High-entropy alloys in electrocatalysis: from fundamentals to applications.

High-entropy alloys (HEAs) comprising five or more elements in near-equiatomic proportions have attracted ever increasing attention for their distinctive properties, such as exceptional strength, corrosion resistance, high hardness, and excellent ductility. The presence of multiple adjacent elements in HEAs provides unique opportunities for novel and adaptable active sites. By carefully selecting the element configuration and composition, these active sites can be optimized for specific purposes. Recently, HEAs have been shown to exhibit remarkable performance in electrocatalytic reactions. Further activity improvement of HEAs is necessary to determine their active sites, investigate the interactions between constituent elements, and understand the reaction mechanisms. Accordingly, a comprehensive review is imperative to capture the advancements in this burgeoning field. In this review, we provide a detailed account of the recent advances in synthetic methods, design principles, and characterization technologies for HEA-based electrocatalysts. Moreover, we discuss the diverse applications of HEAs in electrocatalytic energy conversion reactions, including the hydrogen evolution reaction, hydrogen oxidation reaction, oxygen reduction reaction, oxygen evolution reaction, carbon dioxide reduction reaction, nitrogen reduction reaction, and alcohol oxidation reaction. By comprehensively covering these topics, we aim to elucidate the intricacies of active sites, constituent element interactions, and reaction mechanisms associated with HEAs. Finally, we underscore the imminent challenges and emphasize the significance of both experimental and theoretical perspectives, as well as the potential applications of HEAs in catalysis. We anticipate that this review will encourage further exploration and development of HEAs in electrochemistry-related applications.

Interface Metal Oxides Regulating Electronic State around Nickel Species for Efficient Alkaline Hydrogen Electrocatalysis.

Heterogeneous electrocatalysis typically depends on the surface electronic states of active sites. Modulating the surface charge state of an electrocatalysts can be employed to improve performance. Among all the investigated materials, nickel (Ni)-based catalysts are the only non-noble-metal-based alternatives for both hydrogen oxidation and evolution reactions (HOR and HER) in alkaline electrolyte, while their activities should be further improved because of the unfavorable hydrogen adsorption behavior. Hereto, Ni with exceptional HOR electrocatalytic performance by changing the d-band center by metal oxides interface coupling formed in situ is endowed. The resultant MoO2 coupled Ni heterostructures exhibit an apparent HOR activity, even approaching to that of commercial 20% Pt/C benchmark, but with better long-term stability in alkaline electrolyte. An exceptional HER performance is also achieved by the Ni-MoO2 heterostructures. The experiment results are rationalized by the theoretical calculations, which indicate that coupling MoO2 with Ni results in the downshift of d-band center of Ni, and thus weakens hydrogen adsorption and benefits for hydroxyl adsorption. This concept is further proved by other metal oxides (e.g., CeO2 , V2 O3 , WO3 , Cr2 O3 )-formed Ni-based heterostructures to engineer efficient hydrogen electrocatalysts.

Rational Synthesis of Core-Shell-Structured Nickel Sulfide-Based Nanostructures for Efficient Seawater Electrolysis.

Versatile electrocatalysis at higher current densities for natural seawater splitting to produce hydrogen demands active and robust catalysts to overcome the severe chloride corrosion, competing chlorine evolution, and catalyst poisoning. Hereto, the core-shell-structured heterostructures composed of amorphous NiFe hydroxide layer capped Ni3 S2 nanopyramids which are directly grown on nickel foam skeleton (NiS@LDH/NF) are rationally prepared to regulate cooperatively electronic structure and mass transport for boosting oxygen evolution reaction (OER) performance at larger current densities. The prepared NiS@LDH/NF delivers the anodic current density of 1000 mA cm-2 at the overpotential of 341 mV in 1.0 m KOH seawater. The feasible surface reconstruction of Ni3 S2 -FeNi LDH interfaces improves the chemical stability and corrosion resistance, ensuring the robust electrocatalytic activity in seawater electrolytes for continuous and stable oxygen evolution without any hypochlorite production. Meanwhile, the designed Ni3 S2 nanopyramids coated with FeNi2 P layer (NiS@FeNiP/NF) still exhibit the improved hydrogen evolution reaction (HER) activity in 1.0 m KOH seawater. Furthermore, the NiS@FeNiP/NF||NiS@LDH/NF pair requires cell voltage of 1.636 V to attain 100 mA cm-2 with a 100% Faradaic efficiency, exhibiting tremendous potential for hydrogen production from seawater.

Target coverage and local recurrences after radiotherapy for sinonasal cancer in Denmark 2008-2015. A DAHANCA study.

PURPOSE: The study aimed to investigate the pattern of failure and describe compromises in the definition and coverage of the target for patients treated with curatively intended radiotherapy (RT) for sinonasal cancer (SNC). METHODS AND MATERIAL: Patients treated with curatively intended RT in 2008-2015 in Denmark for SNC were eligible for the retrospective cohort study. Information regarding diagnosis and treatment was retrieved from the national database of the Danish Head and Neck Cancer Group (DAHANCA). Imaging from the diagnosis of recurrences was collected, and the point of origin (PO) of the recurrent tumour was estimated. All treatment plans were collected and reviewed with the focus on target coverage, manual modifications of target volumes, and dose to organs at risk (OARs) above defined constraints. RESULTS: A total of 184 patients were included in the analysis, and 76 (41%) relapsed. The majority of recurrences involved T-site (76%). Recurrence imaging of 39 patients was evaluated, and PO was established. Twenty-nine POs (74%) were located within the CTV, and the minimum dose to the PO was median 64.1 Gy (3.1-70.7). The criteria for target coverage (V95%) was not met in 89/184 (48%) of the CTV and 131/184 (71%) of the PTV. A total of 24% of CTVs had been manually modified to spare OARs of high-dose irradiation. No difference in target volume modifications was observed between patients who suffered recurrence and patients with lasting remission. CONCLUSION: The majority of relapses after radical treatment of SNC were located in the T-site (the primary tumour site). Multiple compromises with regards to target coverage and tolerance levels for OARs in the sinonasal region, as defined from RT guidelines, were taken. No common practice in this respect could be derived from the study.

Comparing different CT, PET and MRI multi-modality image combinations for deep learning-based head and neck tumor segmentation.

BACKGROUND: Manual delineation of gross tumor volume (GTV) is essential for radiotherapy treatment planning, but it is time-consuming and suffers inter-observer variability (IOV). In clinics, CT, PET, and MRI are used to inform delineation accuracy due to their different complementary characteristics. This study aimed to investigate deep learning to assist GTV delineation in head and neck squamous cell carcinoma (HNSCC) by comparing various modality combinations. MATERIALS AND METHODS: This retrospective study had 153 patients with multiple sites of HNSCC including their planning CT, PET, and MRI (T1-weighted and T2-weighted). Clinical delineations of gross tumor volume (GTV-T) and involved lymph nodes (GTV-N) were collected as the ground truth. The dataset was randomly divided into 92 patients for training, 31 for validation, and 30 for testing. We applied a residual 3 D UNet as the deep learning architecture. We independently trained the UNet with four different modality combinations (CT-PET-MRI, CT-MRI, CT-PET, and PET-MRI). Additionally, analogical to post-processing, an average fusion of three bi-modality combinations (CT-PET, CT-MRI, and PET-MRI) was produced as an ensemble. Segmentation accuracy was evaluated on the test set, using Dice similarity coefficient (Dice), Hausdorff Distance 95 percentile (HD95), and Mean Surface Distance (MSD). RESULTS: All imaging combinations including PET provided similar average scores in range of Dice: 0.72-0.74, HD95: 8.8-9.5 mm, MSD: 2.6-2.8 mm. Only CT-MRI had a lower score with Dice: 0.58, HD95: 12.9 mm, MSD: 3.7 mm. The average of three bi-modality combinations reached Dice: 0.74, HD95: 7.9 mm, MSD: 2.4 mm. CONCLUSION: Multimodal deep learning-based auto segmentation of HNSCC GTV was demonstrated and inclusion of the PET image was shown to be crucial. Training on combined MRI, PET, and CT data provided limited improvements over CT-PET and PET-MRI. However, when combining three bimodal trained networks into an ensemble, promising improvements were shown.

Engineering the Core-Shell-Structured NCNTs-Ni2Si@Porous Si Composite with Robust Ni-Si Interfacial Bonding for High-Performance Li-Ion Batteries.

A new strategy has been innovatively proposed for wrapping the Ni-incorporated and N-doped carbon nanotube arrays (Ni-NCNTs) on porous Si with robust Ni-Si interfacial bonding to form the core-shell-structured NCNTs-Ni2Si@Si. The hierarchical porous silicon core was first fabricated via a novel self-templating synthesis route based on two crucial strategies: in situ thermal evaporation of crystal water from the perlite for producing porous SiO2 and subsequent magnesiothermic reduction of porous SiO2 into porous Si. Ni-NCNTs were subsequently constructed based on the Ni-catalyzed tip-growth mechanism and were further engineered to fully wrap the porous Si microparticles by forming the Ni2Si alloy at the heterojunction interface. When the prepared NCNTs-Ni2Si@Si was evaluated as the anode material for Li-ion batteries, the hierarchical porous system in the Si core and the rich void spaces in carbon nanotube arrays contributed to the remarkable accommodation of volume expansion of Si as well as the significant increase of Li+ diffusion and Si utilization. Moreover, the Ni2Si alloy, which chemically linked the Ni-NCNTs and porous Si, not only provided good electronic contact between the Si core and carbon shell but also effectively prevented the CNTs' detachment from the Si core during cycling. The multifunctional structural design rendered the whole electrode highly stable and active in Li storage, and the electrochemically active NCNTs-Ni2Si@Si electrode delivered a high reversible capacity of 1547 mAh g-1 and excellent cycling stability (85% capacity retention after 600 discharge-charge cycles) at a current density of 358 mA g-1 (0.1 C) as well as good rate performance (778 mAh g-1 at 2 C), showing great potential as an efficient and stable anode for high energy density Li-ion batteries.

The RRM domain of human fused in sarcoma protein reveals a non-canonical nucleic acid binding site.

Fused in sarcoma (FUS) is involved in many processes of RNA metabolism. FUS and another RNA binding protein, TDP-43, are implicated in amyotrophic lateral sclerosis (ALS). It is significant to characterize the RNA recognition motif (RRM) of FUS as its nucleic acid binding properties are unclear. More importantly, abolishing the RNA binding ability of the RRM domain of TDP43 was reported to suppress the neurotoxicity of TDP-43 in Drosophila. The sequence of FUS-RRM varies significantly from canonical RRMs, but the solution structure of FUS-RRM determined by NMR showed a similar overall folding as other RRMs. We found that FUS-RRM directly bound to RNA and DNA and the binding affinity was in the micromolar range as measured by surface plasmon resonance and NMR titration. The nucleic acid binding pocket in FUS-RRM is significantly distorted since several critical aromatic residues are missing. An exceptionally positively charged loop in FUS-RRM, which is not found in other RRMs, is directly involved in the RNA/DNA binding. Substituting the lysine residues in the unique KK loop impaired the nucleic acid binding and altered FUS subcellular localization. The results provide insights into the nucleic acid binding properties of FUS-RRM and its potential relevance to ALS.

[Microwave ablation of the canine prostate: an experimental study].

OBJECTIVE: To investigate the effect of transrectal ultrasound-guided microwave ablation of canine prostate tissue. METHODS: Guided by transrectal ultrasound, we conducted microwave ablation on each side of the prostate in 12 male dogs, 6 at 40 W/ 120 s (group A) and the other 6 at 40 W/160 s (group B), and observed the changes in the thermal lesions using grayscale ultrasound. After thermal ablation, we measured the volume of the thermal lesions by contrast-enhanced ultrasound (CEUS). Then we harvested the whole prostate from the animals and determined the lesion volumes in the fresh tissue specimens. RESULTS: Grayscale ultrasound revealed an echogenic area at the initiation of the microwave ablation procedure, which was enlarged with the increase of ablation time. At the end of the procedure, the lesions appeared as an irregular heterogeneous echogenic area. CEUS showed oval non-perfused areas, which appeared as well-defined non-echoic areas in sharp contrast with the surrounding normal prostate parenchyma with bolus injection of contrast material (Sonovue, 2.4 ml), and that the thermal lesion volumes of groups A and B were (1.18 +/- 0.23) cm3 and (1.52 +/- 0.23) cm3, respectively. The thermal lesions of the gross specimen exhibited an elliptical shape, pale color and clear margin, and their volumes were (1.13 +/- 0.20) cm3 and (1.48 +/- 0.20) cm3, respectively, in groups A and B. CONCLUSION: Different combinations of time and power can produce coagulative necrotic lesions of different volumes in the local prostatic tissue. CEUS can accurately manifest the lesion area and thus avoid excessive or inadequate ablation treatment.

Two-dimensional architecture of N,S-codoped nanocarbon composites embedding few-layer MoS2 for efficient lithium storage.

The exploration and advancement of highly efficient anode materials for lithium-ion batteries (LIBs) are critical to meet the growing demands of the energy storage market. In this study, we present an easily scalable synthesis method for the one-pot formation of few-layer MoS2 nanosheets on a N,S dual-doped carbon monolith with a two-dimensional (2D) architecture, termed MoS2/NSCS. Systematic electrochemical measurements demonstrate that MoS2/NSCS, when employed as the anode material in LIBs, exhibits a high capacity of 681 mA h g-1 at 0.2 A g-1 even after 110 cycles. The exceptional electrochemical performance of MoS2/NSCS can be attributed to its unique porous 2D architecture. The few-layer MoS2 sheets with a large interlayer distance reduce ion diffusion pathways and enhance ion mobility rates. Additionally, the N,S-doped porous carbon matrix not only preserves structural integrity but also facilitates electronic conductivity. These combined factors contribute to the reversible electrochemical activities observed in MoS2/NSCS, highlighting its potential as a promising anode material for high-performance LIBs.

Postoperative recurrence of diffuse sclerosing thyroid cancer in an adolescent patient: A case report.

BACKGROUND: Papillary thyroid cancer is an inert malignant tumor with a good response to surgical treatment, low recurrence and metastasis rate, and good prognosis. Diffuse sclerosing thyroid cancer is an invasive subtype that is more common in young people, with a higher rate of lymph node metastasis and recurrence, and a relatively poor prognosis. PATIENT CONCERNS: A 13-year-old girl underwent radical surgery for diffuse sclerosing thyroid cancer. Eight years later, due to a large number of lymph node metastases, she underwent another radical surgery on her neck lymph nodes. METHODS: The patient thyroid ultrasound and neck enhanced CT indicated that the patient had multiple enlarged lymph nodes in the neck with irregular morphology and structure, and the possibility of metastatic lymph nodes was high. Subsequently, the patient underwent thyroid fine-needle aspiration and the results showed that cancer cells were detected in both cervical lymph nodes. DIAGNOSIS: The patient was diagnosed with bilateral cervical lymph node metastases after thyroid surgery. RESULTS: After the second surgery, the patient recovered well, and no residual or focal iodine uptake tissue was found on the enhanced CT examination. CONCLUSION: As diffuse sclerosing thyroid cancer is prone to lymph node and recurrent metastases, once it is diagnosed, radical treatment should be actively performed. Postoperative adjuvant radiation therapy should be administered according to the patient condition and regular follow-ups should be conducted to monitor neck lymph node metastasis.

Taking Advantage of Potential Coincidence Region: Insights into Gas Production Behavior in Advanced Self-Activated Hydrazine-Assisted Alkaline Seawater Electrolysis.

Water electrolysis assisted by hydrazine has emerged as a prospective energy conversion method for achieving efficient hydrogen generation. Due to the potential coincidence region (PCR) between the hydrogen evolution reaction (HER) and the electro-oxidation of hydrazine, the hydrazine oxidation reaction (HzOR) offers distinct advantages in terms of strategy amalgamation, device architecture, and the broadening of application horizons. Herein, we report a bifunctional electrocatalyst of interfacial heterogeneous Fe2P/Co2P microspheres supported on Ni foam (FeCoP/NF). Benefiting from the strong interfacial coupling effect between Fe2P and Co2P and the three-dimensional microsphere structure, FeCoP/NF exhibits outstanding bifunctional electrocatalytic performance, achieving 10 mA cm-2 with low overpotentials of 10 and 203 mV for HER and HzOR, respectively. Utilizing FeCoP/NF for both electrodes in HzOR-assisted water electrolysis results in significantly reduced potentials of 820 mV for 1 A cm-2 in contrast to the electro-oxidation of alternative chemical substrates. The presence of a potential coincidence region makes the application of self-activated seawater electrolysis realistic. The gas production behavior at different current densities in this interesting hydrogen production system is discussed, and some rules that are distinguished from conventional water electrolysis are summarized. Furthermore, a new self-powered hydrogen production system with a direct hydrazine fuel cell, rechargeable Zn-hydrazine battery, and hydrazine-assisted seawater electrolysis is proposed, emphasizing the distinct benefits of HzOR and its potential role in electrochemical energy conversion technologies powered by renewable sources.

RadDeploy: A framework for integrating in-house developed software and artificial intelligence models seamlessly into radiotherapy workflows.

The use of and research in automation and artificial intelligence (AI) in radiotherapy is moving with incredible pace. Many innovations do, however, not make it into the clinic. One technical reason for this may be the lack of a platform to deploy such software into clinical practice. We suggest RadDeploy as a framework for integrating containerized software in clinical workflows outside of treatment planning systems. RadDeploy supports multiple DICOM as input for model containers and can run model containers asynchronously across GPUs and computers. This technical note summarizes the inner workings of RadDeploy and demonstrates three use-cases with varying complexity.

The effect of editing clinical contours on deep-learning segmentation accuracy of the gross tumor volume in glioblastoma.

Background and purpose: Deep-learning (DL) models for segmentation of the gross tumor volume (GTV) in radiotherapy are generally based on clinical delineations which suffer from inter-observer variability. The aim of this study was to compare performance of a DL-model based on clinical glioblastoma GTVs to a model based on a single-observer edited version of the same GTVs. Materials and methods: The dataset included imaging data (Computed Tomography (CT), T1, contrast-T1 (T1C), and fluid-attenuated-inversion-recovery (FLAIR)) of 259 glioblastoma patients treated with post-operative radiotherapy between 2012 and 2019 at a single institute. The clinical GTVs were edited using all imaging data. The dataset was split into 207 cases for training/validation and 52 for testing.GTV segmentation models (nnUNet) were trained on clinical and edited GTVs separately and compared using Surface Dice with 1 mm tolerance (sDSC1mm). We also evaluated model performance with respect to extent of resection (EOR), and different imaging combinations (T1C/T1/FLAIR/CT, T1C/FLAIR/CT, T1C/FLAIR, T1C/CT, T1C/T1, T1C). A Wilcoxon test was used for significance testing. Results: The median (range) sDSC1mm of the clinical-GTV-model and edited-GTV-model both evaluated with the edited contours, was 0.76 (0.43-0.94) vs. 0.92 (0.60-0.98) respectively (p < 0.001). sDSC1mm was not significantly different between patients with a biopsy, partial, and complete resection. T1C as single input performed as good as use of imaging combinations. Conclusions: High segmentation accuracy was obtained by the DL-models. Editing of the clinical GTVs significantly increased DL performance with a relevant effect size. DL performance was robust for EOR and highly accurate using only T1C.

Enhancing the reliability of deep learning-based head and neck tumour segmentation using uncertainty estimation with multi-modal images.

Objective.Deep learning shows promise in autosegmentation of head and neck cancer (HNC) primary tumours (GTV-T) and nodal metastases (GTV-N). However, errors such as including non-tumour regions or missing nodal metastases still occur. Conventional methods often make overconfident predictions, compromising reliability. Incorporating uncertainty estimation, which provides calibrated confidence intervals can address this issue. Our aim was to investigate the efficacy of various uncertainty estimation methods in improving segmentation reliability. We evaluated their confidence levels in voxel predictions and ability to reveal potential segmentation errors.Approach.We retrospectively collected data from 567 HNC patients with diverse cancer sites and multi-modality images (CT, PET, T1-, and T2-weighted MRI) along with their clinical GTV-T/N delineations. Using the nnUNet 3D segmentation pipeline, we compared seven uncertainty estimation methods, evaluating them based on segmentation accuracy (Dice similarity coefficient, DSC), confidence calibration (Expected Calibration Error, ECE), and their ability to reveal segmentation errors (Uncertainty-Error overlap using DSC, UE-DSC).Main results.Evaluated on the hold-out test dataset (n= 97), the median DSC scores for GTV-T and GTV-N segmentation across all uncertainty estimation methods had a narrow range, from 0.73 to 0.76 and 0.78 to 0.80, respectively. In contrast, the median ECE exhibited a wider range, from 0.30 to 0.12 for GTV-T and 0.25 to 0.09 for GTV-N. Similarly, the median UE-DSC also ranged broadly, from 0.21 to 0.38 for GTV-T and 0.22 to 0.36 for GTV-N. A probabilistic network-PhiSeg method consistently demonstrated the best performance in terms of ECE and UE-DSC.Significance.Our study highlights the importance of uncertainty estimation in enhancing the reliability of deep learning for autosegmentation of HNC GTV. The results show that while segmentation accuracy can be similar across methods, their reliability, measured by calibration error and uncertainty-error overlap, varies significantly. Used with visualisation maps, these methods may effectively pinpoint uncertainties and potential errors at the voxel level.

Artificial Heterointerfaces with Regulated Charge Distribution of Ni Active Sites for Urea Oxidation Reaction.

In contrast to the thermodynamically unfavorable anodic oxygen evolution reaction, the electrocatalytic urea oxidation reaction (UOR) presents a more favorable thermodynamic potential. However, the practical application of UOR has been hindered by sluggish kinetics. In this study, hierarchical porous nanosheet arrays featuring abundant Ni-WO3 heterointerfaces on nickel foam (Ni-WO3/NF) is introduced as a monolith electrode, demonstrating exceptional activity and stability toward UOR. The Ni-WO3/NF catalyst exhibits unprecedentedly rapid UOR kinetics (200 mA cm-2 at 1.384 V vs. RHE) and a high turnover frequency (0.456 s-1), surpassing most previously reported Ni-based catalysts, with negligible activity decay observed during a durability test lasting 150 h. Ex situ X-ray photoelectron spectroscopy and density functional theory calculations elucidate that the WO3 interface significantly modulates the local charge distribution of Ni species, facilitating the generation of Ni3+ with optimal affinity for interacting with urea molecules and CO2 intermediates at heterointerfaces during UOR. This mechanism accelerates the interfacial electrocatalytic kinetics. Additionally, in situ Fourier transform infrared spectroscopy provides deep insights into the substantial contribution of interfacial Ni-WO3 sites to UOR electrocatalysis, unraveling the underlying molecular-level mechanisms. Finally, the study explores the application of a direct urea fuel cell to inspire future practical implementations.

Changes in α-Dicarbonyl Compound Contents during Storage of Various Fruits and Juices.

α-Dicarbonyl compounds (α-DCs) are commonly present in various foods. We conducted the investigation into concentration changes of α-DCs including 3-deoxyglucosone (3-DG), glyoxal (GO), and methylglyoxal (MGO) in fresh fruits and decapped commercial juices during storage at room temperature and 4 °C, as well as in homemade juices during storage at 4 °C. The studies indicate the presence of α-DCs in all samples. The initial contents of 3-DG in the commercial juices (6.74 to 65.61 µg/mL) are higher than those in the homemade ones (1.97 to 4.65 µg/mL) as well as fruits (1.58 to 3.33 µg/g). The initial concentrations of GO and MGO are normally less than 1 µg/mL in all samples. During storage, the α-DC levels in the fruits exhibit an initial increase followed by a subsequent decrease, whereas, in all juices, they tend to accumulate continuously over time. As expected, 4 °C storage reduces the increase rates of the α-DC concentrations in most samples. From the viewpoint of the α-DC contents, fruits and homemade juices should always be the first choice for daily intake of nutrients and commercial juices ought to be mostly avoided.

Artificial intelligence uncertainty quantification in radiotherapy applications - A scoping review.

BACKGROUND/PURPOSE: The use of artificial intelligence (AI) in radiotherapy (RT) is expanding rapidly. However, there exists a notable lack of clinician trust in AI models, underscoring the need for effective uncertainty quantification (UQ) methods. The purpose of this study was to scope existing literature related to UQ in RT, identify areas of improvement, and determine future directions. METHODS: We followed the PRISMA-ScR scoping review reporting guidelines. We utilized the population (human cancer patients), concept (utilization of AI UQ), context (radiotherapy applications) framework to structure our search and screening process. We conducted a systematic search spanning seven databases, supplemented by manual curation, up to January 2024. Our search yielded a total of 8980 articles for initial review. Manuscript screening and data extraction was performed in Covidence. Data extraction categories included general study characteristics, RT characteristics, AI characteristics, and UQ characteristics. RESULTS: We identified 56 articles published from 2015 to 2024. 10 domains of RT applications were represented; most studies evaluated auto-contouring (50 %), followed by image-synthesis (13 %), and multiple applications simultaneously (11 %). 12 disease sites were represented, with head and neck cancer being the most common disease site independent of application space (32 %). Imaging data was used in 91 % of studies, while only 13 % incorporated RT dose information. Most studies focused on failure detection as the main application of UQ (60 %), with Monte Carlo dropout being the most commonly implemented UQ method (32 %) followed by ensembling (16 %). 55 % of studies did not share code or datasets. CONCLUSION: Our review revealed a lack of diversity in UQ for RT applications beyond auto-contouring. Moreover, we identified a clear need to study additional UQ methods, such as conformal prediction. Our results may incentivize the development of guidelines for reporting and implementation of UQ in RT.

Artificial Intelligence Uncertainty Quantification in Radiotherapy Applications - A Scoping Review.

Background/purpose: The use of artificial intelligence (AI) in radiotherapy (RT) is expanding rapidly. However, there exists a notable lack of clinician trust in AI models, underscoring the need for effective uncertainty quantification (UQ) methods. The purpose of this study was to scope existing literature related to UQ in RT, identify areas of improvement, and determine future directions. Methods: We followed the PRISMA-ScR scoping review reporting guidelines. We utilized the population (human cancer patients), concept (utilization of AI UQ), context (radiotherapy applications) framework to structure our search and screening process. We conducted a systematic search spanning seven databases, supplemented by manual curation, up to January 2024. Our search yielded a total of 8980 articles for initial review. Manuscript screening and data extraction was performed in Covidence. Data extraction categories included general study characteristics, RT characteristics, AI characteristics, and UQ characteristics. Results: We identified 56 articles published from 2015-2024. 10 domains of RT applications were represented; most studies evaluated auto-contouring (50%), followed by image-synthesis (13%), and multiple applications simultaneously (11%). 12 disease sites were represented, with head and neck cancer being the most common disease site independent of application space (32%). Imaging data was used in 91% of studies, while only 13% incorporated RT dose information. Most studies focused on failure detection as the main application of UQ (60%), with Monte Carlo dropout being the most commonly implemented UQ method (32%) followed by ensembling (16%). 55% of studies did not share code or datasets. Conclusion: Our review revealed a lack of diversity in UQ for RT applications beyond auto-contouring. Moreover, there was a clear need to study additional UQ methods, such as conformal prediction. Our results may incentivize the development of guidelines for reporting and implementation of UQ in RT.

Towards interactive deep-learning for tumour segmentation in head and neck cancer radiotherapy.

Background and purpose: With deep-learning, gross tumour volume (GTV) auto-segmentation has substantially been improved, but still substantial manual corrections are needed. With interactive deep-learning (iDL), manual corrections can be used to update a deep-learning tool while delineating, minimising the input to achieve acceptable segmentations. We present an iDL tool for GTV segmentation that took annotated slices as input and simulated its performance on a head and neck cancer (HNC) dataset. Materials and methods: Multimodal image data of 204 HNC patients with clinical tumour and lymph node GTV delineations were used. A baseline convolutional neural network (CNN) was trained (n = 107 training, n = 22 validation) and tested (n = 24). Subsequently, user input was simulated on initial test set by replacing one or more of predicted slices with ground truth delineation, followed by re-training the CNN. The objective was to optimise re-training parameters and simulate slice selection scenarios while limiting annotations to maximally-five slices. The remaining 51 patients were used as an independent test set, where Dice similarity coefficient (DSC), mean surface distance (MSD), and 95% Hausdorff distance (HD95%) were assessed at baseline and after every update. Results: Median segmentation accuracy at baseline was DSC = 0.65, MSD = 4.3 mm, HD95% = 17.5 mm. Updating CNN using three slices equally sampled from the craniocaudal axis of the GTV in the first round, followed by two rounds of annotating one extra slice, gave the best results. The accuracy improved to DSC = 0.82, MSD = 1.6 mm, HD95% = 4.8 mm. Every CNN update took 30 s. Conclusions: The presented iDL tool achieved substantial segmentation improvement with only five annotated slices.

Mortality and excess life-years lost in patients with schizophrenia under community care: a 5-year follow-up cohort study.

OBJECTIVE: Mortality rate is a general indicator which can be used to measure care and management of schizophrenia. This cohort study evaluated the standardized mortality ratios (SMRs) of all-cause mortality and life-years lost (LYLs) in patients with schizophrenia under a community care program in China. METHODS: Data were obtained from the National Community Care Program System for Severe Mental Disorders. A total of 99,214 patients diagnosed with schizophrenia were enrolled before December 2014 and followed between 2015 and 2019. A total of 9,483 patients died. Crude mortality rates (CMRs) and SMRs were then stratified by natural vs. unnatural causes, and major groups of death were standardized according to the 2010 National Population SMRs. The corresponding LYLs at birth were also calculated by gender and age. RESULTS: The SMRs of patients with schizophrenia were significantly elevated during the study period, with an overall SMR of 4.98 (95%CI 2.67-7.32). Neoplasms, cardiovascular diseases, cerebrovascular diseases, external injuries, and poisonings were the most significant causes of death among patients with schizophrenia compared to the general population. The mean LYLs of patients with schizophrenia were 15.28 (95%CI 13.26-17.30). Males with schizophrenia lost 15.82 life-years (95%CI 13.48-18.16), and females lost 14.59 life-years (95%CI 13.12-16.06). CONCLUSIONS: Patients with schizophrenia under community care had a high mortality rate in our study, even though mental health services have been integrated into the general healthcare system in China to narrow treatment gaps in mental health for > 10 years. In terms of mortality outcome indicators, effective and quality mental health services still have a long way to go. The current study demonstrates the potential for improved prevention and treatment of individuals with schizophrenia under community care.