESA (including darbepoetin and epoetin) dose recommendation, the simulation results by our model revealed a higher rate of achieved hemoglobin targets (physician prescription vs. model: 86.3 % vs. 92.7 %, p < 0.001), a lower rate of hemoglobin levels below 10 g/dL (13.7 % vs. 7.3 %, p < 0.001) and smaller change in hemoglobin levels (0.6 g/dL vs. 0.4 g/dL, p < 0.001) in all patients. Our model holds great potential for individualized anemia management as a computerized clinical decision support system for hemodialysis patients. Further external validation with other datasets and prospective clinical utility studies are warranted.

Investigation of Parameter Sensitivity and the Physical Mechanism for the Formation of a Core-Skin-Core (CSC) Structure in Two-Stage Co-Injection Molding.

One of the main challenges in co-injection molding is how to predict the skin to core morphology accurately and then manage it properly, especially after skin material has been broken through. In this study, the formation of the Core-Skin-Core (CSC) structure and its physical mechanism in a two-stage co-injection molding has been studied based on the ASTM D638 TYPE V system by using both numerical simulation and experimental observation. Results showed that when the skin to core ratio is selected properly (say 30/70), the CSC structure can be observed clearly at central location for 30SFPP/30SFPP system. When the skin to core ratio and operation conditions are fixed, regardless of material arrangement (including 30SFPP/30SFPP; PP/PP; 30SFPP/PP; and PP/30SFPP systems), the morphologies of the CSC structures are very close for all systems. This CSC structure can be further validated by using µ-CT scan and image analysis technologies perfectly. Furthermore, the influences of various operation parameters on the CSC structure variation have been investigated. Results exhibited that the CSC structure does not change significantly irrespective of the flow rate changing, melt temperature varying, or even mold temperature being modified. Moreover, the mechanism to generate the CSC structure can be derived using the melt front movement of the numerical simulation. It is worth noting that after the skin material was broken through, the core material travelled ahead with fountain flow to occupy the flow front. In the same period, the proper amount of skin material with certain inertia of enough kinetic energy will keep going to penetrate the new coming core material to travel until the end of filling. It ends up with this special CSC structure.

Adjuvant Chemoradiotherapy Associated with Improved Overall Survival in Resected Esophageal Squamous Cell Carcinoma after Neoadjuvant Chemoradiotherapy in Intensity-Modulated Radiotherapy Era.

BACKGROUND: The prognosis of patients with resected esophageal squamous cell carcinoma after neoadjuvant chemoradiotherapy is particularly poor in those who were staged as ypT3/T4 and/or ypN+. This study investigated whether adjuvant chemoradiotherapy was associated with improved clinical outcomes in these patients. METHODS: we identified patients with esophageal squamous cell carcinoma who were staged as ypT3/T4 and/or ypN+ after being treated with neoadjuvant chemoradiotherapy followed by esophagectomy between the years 2013 and 2019. Patients were divided into two groups based on whether they received adjuvant chemoradiotherapy. The Kaplan-Meier method and Cox regression modeling were performed for survival analyses and multivariable analysis, respectively. RESULTS: 76 eligible patients were included in the analyses. The median follow-up for the study cohort was 43.4 months. On Kaplan-Meier analyses of the overall population, adjuvant chemoradiotherapy was associated with significantly improved median overall survival (31.7 months vs. 16.3 months, p = 0.036). On Kaplan-Meier analyses of the 35 matched pairs generated by propensity score matching, adjuvant chemoradiotherapy was associated with significantly longer median overall survival (31.7 months vs. 14.3 months; p = 0.004) and median recurrence-free survival (18.9 months vs. 11.7 months; p = 0.020). In multivariable analysis, adjuvant chemoradiotherapy was independently associated with a 60% reduction in mortality (p = 0.003) and a 48% reduction in risk of recurrence (p = 0.035) after adjusting for putative confounders. In addition, microscopic positive resection margin and Mandard tumor regression grade 3-4 were independently associated with increased mortality and risk of recurrence. While a greater number of lymph nodes dissected was independently associated with significantly improved overall survival, the number of positive lymph nodes was independently associated with significantly worse overall survival and a trend (p = 0.058) towards worse recurrence-free survival. CONCLUSIONS: This study demonstrated that adjuvant CRT was independently associated with a significantly improved survival and lower risk of recurrence than observation in esophageal squamous cell carcinoma patients staged as ypT3 and/or ypN+ after receiving neoadjuvant chemoradiotherapy and radical surgery. The results of this study have implications for the design of future clinical trials and may improve treatment outcomes of patients in this setting who cannot afford or are without access to adjuvant nivolumab.

Predictive Value of Albumin and Neutrophil-to-Lymphocyte Ratio Score for Treatment Completeness and Safety Profiles in Patients With Head and Neck Cancer Receiving Definitive Concurrent Chemoradiotherapy.

BACKGROUND/AIM: Malnutrition and inflammation are common conditions in patients with head and neck cancer (HNC). This study aimed to evaluate the predictive value of albumin combined with neutrophil-lymphocyte ratio (NLR), referring to the albumin-NLR score (ANS), in the prediction of treatment completeness and safety profiles in HNC patients receiving definitive concurrent chemoradiotherapy (CCRT). PATIENTS AND METHODS: 461 consecutive HNC patients who received CCRT between 2016 and 2017 at three medical centers in Taiwan were prospectively enrolled and divided into three different groups based on their pretreatment ANS (ANS 0, high albumin and low NLR; ANS 1, low albumin or high NLR; and ANS 2, low albumin and high NLR) for treatment completeness and safety profiles comparison. RESULTS: Overall, 46 patients (10.0%) had incomplete CCRT treatment. Patients in the ANS 2 group experienced a higher rate of incomplete CCRT (20.9%) than those in the ANS 1 (7.4%) and ANS 0 (3.5%) groups. ANS had a better discriminatory ability in predicting CCRT completeness in terms of -2 log-likelihood value, chi-square value, and c-index than the prognostic nutritional index. Patients in the ANS 2 group had significantly higher incidences of grade 3 or higher leukopenia, anemia, neutropenia, thrombocytopenia, non-neutropenic infection, and hypokalemia than those in the other two ANS groups. CONCLUSION: Our study showed that the ANS can accurately predict the treatment completeness of CCRT in patients with HNC and can be widely used as a simple predictor of treatment tolerance and safety profiles in patients with HNC undergoing CCRT.

Potential Utility of the Mallampati Score for Prediction of Treatment Compliance and Safety Profiles of Patients With Head and Neck Cancer Undergoing Definitive Concurrent Chemoradiotherapy.

BACKGROUND/AIM: Restriction of mouth opening (RMO) is a common manifestation of head and neck cancer (HNC) and a poor prognostic factor following concurrent chemoradiotherapy (CCRT) of patients. This study aimed to explore whether the Mallampati score, a visual assessment of the distance from the tongue base to the roof of the mouth, can be used as a surrogate for RMO in predicting treatment outcomes in patients with HNC undergoing CCRT. PATIENTS AND METHODS: A total of 461 consecutive patients who received definitive CCRT for the treatment of locally advanced HNC between August 2016 and December 2017 at Chang Gung Memorial Hospital in Taiwan (Linkou, Keelung, and Kaohsiung branches) were enrolled in this prospective study. Patients were allocated by the pre-treatment Mallampati score of 1 or 2 (n=24) vs. 3 or 4 (n=207) to compare treatment compliance and treatment-related complications. RESULTS: Patients in the Mallampati score of 3 or 4 group had a higher prevalence of betel quid chewing, oral cavity and oropharynx cancers, advanced tumor stage, poorer performance status, and were more likely to receive platinum monotherapy during CCRT. Patients in the Mallampati score of 3 or 4 group had a 2.08-fold (p=0.002) hazard ratio (HR) for overall survival compared to those in the score of 1 or 2 group in the univariate analysis, the difference remained significant in multivariate analysis (adjusted HR=1.61; 95% CI=1.02-2.61; p=0.047). Patients in the Mallampati score 3 or 4 group had a 2.36-fold (95% CI=1.07-5.19; p=0.033) increased likelihood of incomplete chemotherapy, 2.44-fold (95% CI=1.17-5.06; p=0.017) increased likelihood of incomplete radiotherapy, and 1.84-fold (95% CI=1.18-2.87; p=0.007) risk of unexpected hospitalization compared to those with a Mallampati score of 1 or 2 in multivariate analysis. CONCLUSION: Patients with HNC with higher pre-treatment Mallampati scores had poorer survival outcomes and were at a higher risk of treatment incompletion and treatment-related toxicities when undergoing CCRT. Our results support the utility of Mallampati score as a surrogate for measuring RMO to predict survival outcomes, treatment compliance, and safety profiles in patients with HNC undergoing CCRT.

ADAM Challenge: Detecting Age-Related Macular Degeneration From Fundus Images.

Age-related macular degeneration (AMD) is the leading cause of visual impairment among elderly in the world. Early detection of AMD is of great importance, as the vision loss caused by this disease is irreversible and permanent. Color fundus photography is the most cost-effective imaging modality to screen for retinal disorders. Cutting edge deep learning based algorithms have been recently developed for automatically detecting AMD from fundus images. However, there are still lack of a comprehensive annotated dataset and standard evaluation benchmarks. To deal with this issue, we set up the Automatic Detection challenge on Age-related Macular degeneration (ADAM), which was held as a satellite event of the ISBI 2020 conference. The ADAM challenge consisted of four tasks which cover the main aspects of detecting and characterizing AMD from fundus images, including detection of AMD, detection and segmentation of optic disc, localization of fovea, and detection and segmentation of lesions. As part of the ADAM challenge, we have released a comprehensive dataset of 1200 fundus images with AMD diagnostic labels, pixel-wise segmentation masks for both optic disc and AMD-related lesions (drusen, exudates, hemorrhages and scars, among others), as well as the coordinates corresponding to the location of the macular fovea. A uniform evaluation framework has been built to make a fair comparison of different models using this dataset. During the ADAM challenge, 610 results were submitted for online evaluation, with 11 teams finally participating in the onsite challenge. This paper introduces the challenge, the dataset and the evaluation methods, as well as summarizes the participating methods and analyzes their results for each task. In particular, we observed that the ensembling strategy and the incorporation of clinical domain knowledge were the key to improve the performance of the deep learning models.

Semantic segmentation guided detector for segmentation, classification, and lesion mapping of acute ischemic stroke in MRI images.

BACKGROUND AND PURPOSE: MRI images timely and accurately reflect ischemic injuries to the brain tissues and, therefore, can support clinical decision-making of acute ischemic stroke (AIS). To maximize the information provided by the MRI images, we leverage deep learning models to segment, classify, and map lesion distributions of AIS. METHODS: We evaluated brain MRI images of AIS patients from 2017 to 2020 at a tertiary teaching hospital and developed the Semantic Segmentation Guided Detector Network (SGD-Net), composed of the first U-shaped model for segmentation in diffusion-weighted imaging (DWI) and the second model for binary classification of lesion size (lacune vs. non-lacune) and circulatory territory of lesion location (anterior vs. posterior circulation). Next, we modified the two-stage deep learning model into SGD-Net Plus by automatically segmenting AIS lesions in DWI images and registering the lesion in T1-weighted images and the brain atlases. RESULTS: The final enrollment (216 patients with 4606 slices) was divided into 80% for model development and 20% for testing. S1 model segmented AIS lesions in DWI images accurately with a pixel accuracy > 99% (Dice 0.806-0.828 and IoU 0.675-707). In comprehensive evaluation of classification performance, the two-stage SGD-Net outperformed the traditional one-stage models in classifying AIS lesion size (accuracy 0.867-0.956 vs. 0.511-0.867, AUROC 0.962-0.992 vs. 0.528-0.937, AUPRC 0.964-0.994 vs. 0.549-0.938) and location (accuracy 0.860-0.930 vs. 0.326-0.721, AUROC 0.936-0.988 vs. 0.493-0.833, AUPRC 0.883-0.978 vs. 0.365-0.695). The precise lesion segmentation at the first stage of the deep learning model was the basis for further application. After that, the modified two-stage model SGD-Net Plus accurately reported the volume, region percentage, and lesion percentage of each region on the selected brain atlas. Its reports provided clear descriptions and quantifications of the AIS-related brain injuries on white matter tracts, Brodmann areas, and cytoarchitectonic areas. CONCLUSION: Domain knowledge-oriented design of artificial intelligence applications can deepen our understanding of patients' conditions and strengthen the use of MRI for patient care. SGD-Net precisely segments AIS lesions on DWI and accurately classifies the lesions. In addition, SGD-Net Plus maps the AIS lesions and quantifies their occupancy in each brain region. They are practical tools to meet the clinical needs and enrich educational resources of neuroimage.

Development and validation of a prognostic nomogram to predict survival in patients with advanced pancreatic cancer receiving second-line palliative chemotherapy.

BACKGROUND AND AIM: Given that a wide variation in tumor response rates and survival times suggests heterogeneity among the patients with advanced pancreatic cancer (APC) who underwent second-line (L2) chemotherapy, it is a challenge in clinical practice to identify patients who will receive the most benefit from L2 treatment. METHODS: We selected 183 APC patients who received L2 palliative chemotherapy between 2010 and 2016 from a medical center as the development cohort. A Cox proportional hazard model was used to identify the prognostic factors and construct the nomogram. An independent cohort of 166 patients from three other hospitals was selected for external validation. RESULTS: The nomogram was based on eight independent prognostic factors from the multivariate Cox model: sex, Eastern Cooperative Oncology Group performance status, reason for first-line treatment discontinuation, duration of first-line treatment, neutrophil-to-lymphocyte ratio, tumor stage, body mass index, and serum carbohydrate antigen 19-9 levels at the beginning of L2 treatment. The model exhibited good discrimination ability, with a C-index of 0.733 (95% confidence interval, 0.681-0.785) and 0.724 (95% confidence interval, 0.661-0.787) in the development and validation cohorts, respectively. The calibration plots of the development and validation cohorts showed optimal agreement between model prediction and actual observation in predicting survival probability at 6 months, 1 year, and 2 years. CONCLUSIONS: This study developed and externally validated a prognostic model that accurately predicts the survival outcome of APC patients before L2 palliative chemotherapy, which could assist in clinical decision-making, counseling for treatment, and most importantly, prognostic stratification of patients.

SiGAN: Siamese Generative Adversarial Network for Identity-Preserving Face Hallucination.

Though generative adversarial networks (GANs) can hallucinate high-quality high-resolution (HR) faces from low-resolution (LR) faces, they cannot ensure identity preservation during face hallucination, making the HR faces difficult to recognize. To address this problem, we propose a Siamese GAN (SiGAN) to reconstruct HR faces that visually resemble their corresponding identities. On top of a Siamese network, the proposed SiGAN consists of a pair of two identical generators and one discriminator. We incorporate reconstruction error and identity label information in the loss function of SiGAN in a pairwise manner. By iteratively optimizing the loss functions of the generator pair and the discriminator of SiGAN, we not only achieve visually-pleasing face reconstruction but also ensure that the reconstructed information is useful for identity recognition. Experimental results demonstrate that SiGAN significantly outperforms existing face hallucination GANs in objective face verification performance while achieving promising visual-quality reconstruction. Moreover, for input LR faces with unseen identities that are not part of the training dataset, SiGAN can still achieve reasonable performance.

The Effect of Music Listening on Pain, Heart Rate Variability, and Range of Motion in Older Adults After Total Knee Replacement.

The purpose of this study was to investigate the effects that listening and not listening to music had on pain relief, heart rate variability (HRV), and knee range of motion in total knee replacement (TKR) patients who underwent continuous passive motion (CPM) rehabilitation. We adopted a single-group quasi-experimental design. A sample of 49 TKR patients listened to music for 25 min during one session of CPM and no music during another session of CPM the same day for a total of 2 days. Results indicated that during CPM, patients exhibited a significant decrease in the pain level ( p < .05), an increase in the CPM knee flexion angle ( p < .05), a decrease in the low-frequency/high-frequency ratio (LF/HF) and normalized LF (nLF) of the HRV ( p < .01), and an increase in the normalized HF (nHF) and standard deviation of all normal-to-normal intervals (SDNN; p < .01) when listening to music compared with no music. This study demonstrated that listening to music can effectively decrease pain during CPM rehabilitation and improve the joint range of motion in patients who underwent TKR surgery.

Is Intensity Inhomogeneity Correction Useful for Classification of Breast Cancer in Sonograms Using Deep Neural Network?

The sonogram is currently an effective cancer screening and diagnosis way due to the convenience and harmlessness in humans. Traditionally, lesion boundary segmentation is first adopted and then classification is conducted, to reach the judgment of benign or malignant tumor. In addition, sonograms often contain much speckle noise and intensity inhomogeneity. This study proposes a novel benign or malignant tumor classification system, which comprises intensity inhomogeneity correction and stacked denoising autoencoder (SDAE), and it is suitable for small-size dataset. A classifier is established by extracting features in the multilayer training of SDAE; automatic analysis of imaging features by the deep learning algorithm is applied on image classification, thus allowing the system to have high efficiency and robust distinguishing. In this study, two kinds of dataset (private data and public data) are used for deep learning models training. For each dataset, two groups of test images are compared: the original images and the images after intensity inhomogeneity correction, respectively. The results show that when deep learning algorithm is applied on the sonograms after intensity inhomogeneity correction, there is a significant increase of the tumor distinguishing accuracy. This study demonstrated that it is important to use preprocessing to highlight the image features and further give these features for deep learning models. In this way, the classification accuracy will be better to just use the original images for deep learning.

Effectiveness of Music Listening in Patients With Total Knee Replacement During CPM Rehabilitation.

AIM: This study investigated the effects of music listening on the anxiety, heart rate variability (HRV), and joint range of motion (ROM) of patients undergoing continuous passive motion (CPM) after total knee replacement surgery. METHOD: An experimental design was used. Participants in the experimental group (n = 49) listened to music from 10 min before receiving CPM until the end of the session (25 min in total) on the first and second day following surgery, whereas participants in the control group (n = 42) did not listen to music but rested quietly in bed starting 10 min before and throughout CPM. RESULTS: Compared with the control group, the experimental group exhibited significantly lower anxiety levels (p < .05) and increased CPM angles (p < .05) during treatment and increased active flexion ROM (p < .05) upon discharge. The low-frequency (LF)/high-frequency (HF) power ratio, normalized LF HRV, and normalized HF HRV of the two groups differed significantly, indicating that the patients in the experimental group had greater parasympathetic activity compared with those in the control group. CONCLUSION: Music listening can effectively reduce patient anxiety and enhance the ROM of their joints during postoperative rehabilitation. Health-care practitioners should consider including music listening as a routine practice for postoperative rehabilitation following orthopedic surgery.

Temporally coherent superresolution of textured video via dynamic texture synthesis.

This paper addresses the problem of hallucinating the missing high-resolution (HR) details of a low-resolution (LR) video while maintaining the temporal coherence of the reconstructed HR details using dynamic texture synthesis (DTS). Most existing multiframe-based video superresolution (SR) methods suffer from the problem of limited reconstructed visual quality due to inaccurate subpixel motion estimation between frames in an LR video. To achieve high-quality reconstruction of HR details for an LR video, we propose a texture-synthesis (TS)-based video SR method, in which a novel DTS scheme is proposed to render the reconstructed HR details in a temporally coherent way, which effectively addresses the temporal incoherence problem caused by traditional TS-based image SR methods. To further reduce the complexity of the proposed method, our method only performs the TS-based SR on a set of key frames, while the HR details of the remaining nonkey frames are simply predicted using the bidirectional overlapped block motion compensation. After all frames are upscaled, the proposed DTS-SR is applied to maintain the temporal coherence in the HR video. Experimental results demonstrate that the proposed method achieves significant subjective and objective visual quality improvement over state-of-the-art video SR methods.