Huawei German Research Center

225Ac-labeled prostate-specific membrane antigen (PSMA) radiopharmaceutical therapy (RPT) has emerged as a promising treatment option in advanced metastatic castration-resistant prostate cancer. We aimed to identify predictors for outcome including established clinical and PSMA-directed imaging parameters at baseline. Methods: Twenty-six patients with metastatic castration-resistant prostate cancer who were receiving [225Ac]Ac-PSMA-I&T RPT were included in this retrospective monocentric study. In each patient, all metastases on pretherapeutic [18F]F-PSMA-1007 PET/CT were segmented, enabling assessment of averaged SUVmax, SUVmean, as well as summed PSMA tumor volume (PSMA-TV) and PSMA-total lesion quotient (PSMA-TLQ; defined as PSMA-TV divided by SUVmean). PET parameters were then correlated with the relative prostate-specific antigen (PSA) change after 2 cycles of [225Ac]Ac-PSMA-I&T RPT. In addition, the predictive values for early progressive disease (PD; defined as a PSA increase of more than 25% or PD according to RECIP 1.0) after 2 cycles, progression-free survival (PFS), and overall survival (OS) were explored. Results: SUVmax (r = -0.42, P = 0.031) and SUVmean (r = -0.4, P = 0.046) correlated significantly with PSA change after 2 cycles, but PSMA-TV (P = 0.51) and PSMA-TLQ (P = 0.83) did not. Eleven patients (42%) demonstrated early PD. SUVmax (odds ratio, 0.769; P = 0.032) and SUVmean (odds ratio, 0.427; P = 0.041) were predictive for early PD. SUVmean emerged as the strongest predictor for prolonged PFS (hazard ratio [HR], 0.501; P = 0.006) with SUVmax also trending toward significance (HR, 0.914; P = 0.054). Patients with a high SUVmean achieved a longer median PFS of 134 d compared with 39 d in patients with low SUVmean (HR, 0.369; P = 0.041). PSMA-TLQ (HR, 1.007; P = 0.027) and PSMA-TV (HR, 1.001; P = 0.037) were predictive for OS. Patients with a low PSMA-TLQ (reflecting low PSMA-TV in combination with high SUVmean) achieved a longer median OS of 375 d compared with 148 d in patients with high PSMA-TLQ (HR, 2.84; P = 0.043). Conclusion: In patients scheduled for [225Ac]Ac-PSMA-I&T RPT, SUVmean on pretherapeutic [18F]F-PSMA-1007 PET/CT is predictive for early treatment response by identifying individuals prone to early PD and shorter PFS. In addition, PSMA-TLQ is associated with OS. As such, pretherapeutic PSMA PET-based quantification may optimize patient selection for targeted α-RPT.

The generation of high-fidelity view synthesis is essential for robotic navigation and interaction but remains challenging, particularly in indoor environments and real-time scenarios. Existing techniques often require significant computational resources for both training and rendering, and they frequently result in suboptimal 3D representations due to insufficient geometric structuring. To address these limitations, we introduce VoxNeRF, a novel approach that utilizes easy-to-obtain geometry priors to enhance both the quality and efficiency of neural indoor reconstruction and novel view synthesis. We propose an efficient voxel-guided sampling technique that allocates computational resources selectively to the most relevant segments of rays based on a voxel-encoded geometry prior, significantly reducing training and rendering time. Additionally, we incorporate a robust depth loss to improve reconstruction and rendering quality in sparse view settings. Our approach is validated with extensive experiments on ScanNet and ScanNet++ where VoxNeRF outperforms existing state-of-the-art methods and establishes a new benchmark for indoor immersive interpolation and extrapolation settings.

Tumor asphericity in 18F-FDG PET is a prognostic marker that has been investigated in small pilot studies of patients with head and neck squamous cell carcinoma (HNSCC). Here, we investigated the prognostic role of asphericity in a large multicenter database of patients with HNSCC treated with primary radiotherapy or chemoradiation and assessed its independent prognostic value. Methods: In total, 1,104 patients were included in this analysis. All received pretreatment 18F-FDG PET scans. Clinical risk factors were evaluated, and quantitative PET parameters SUVmax, metabolic tumor volume (MTV), total lesion glycolysis, and asphericity were calculated. Primary study endpoints were overall survival (OS) and locoregional control (LRC). Uni- and multivariate Cox regression analyses were performed. Additionally, asphericity was combined with the best-established quantitative PET parameter of MTV, and the combinatory approach of using asphericity and MTV was compared with the use of only asphericity or MTV by bootstrap analyses. Results: Asphericity showed only a modest correlation with the established PET parameters of MTV, SUVmax, and total lesion glycolysis. On univariate testing asphericity was strongly associated with the outcome of patients (LRC and OS with P < 0.001). In multivariate testing of all imaging parameters that were not highly correlated, both MTV and asphericity showed a significant association with LRC (P < 0.001 for MTV and P = 0.021 for asphericity) and OS (P < 0.001 for MTV and asphericity). Asphericity and MTV were binarized and combined for risk stratification, and the prognostic value of the combination was compared with the prognostic value of individual parameters. Bootstrapping revealed significantly better performance by the combinatory approach when compared with MTV (P = 0.012 for LRC and P < 0.001 for OS) and asphericity with regard to OS (P < 0.001) but not for LRC (P = 0.53). Conclusion: We were able to show that asphericity bears independent prognostic value and significantly improves risk stratification when combined with MTV in a comprehensive retrospective cohort of HNSCC patients.

Abstract Accurate channel characterization is extremely helpful in channel estimation, channel coding, and many other parts of communication system design and can effectively reduce overhead. Ray tracing (RT) shows accurate channel reconstruction for specific maps, but the multipath propagation in indoor scenes is far more complex than in outdoor scenes leading to a challenge for RT. This work presents and validates an RT tool for a massive multiple-input multiple-output (MIMO) system in the millimeter-wave frequency bands with the associated channel beamforming algorithm and provides ideas for channel estimation algorithm in subsequent MIMO systems. The impact of the order of interactions, e.g. reflections and diffractions on the channel impulse response reconstruction are analyzed in the RT simulation. The comparison between RT simulated and measured results shows a reasonable level of agreement. The presented RT tool that can provide complete and accurate channel information is of high value for the design of reliable communication systems.

The emergence of 6G networks paves the way for a new era of Artificial Intelligence (AI)-empowered integration of Internet of Things (IoT), autonomous mobile robots (AMRs) and extended reality (XR). This will revolutionise the ways users (e.g., humans, robots) interact with each other and their surroundings expanding their awareness beyond immediate boundaries. Driven by demand for more effective emergency response fuelled by adverse and increasing impacts of climate change, in this paper, we conceptually explore an architectural framework for intelligent emergency response services in 6G networks. The proposed architecture bridges an evolution of supporting technologies in 5G with future intelligent emergency response service requirements. We identify integration of 3D networking (3DN) and ambient IoT (A-IoT), empowered by native AI, network sensing and localisation, as key technologies for the proposed framework. We conclude the paper with discussion on current technology gaps and recommendations that would lead towards future 6G intelligent emergency response.

Judder is a motion artifact describing a perceptual mismatch between human visual system and discrete movements on a display. Judder is correlated with video frame rates, motion speeds and brightness. It significantly degrades the perceived video quality. We present a framework for modeling judder, where the input frames are processed to generate optical flow and a sensitivity map. Concurrently, a specifically designed attention map is integrated into the process. These components collectively contribute to the creation of a feature map, which is utilized to compute judder scores. The proposed approach explicitly takes into account the video frame rate, resulting in an accurate judder score to predict the subjective mean opinion scores ultimately. Our assessment of the proposed framework on an HDR video dataset shows that judderness is highly influenced by the frame rate and, to some extent, by motion speed and brightness. The prediction of perceived quality scores shows improvement compared to the baseline framework.

As learned image codecs (LICs) become more prevalent, their low coding efficiency for out-of-distribution data becomes a bottleneck for some applications. To improve the performance of LICs for screen content (SC) images without breaking backwards compatibility, we propose to introduce parameterized and invertible linear transformations into the coding pipeline without changing the underlying baseline codec's operation flow. We design two neural networks to act as prefilters and postfilters in our setup to increase the coding efficiency and help with the recovery from coding artifacts. Our end-to-end trained solution achieves up to 10% bitrate savings on SC compression compared to the baseline LICs while introducing only 1% extra parameters.

Entropy estimation is essential for the performance of learned image compression. It has been demonstrated that a transformer-based entropy model is of critical importance for achieving a high compression ratio, however, at the expense of a significant computational effort. In this work, we introduce the Efficient Contextformer (eContextformer) -a computationally efficient transformer-based autoregressive context model for learned image compression. The eContextformer efficiently fuses the patch-wise, checkered, and channel-wise grouping techniques for parallel context modeling, and introduces a shifted window spatio-channel attention mechanism. We explore better training strategies and architectural designs and introduce additional complexity optimizations. During decoding, the proposed optimization techniques dynamically scale the attention span and cache the previous attention computations, drastically reducing the model and runtime complexity. Compared to the non-parallel approach, our proposal has ∼145x lower model complexity and ∼210x faster decoding speed, and achieves higher average bit savings on Kodak, CLIC2020, and Tecnick datasets. Additionally, the low complexity of our context model enables online ratedistortion algorithms, which further improve the compression performance. We achieve up to 17% bitrate savings over the intra coding of Versatile Video Coding (VVC) Test Model (VTM) 16.2 and surpass various learning-based compression models.