Treatment outcome of elderly patients (≥ 78 years) with head and neck squamous cell carcinoma: A single center experience.

Patients older than the expected age of the local population generally have limited life expectancy. The optimal treatment approach for very elderly patients with head and neck cancer remains uncertain. This study retrospectively analyzed patients over 78 years old, the expected age in 2019 for Chinese individuals, who underwent treatment for head and neck cancer at a tertiary cancer center in China. The study compared the overall survival rates among different treatment groups. The findings revealed that among patients eligible for surgery, radical resection yielded better outcomes compared to radiotherapy-based treatments, with a hazard ratio of 0.362 (95% CI 0.160-0.819, P = 0.015). Among patients who received radiotherapy, those who received a total dose exceeding 60 Gy had a significantly longer survival compared to those who received palliative doses, with median survival time of 31 months versus 14 months (P = 0.003). Among 78 patients who underwent conventional fractionated radiotherapy (CFRT), 15 patients (19.23%) experienced unscheduled treatment breaks with a median duration of 12 days. However, these treatment breaks did not appear to impact survival (P > 0.1). The study also suggested that altered fractionated radiotherapy, including hypofractionated radiotherapy (hypo-RT), could be a viable alternative to CFRT, offering similar survival outcomes with reduced treatment duration. In conclusion, eligible patients should be treated with curative intent, even if they are older than the expected age of the local population. When radiotherapy is indicated, altered fractionation, particularly hypo-RT, may be a favorable option to consider.

Prognostic effects of different treatment modalities for hypopharyngeal squamous cell carcinoma: Experience of two tertiary hospitals in Southwestern China.

Background: The prognostic effects of different treatment modalities on patients with hypopharyngeal squamous cell carcinoma (HPSCC) remain unclear. Methods: HPSCC patients diagnosed and treated at either West China Hospital or Sichuan Cancer Hospital between January 1, 2009, and December 31, 2019, were enrolled in this retrospective, real-world study. Survival rates were presented using Kaplan-Meier curves and compared using log-rank tests. Univariable and multivariable Cox proportional hazards regression models were used to identify the predictors of overall survival (OS). Subgroup analyses were conducted for patients with advanced-stage HPSCC (stages III and IV and category T4). Results: A total of 527 patients with HPSCC were included. Patients receiving SRC (surgery, radiotherapy [RT], and chemotherapy) showed the best OS (p < 0.0001). In comparison with RT alone, both surgery alone (all cases: hazard ratio [HR] = 0.39, p = 0.0018; stage IV cases: HR = 0.38, p = 0.0085) and surgery-based multimodality treatment (SBMT; all cases: HR = 0.27, p < 0.0001; stage IV cases: HR = 0.30, p = 0.00025) showed prognostic benefits, while SBMT also showed survival priority over chemoradiotherapy (CRT; all cases: HR = 0.52, p < 0.0001; stage IV cases: HR = 0.59, p = 0.0033). Moreover, patients who underwent surgery alone had comparable OS to those who underwent SBMT (all patients: p = 0.13; stage IV cases: p = 0.34), while CRT yielded similar prognostic outcomes as RT alone (all patients: p = 0.054; stage IV cases: p = 0.11). Conclusions: Surgery alone was comparable to SBMT and superior to RT/CRT in terms of OS in patients with HPSCC. We suggest that surgery should be encouraged for the treatment of HPSCC, even in patients with advanced-stage disease.

The roles of tissue resident macrophages in health and cancer.

As integral components of the immune microenvironment, tissue resident macrophages (TRMs) represent a self-renewing and long-lived cell population that plays crucial roles in maintaining homeostasis, promoting tissue remodeling after damage, defending against inflammation and even orchestrating cancer progression. However, the exact functions and roles of TRMs in cancer are not yet well understood. TRMs exhibit either pro-tumorigenic or anti-tumorigenic effects by engaging in phagocytosis and secreting diverse cytokines, chemokines, and growth factors to modulate the adaptive immune system. The life-span, turnover kinetics and monocyte replenishment of TRMs vary among different organs, adding to the complexity and controversial findings in TRMs studies. Considering the complexity of tissue associated macrophage origin, macrophages targeting strategy of each ontogeny should be carefully evaluated. Consequently, acquiring a comprehensive understanding of TRMs' origin, function, homeostasis, characteristics, and their roles in cancer for each specific organ holds significant research value. In this review, we aim to provide an outline of homeostasis and characteristics of resident macrophages in the lung, liver, brain, skin and intestinal, as well as their roles in modulating primary and metastatic cancer, which may inform and serve the future design of targeted therapies.

Masked Conditional Variational Autoencoders for Chromosome Straightening.

Karyotyping is of importance for detecting chromosomal aberrations in human disease. However, chromosomes easily appear curved in microscopic images, which prevents cytogeneticists from analyzing chromosome types. To address this issue, we propose a framework for chromosome straightening, which comprises a preliminary processing algorithm and a generative model called masked conditional variational autoencoders (MC-VAE). The processing method utilizes patch rearrangement to address the difficulty in erasing low degrees of curvature, providing reasonable preliminary results for the MC-VAE. The MC-VAE further straightens the results by leveraging chromosome patches conditioned on their curvatures to learn the mapping between banding patterns and conditions. During model training, we apply a masking strategy with a high masking ratio to train the MC-VAE with eliminated redundancy. This yields a non-trivial reconstruction task, allowing the model to effectively preserve chromosome banding patterns and structure details in the reconstructed results. Extensive experiments on three public datasets with two stain styles show that our framework surpasses the performance of state-of-the-art methods in retaining banding patterns and structure details. Compared to using real-world bent chromosomes, the use of high-quality straightened chromosomes generated by our proposed method can improve the performance of various deep learning models for chromosome classification by a large margin. Such a straightening approach has the potential to be combined with other karyotyping systems to assist cytogeneticists in chromosome analysis.

Gradient-aware learning for joint biases: Label noise and class imbalance.

Data biases such as class imbalance and label noise always exist in large-scale datasets in real-world. These problems bring huge challenges to deep learning methods. Some previous works adopted loss re-weighting, sample re-weighting, or data-dependent regularization to mitigate the influence of these training biases. But these methods usually pay more attention to class imbalance problem when both the class imbalance and label noise exist in training set simultaneously. These methods may overfit noisy labels, which leads to a great degradation in performance. In this paper, we propose a gradient-aware learning method for the combination of the two biases. During the training process, we update only a part of crucial parameters regularly and rectify the update direction of the rest redundant parameters. This update rule is conducted both in the encoder and classifier of the deep network to decouple label noise and class imbalance implicitly. The experimental results verify the effectiveness of the proposed method on synthetic and real-world data biases.

Long-term outcomes of metastasis-directed stereotactic body radiation therapy in metastatic nasopharyngeal carcinoma.

BACKGROUND: The study aims to evaluate the outcomes of metastasis-directed stereotactic body radiation therapy (SBRT) in metastatic nasopharyngeal carcinoma (mNPC). METHODS: We reviewed all SBRT conducted in patients with mNPC in our institution between 2013 and 2022. Systemic therapy was performed with chemotherapy with or without anti-programmed death-1 (PD-1) therapy. Local treatment delivered with ablative purpose in stereotactic setting with dose/fraction ≥5 Gy was evaluated. Kaplan-Meier analyses were used to determine the rates of local control (LC), progression-free survival (PFS), and overall survival (OS). Univariate and multivariate analyses were performed by Cox regression. RESULTS: A total of 54 patients with 76 metastatic sites receiving SBRT were analyzed. Median follow-up was 49 months. The 3-year LC, PFS, and OS rates were 89.1%, 29.4%, and 57.9%, respectively. Adding a PD-1 inhibitor to SBRT tended to prolong median OS (50.1 vs. 32.2 months, p = 0.068). Patients receiving a biological effective dose (BED, α/ß = 10) ≥ 80 Gy had a significantly longer median OS compared to those who received a lower dose (not reached vs. 29.5 months, p = 0.004). Patients with oligometastases (1-5 metastases) had a better median OS (not reached vs. 29.5 months, p < 0.001) and PFS (34.3 vs. 4.6 months, p < 0.001). Pretreatment EBV-DNA and maintenance therapy were also significant predictors for OS. CONCLUSIONS: Metastatic NPC patients could benefit from metastases-directed SBRT in combination with systemic therapy.

CDDSA: Contrastive domain disentanglement and style augmentation for generalizable medical image segmentation.

Generalization to previously unseen images with potential domain shifts is essential for clinically applicable medical image segmentation. Disentangling domain-specific and domain-invariant features is key for Domain Generalization (DG). However, existing DG methods struggle to achieve effective disentanglement. To address this problem, we propose an efficient framework called Contrastive Domain Disentanglement and Style Augmentation (CDDSA) for generalizable medical image segmentation. First, a disentangle network decomposes the image into domain-invariant anatomical representation and domain-specific style code, where the former is sent for further segmentation that is not affected by domain shift, and the disentanglement is regularized by a decoder that combines the anatomical representation and style code to reconstruct the original image. Second, to achieve better disentanglement, a contrastive loss is proposed to encourage the style codes from the same domain and different domains to be compact and divergent, respectively. Finally, to further improve generalizability, we propose a style augmentation strategy to synthesize images with various unseen styles in real time while maintaining anatomical information. Comprehensive experiments on a public multi-site fundus image dataset and an in-house multi-site Nasopharyngeal Carcinoma Magnetic Resonance Image (NPC-MRI) dataset show that the proposed CDDSA achieved remarkable generalizability across different domains, and it outperformed several state-of-the-art methods in generalizable segmentation. Code is available at https://github.com/HiLab-git/DAG4MIA.

Comprehensive Evaluation of a Deep Learning Model for Automatic Organs-at-Risk Segmentation on Heterogeneous Computed Tomography Images for Abdominal Radiation Therapy.

PURPOSE: Our purpose was to develop a deep learning model (AbsegNet) that produces accurate contours of 16 organs at risk (OARs) for abdominal malignancies as an essential part of fully automated radiation treatment planning. METHODS AND MATERIALS: Three data sets with 544 computed tomography scans were retrospectively collected. Data set 1 was split into 300 training cases and 128 test cases (cohort 1) for AbsegNet. Data set 2, including cohort 2 (n = 24) and cohort 3 (n = 20), were used to validate AbsegNet externally. Data set 3, including cohort 4 (n = 40) and cohort 5 (n = 32), were used to clinically assess the accuracy of AbsegNet-generated contours. Each cohort was from a different center. The Dice similarity coefficient and 95th-percentile Hausdorff distance were calculated to evaluate the delineation quality for each OAR. Clinical accuracy evaluation was classified into 4 levels: no revision, minor revisions (0% < volumetric revision degrees [VRD] ≤ 10%), moderate revisions (10% ≤ VRD < 20%), and major revisions (VRD ≥20%). RESULTS: For all OARs, AbsegNet achieved a mean Dice similarity coefficient of 86.73%, 85.65%, and 88.04% in cohorts 1, 2, and 3, respectively, and a mean 95th-percentile Hausdorff distance of 8.92, 10.18, and 12.40 mm, respectively. The performance of AbsegNet outperformed SwinUNETR, DeepLabV3+, Attention-UNet, UNet, and 3D-UNet. When experts evaluated contours from cohorts 4 and 5, 4 OARs (liver, kidney_L, kidney_R, and spleen) of all patients were scored as having no revision, and over 87.5% of patients with contours of the stomach, esophagus, adrenals, or rectum were considered as having no or minor revisions. Only 15.0% of patients with colon and small bowel contours required major revisions. CONCLUSIONS: We propose a novel deep-learning model to delineate OARs on diverse data sets. Most contours produced by AbsegNet are accurate and robust and are, therefore, clinically applicable and helpful to facilitate radiation therapy workflow.

Developing a predictive nomogram and web-based survival calculator for locally advanced hypopharyngeal cancer: A propensity score-adjusted, population-based study.

Understanding the clinical features and accurately predicting the prognosis of patients with locally advanced hypopharyngeal squamous cell carcinoma (LA-HPSCC) is important for patient centered decision-making. This study aimed to create a multi-factor nomogram predictive model and a web-based calculator to predict post-therapy survival for patients with LA-HPSCC. A retrospective cohort study analyzing Surveillance, Epidemiology, and End Results (SEER) database from 2004 to 2015 for patients diagnosed with LA-HPSCC was conducted and randomly divided into a training and a validation group (7:3 ratio). The external validation cohort included 276 patients from Sichuan Cancer Hospital, China. The Least Absolute Shrinkage and Selection Operator (LASSO)-Cox regression analysis was used to identify independent factors associated with overall survival (OS) and cancer-specific survival (CSS), and nomogram models and web-based survival calculators were constructed. Propensity score matching (PSM) was used to compare survival with different treatment options. A total of 2526 patients were included in the prognostic model. The median OS and CSS for the entire cohort were 20 (18.6-21.3) months and 24 (21.7-26.2) months, respectively. Nomogram models integrating the seven factors demonstrated high predictive accuracy for 3-year and 5-year survival. PSM found that patients who received surgery-based curative therapy had better OS and CSS than those who received radiotherapy-based treatment (median survival times: 33 months vs 18 months and 40 months vs 22 months, respectively). The nomogram model accurately predicted patient survival from LA-HPSCC. Surgery with adjuvant therapy yielded significantly better survival than definitive radiotherapy. and should be prioritized over definitive radiotherapy.

Antimicrobial properties of heterojunction BiSnSbO6-ZnO composites in wastewater treatment.

BiSnSbO6-ZnO composite photocatalytic material with type II heterojunction structure was synthesized by a simple solid-phase sintering method, it was characterized by XRD, UV-vis, and PT methods. The photocatalytic antibacterial experiments were carried out under LED light irradiation. The experimental results showed that the photocatalytic antibacterial properties of BiSnSbO6-ZnO composites against bacteria and fungi were significantly stronger than those of single BiSnSbO6 and ZnO. Under light conditions, the antibacterial efficiencies of 500 mg/L BiSnSbO6-ZnO composites against E. coli, S. aureus, and P. aeruginosa reached 99.63%, 100%, and 100% for 6 h, 4 h, and 4 h, respectively. The best antibacterial concentration of BiSnSbO6-ZnO composite against the eukaryotic microorganism Candida albicans was 250 mg/L, and the antibacterial efficiency reached the highest 63.8% at 6 h. Antibacterial experiments were carried out on domestic livestock and poultry wastewater, which showed that the BiSnSbO6-ZnO composite photocatalytic material has broad-spectrum antibacterial activity against bacteria, and the antibacterial effect has species differences. Through the MTT experiment, it is proved that the prepared BiSnSbO6-ZnO composite photocatalytic material has no toxicity at the experimental concentration. According to the free radical scavenging experiment and SEM observation of the morphological changes of the bacteria after light treatment, the prepared BiSnSbO6-ZnO composite photocatalytic material can generate active species OH, h+, and e- through light irradiation to achieve the purpose of sterilization, where e- play a major role, indicating that the BiSnSbO6-ZnO composite photocatalytic material has broad application prospects in the actual antibacterial field.

Development of a tongue image-based machine learning tool for the diagnosis of gastric cancer: a prospective multicentre clinical cohort study.

Background: Tongue images (the colour, size and shape of the tongue and the colour, thickness and moisture content of the tongue coating), reflecting the health state of the whole body according to the theory of traditional Chinese medicine (TCM), have been widely used in China for thousands of years. Herein, we investigated the value of tongue images and the tongue coating microbiome in the diagnosis of gastric cancer (GC). Methods: From May 2020 to January 2021, we simultaneously collected tongue images and tongue coating samples from 328 patients with GC (all newly diagnosed with GC) and 304 non-gastric cancer (NGC) participants in China, and 16 S rDNA was used to characterize the microbiome of the tongue coating samples. Then, artificial intelligence (AI) deep learning models were established to evaluate the value of tongue images and the tongue coating microbiome in the diagnosis of GC. Considering that tongue imaging is more convenient and economical as a diagnostic tool, we further conducted a prospective multicentre clinical study from May 2020 to March 2022 in China and recruited 937 patients with GC and 1911 participants with NGC from 10 centres across China to further evaluate the role of tongue images in the diagnosis of GC. Moreover, we verified this approach in another independent external validation cohort that included 294 patients with GC and 521 participants with NGC from 7 centres. This study is registered at ClinicalTrials.gov, NCT01090362. Findings: For the first time, we found that both tongue images and the tongue coating microbiome can be used as tools for the diagnosis of GC, and the area under the curve (AUC) value of the tongue image-based diagnostic model was 0.89. The AUC values of the tongue coating microbiome-based model reached 0.94 using genus data and 0.95 using species data. The results of the prospective multicentre clinical study showed that the AUC values of the three tongue image-based models for GCs reached 0.88-0.92 in the internal verification and 0.83-0.88 in the independent external verification, which were significantly superior to the combination of eight blood biomarkers. Interpretation: Our results suggest that tongue images can be used as a stable method for GC diagnosis and are significantly superior to conventional blood biomarkers. The three kinds of tongue image-based AI deep learning diagnostic models that we developed can be used to adequately distinguish patients with GC from participants with NGC, even early GC and precancerous lesions, such as atrophic gastritis (AG). Funding: The National Key R&D Program of China (2021YFA0910100), Program of Zhejiang Provincial TCM Sci-tech Plan (2018ZY006), Medical Science and Technology Project of Zhejiang Province (2022KY114, WKJ-ZJ-2104), Zhejiang Provincial Research Center for Upper Gastrointestinal Tract Cancer (JBZX-202006), Natural Science Foundation of Zhejiang Province (HDMY22H160008), Science and Technology Projects of Zhejiang Province (2019C03049), National Natural Science Foundation of China (82074245, 81973634, 82204828), and Chinese Postdoctoral Science Foundation (2022M713203).

Efficacy and safety of PLDR-IMRT for the re-irradiation of recurrent NPC: A prospective, single-arm, multicenter trial.

Salvage treatment of locoregionally recurrent nasopharyngeal carcinoma (NPC) requires weighing the benefits of re-irradiation against increased risks of toxicity. Here, we evaluated the outcomes of patients treated with intensity-modulated-based pulsed low-dose-rate radiotherapy (PLDR-IMRT) to enhance the curative effect of salvage treatment and reduce RT-related SAEs. A prospective clinical trial was conducted from March 2018 to March 2020 at multiple institutions. NPC patients who experienced relapse after radical therapy were re-irradiated with a median dose of 60 Gy (50.4-70 Gy)/30 f (28-35 f) using PLDR-IMRT. Thirty-six NPC patients who underwent PLDR-IMRT for locoregional recurrence were identified. With a median follow-up of 26.2 months, the objective response rate (ORR) of the entire cohort was 91.6%. The estimated mPFS duration was 28 months (95% CI: 24.9-31.1), and the estimated mLRFS duration was 30.4 months (95% CI: 25.2-35.5). The overall survival (OS) rate for all patients was 80.6%, the progression-free survival (PFS) rate was 75% and the cancer-specific survival (CSS) rate was 88.9% at 1 year. The LRFS and DMFS rates were 88.9% and 91.7%, respectively, at 1 year. A combination of systematic therapies could provide survival benefits to patients who experience NPC relapse (p < 0.05), and a Karnofsky performance status (KPS) score of ≥90 was a favorable factor for local control (p < 0.05). The incidence of acute SAEs (grade 3+) from PLDR was 22.2%, and the incidence of chronic SAEs was 19.4% among all patients. PLDR-IMRT combined with systematic therapy can effectively treat patients with locoregionally recurrent nasopharyngeal carcinoma and causes fewer adverse events than the rates expected with IMRT.

Segmentation of multiple Organs-at-Risk associated with brain tumors based on coarse-to-fine stratified networks.

BACKGROUND: Delineation of Organs-at-Risks (OARs) is an important step in radiotherapy treatment planning. As manual delineation is time-consuming, labor-intensive and affected by inter- and intra-observer variability, a robust and efficient automatic segmentation algorithm is highly desirable for improving the efficiency and repeatability of OAR delineation. PURPOSE: Automatic segmentation of OARs in medical images is challenged by low contrast, various shapes and imbalanced sizes of different organs. We aim to overcome these challenges and develop a high-performance method for automatic segmentation of 10 OARs required in radiotherapy planning for brain tumors. METHODS: A novel two-stage segmentation framework is proposed, where a coarse and simultaneous localization of all the target organs is obtained in the first stage, and a fine segmentation is achieved for each organ, respectively, in the second stage. To deal with organs with various sizes and shapes, a stratified segmentation strategy is proposed, where a High- and Low-Resolution Residual Network (HLRNet) that consists of a multiresolution branch and a high-resolution branch is introduced to segment medium-sized organs, and a High-Resolution Residual Network (HRRNet) is used to segment small organs. In addition, a label fusion strategy is proposed to better deal with symmetric pairs of organs like the left and right cochleas and lacrimal glands. RESULTS: Our method was validated on the dataset of MICCAI ABCs 2020 challenge for OAR segmentation. It obtained an average Dice of 75.8% for 10 OARs, and significantly outperformed several state-of-the-art models including nnU-Net (71.6%) and FocusNet (72.4%). Our proposed HLRNet and HRRNet improved the segmentation accuracy for medium-sized and small organs, respectively. The label fusion strategy led to higher accuracy for symmetric pairs of organs. CONCLUSIONS: Our proposed method is effective for the segmentation of OARs of brain tumors, with a better performance than existing methods, especially on medium-sized and small organs. It has a potential for improving the efficiency of radiotherapy planning with high segmentation accuracy.

Deep learning-based accurate delineation of primary gross tumor volume of nasopharyngeal carcinoma on heterogeneous magnetic resonance imaging: A large-scale and multi-center study.

BACKGROUND AND PURPOSE: The problem of obtaining accurate primary gross tumor volume (GTVp) segmentation for nasopharyngeal carcinoma (NPC) on heterogeneous magnetic resonance imaging (MRI) images with deep learning remains unsolved. Herein, we reported a new deep-learning method than can accurately delineate GTVp for NPC on multi-center MRI scans. MATERIAL AND METHODS: We collected 1057 patients with MRI images from five hospitals and randomly selected 600 patients from three hospitals to constitute a mixed training cohort for model development. The resting patients were used as internal (n = 259) and external (n = 198) testing cohorts for model evaluation. An augmentation-invariant strategy was proposed to delineate GTVp from multi-center MRI images, which encouraged networks to produce similar predictions for inputs with different augmentations to learn invariant anatomical structure features. The Dice similarity coefficient (DSC), 95 % Hausdorff distance (HD95), average surface distance (ASD), and relative absolute volume difference (RAVD) were used to measure segmentation performance. RESULTS: The model-generated predictions had a high overlap ratio with the ground truth. For the internal testing cohorts, the average DSC, HD95, ASD, and RAVD were 0.88, 4.99 mm, 1.03 mm, and 0.13, respectively. For external testing cohorts, the average DSC, HD95, ASD, and RAVD were 0.88, 3.97 mm, 0.97 mm, and 0.10, respectively. No significant differences were found in DSC, HD95, and ASD for patients with different T categories, MRI thickness, or in-plane spacings. Moreover, the proposed augmentation-invariant strategy outperformed the widely-used nnUNet, which uses conventional data augmentation approaches. CONCLUSION: Our proposed method showed a highly accurate GTVp segmentation for NPC on multi-center MRI images, suggesting that it has the potential to act as a generalized delineation solution for heterogeneous MRI images.

Lymph drainage and cervical fascia anatomy-oriented differential nodal CTV delineation at the supraclavicular region for esophageal cancer and nasopharyngeal cancer.

PURPOSE: To determine the differences in supraclavicular lymph node metastasis between esophageal cancer (EC) and nasopharyngeal cancer (NPC) and explore the feasibility of differential supraclavicular clinical target volume (CTV) contouring between these two diseases based on the involvement of different fascial spaces. MATERIALS AND METHODS: One hundred patients with supraclavicular nodes positive for EC or NPC were enrolled, and their pre-treatment images were reviewed. The distribution patterns of nodes between the two diseases were compared in the context of node levels defined by the 2017 Japanese Esophageal Society and 2013 International Consensus on Cervical Lymph Node Level Classification. Grouping supraclavicular nodes based on sub-compartments formed by the cervical fascia was discussed, and the feasibility of differential CTV contouring based on the differences in the involvement of these sub-compartments between EC and NPC was explored. RESULTS: The 2013 Consensus on cervical node levels and 2017 Japanese Esophageal Society node station could not practically guide supraclavicular CTV contouring. We divided the supraclavicular space into six sub-compartments: the para-esophageal space (PES), carotid sheath space (CSS), sub-thyroid pre-trachea space (STPTS), pre-vascular space (PVS), and vascular lateral space (VLS) I and II. EC mainly spread to the PES, STPTS, CSS, and VLS I, whereas NPC tended to spread to the CSS, VLS I, and VLS II. These combinations of sub-compartments may help constitute the supraclavicular CTVs for EC and NPC. CONCLUSIONS: The fascia anatomy-based sub-compartments sufficiently distinguished metastasis to the supraclavicular space between EC and NPC, thus facilitating differential CTV contouring between these two diseases.

HMRNet: High and Multi-Resolution Network With Bidirectional Feature Calibration for Brain Structure Segmentation in Radiotherapy.

Accurate segmentation of Anatomical brain Barriers to Cancer spread (ABCs) plays an important role for automatic delineation of Clinical Target Volume (CTV) of brain tumors in radiotherapy. Despite that variants of U-Net are state-of-the-art segmentation models, they have limited performance when dealing with ABCs structures with various shapes and sizes, especially thin structures (e.g., the falx cerebri) that span only few slices. To deal with this problem, we propose a High and Multi-Resolution Network (HMRNet) that consists of a multi-scale feature learning branch and a high-resolution branch, which can maintain the high-resolution contextual information and extract more robust representations of anatomical structures with various scales. We further design a Bidirectional Feature Calibration (BFC) block to enable the two branches to generate spatial attention maps for mutual feature calibration. Considering the different sizes and positions of ABCs structures, our network was applied after a rough localization of each structure to obtain fine segmentation results. Experiments on the MICCAI 2020 ABCs challenge dataset showed that: 1) Our proposed two-stage segmentation strategy largely outperformed methods segmenting all the structures in just one stage; 2) The proposed HMRNet with two branches can maintain high-resolution representations and is effective to improve the performance on thin structures; 3) The proposed BFC block outperformed existing attention methods using monodirectional feature calibration. Our method won the second place of ABCs 2020 challenge and has a potential for more accurate and reasonable delineation of CTV of brain tumors.

Semi-supervised medical image segmentation via uncertainty rectified pyramid consistency.

Despite that Convolutional Neural Networks (CNNs) have achieved promising performance in many medical image segmentation tasks, they rely on a large set of labeled images for training, which is expensive and time-consuming to acquire. Semi-supervised learning has shown the potential to alleviate this challenge by learning from a large set of unlabeled images and limited labeled samples. In this work, we present a simple yet efficient consistency regularization approach for semi-supervised medical image segmentation, called Uncertainty Rectified Pyramid Consistency (URPC). Inspired by the pyramid feature network, we chose a pyramid-prediction network that obtains a set of segmentation predictions at different scales. For semi-supervised learning, URPC learns from unlabeled data by minimizing the discrepancy between each of the pyramid predictions and their average. We further present multi-scale uncertainty rectification to boost the pyramid consistency regularization, where the rectification seeks to temper the consistency loss at outlier pixels that may have substantially different predictions than the average, potentially due to upsampling errors or lack of enough labeled data. Experiments on two public datasets and an in-house clinical dataset showed that: 1) URPC can achieve large performance improvement by utilizing unlabeled data and 2) Compared with five existing semi-supervised methods, URPC achieved better or comparable results with a simpler pipeline. Furthermore, we build a semi-supervised medical image segmentation codebase to boost research on this topic: https://github.com/HiLab-git/SSL4MIS.

Level Ib CTV delineation in nasopharyngeal carcinoma based on lymph node distribution and topographic anatomy.

BACKGROUND AND PURPOSE: To analyze the distribution pattern of lymph nodes (LNs) metastasis of level Ib in nasopharyngeal cancer (NPC) and propose shrinkage of clinical target volume (CTV) boundaries to avoid unnecessary radiation for some space with very low-risk of involvement. MATERIALS AND METHODS: Pretreatment images of pathologically proven NPC patients were reviewed and those with positive level Ib LN metastasis was enrolled. The geometric center of each level Ib LN in the neck was marked on a template CT. The spatial relationship of nodes with key structures in level Ib was analyzed. Modified level Ib CTV according to the 2013 International CTV consensus was proposed based on the LN distribution pattern. A PlanIQ Feasibility DVH module was implemented to evaluate the feasibility analysis of the best possible sparing of organs at risk (OAR) with modified Ib CTV. RESULTS: A total of 1518 NPC patients were reviewed and 54 with positive level Ib nodes were enrolled. Four sub-level anatomical regions were defined within the gross area of level Ib. Of 106 positive nodes identified, none, one, 88, and 17 were found in the intraglandular (IG), medial mandibular (MM), supra perivascular (SP), and infra perivascular (IP) sub-level, respectively. This study proposes sparing the IG and MM sub-level and including the area within a specified distance from the submandibular gland (11 mm for SP, 17 mm for IP) for CTV coverage. Compared with planning based on CTV-consensus, planning based on CTV-proposed results in a significantly reduced CTV volume, and mean dose (Dmean) of both the ipsilateral SMG and bilateral SLG. CONCLUSIONS: Based on detailed analysis of the relationship between positive node distribution and several important anatomical structures, modified level Ib CTV for prophylactic irradiation was proposed to reduce the dose of OAR irradiation.

A Self-Supporting, Conductor-Exposing, Stretchable, Ultrathin, and Recyclable Kirigami-Structured Liquid Metal Paper for Multifunctional E-Skin.

Electronic skin (E-skin) is a crucial seamless human-machine interface (HMI), holding promise in healthcare monitoring and personal electronics. Liquid metal (LM) has been recognized as an ideal electrode material to fabricate E-skins. However, conventional sealed LM electrodes cannot expose the LM layer for direct contact with the skin resulting in the low performance of electrophysiological monitoring. Furthermore, traditional printed LM electrodes are difficult to transfer or recycle, and fractures easily occur under stretching of the substrate. Here, we report a kind of LM electrode that we call a kirigami-structured LM paper (KLP), which is self-supporting, conductor-exposing, stretchable, ultrathin, and recyclable for multifunctional E-skin. The KLP is fabricated by the kirigami paper cutting art with three types of structures including uniaxial, biaxial, and square spiral. The KLP can act as an E-skin to acquire high-quality electrophysiological signals, such as electroencephalogram (EEG), electrocardiogram (ECG), and electromyogram (EMG). Upon integration with a triboelectric nanogenerator (TENG), the KLP can also operate as a self-powered E-skin. On the basis of the self-powered E-skin, we further developed a smart dialing communication system, which is applied on human skin to call a cellphone. Compared with conventional sealed or printed LM electrodes, the KLP can simultaneously achieve self-supporting, conductor-exposing, stretchable, ultrathin, and recyclable features. Such KLP offers potential for E-skins in healthcare monitoring and intelligent control, as well as smart robots, virtual reality, on-skin personal electronics, etc.

Automatic segmentation of organs-at-risks of nasopharynx cancer and lung cancer by cross-layer attention fusion network with TELD-Loss.

PURPOSE: Radiotherapy is one of the main treatments of nasopharyngeal cancer (NPC) and lung cancer. Accurate segmentation of organs at risks (OARs) in CT images is a key step in radiotherapy planning for NPC and lung cancer. However, the segmentation of OARs is influenced by the highly imbalanced size of organs, which often results in very poor segmentation results for small and difficult-to-segment organs. In addition, the complex morphological changes and fuzzy boundaries of OARs also pose great challenges to the segmentation task. In this paper, we propose a cross-layer attention fusion network (CLAF-CNN) to solve the problem of accurately segmenting OARs. METHODS: In CLAF-CNN, we integrate the spatial attention maps of the adjacent spatial attention modules to make the segmentation targets more accurately focused, so that the network can capture more target-related features. In this way, the spatial attention modules in the network can be learned and optimized together. In addition, we introduce a new Top-K exponential logarithmic Dice loss (TELD-Loss) to solve the imbalance problem in OAR segmentation. The TELD-Loss further introduces a Top-K optimization mechanism based on Dice loss and exponential logarithmic loss, which makes the network pay more attention to small organs and difficult-to-segment organs, so as to enhance the overall performance of the segmentation model. RESULTS: We validated our framework on the OAR segmentation datasets of the head and neck and lung CT images in the StructSeg 2019 challenge. Experiments show that the CLAF-CNN outperforms the state-of-the-art attention-based segmentation methods in the OAR segmentation task with average Dice coefficient of 79.65% for head and neck OARs and 88.39% for lung OARs. CONCLUSIONS: This work provides a new network named CLAF-CNN which contains cross-layer spatial attention map fusion architecture and TELD-Loss for OAR segmentation. Results demonstrated that the proposed method could obtain accurate segmentation results for OARs, which has a potential of improving the efficiency of radiotherapy planning for nasopharynx cancer and lung cancer.