Vision Data Research Articles

A Staged Framework for Computer Vision Education: Integrating AI, Data Science, and Computational Thinking

Computer vision education is increasingly important in modern technology curricula; yet, it often lacks a systematic approach integrating both theoretical concepts and practical applications. This study proposes a staged framework for computer vision education designed to progressively build learners’ competencies across four levels. This study proposes a four-staged framework for computer vision education, progressively introducing concepts from basic image recognition to advanced video analysis. Validity assessments were conducted twice with 25 experts in the field of AI education and curricula. The results indicated high validity of the staged framework. Additionally, a pilot program, applying computer vision to acid–base titration activities, was implemented with 40 upper secondary school students to evaluate the effectiveness of the staged framework. The pilot program showed significant improvements in students’ understanding and interest in both computer vision and scientific inquiry. This research contributes to the AI educational field by offering a structured, adaptable approach to computer vision education, integrating AI, data science, and computational thinking. It provides educators with a structured guide for implementing progressive, hands-on learning experiences in computer vision, while also highlighting areas for future research and improvement in educational methodologies.

Transparent, Flexible, Responsive Switching "Delayed" Amphiphilic Coatings Designed on the Basis of the Full-Cycle Antifouling Strategy.

Marine fouling on the surface of ships and equipment not only creates problems of enhanced resistance to navigation and increased energy consumption but also leads to unclear vision and inaccurate data collection. Antifouling coatings to resist fouling are effective, but it is difficult to achieve long-lasting fouling protection with a single interface state. Switching the status of the interface by intelligent response is a reasonable way to achieve full-cycle efficient antifouling. In this study, the hydrophobic and active antifouling interface in the initial state was achieved by adopting the fluorine-containing group and the natural extract (citronellol) as the antifouling active site. The switching of the interface relies on silanes, which respond to the generation of zwitterions in a seawater environment. Eventually, the interface switched from the hydrophobic state to the amphiphilic state with delayed formation, which achieved continued antifouling. Based on the full-cycle antifouling concept, the combination of low surface energy and antifouling active ingredients in the initial state sustainably switched surfaces in the midterm (free radicals generated during the hydrolysis process), and amphiphilic interfaces formed by "delays" produced an antifouling effect from the initial stage to the subsequent stage. The excellent antifouling activity (bacterial and diatom attachment inhibition by over 90% and significantly reduced mussel adhesion force), optical transparency, and flexibility of these coatings indicate the potential for the application of antifouling coatings prepared from hyperbranched silicone-based resins; they can also be used for data extraction sensors, underwater probes, marine photovoltaics, and other areas where transparency is required.

Estimation of Lower Limb Joint Angles Using sEMG Signals and RGB-D Camera.

Estimating human joint angles is a crucial task in motion analysis, gesture recognition, and motion intention prediction. This paper presents a novel model-based approach for generating reliable and accurate human joint angle estimation using a dual-branch network. The proposed network leverages combined features derived from encoded sEMG signals and RGB-D image data. To ensure the accuracy and reliability of the estimation algorithm, the proposed network employs a convolutional autoencoder to generate a high-level compression of sEMG features aimed at motion prediction. Considering the variability in the distribution of sEMG signals, the proposed network introduces a vision-based joint regression network to maintain the stability of combined features. Taking into account latency, occlusion, and shading issues with vision data acquisition, the feature fusion network utilizes high-frequency sEMG features as weights for specific features extracted from image data. The proposed method achieves effective human body joint angle estimation for motion analysis and motion intention prediction by mitigating the effects of non-stationary sEMG signals.

Data Harmonization, Standardization, and Collaboration for Diabetic Retinal Disease (DRD) Research: Report From the 2024 Mary Tyler Moore Vision Initiative Workshop on Data.

The 2024 Mary Tyler Moore Vision Initiative (MTM Vision) Workshop on Data convened to discuss best practices and specific considerations for building a comprehensive, shareable MTM Vision data lake. The workshop aimed to accelerate the development of new indications, therapies, and regulatory pathways for diabetic retinal disease (DRD) by standardizing and harmonizing clinical data and ocular 'omics analyses. Standardization of data collection, the use of common data elements, and data interoperability were emphasized, alongside federated learning approaches to promote data sharing and collaboration while maintaining data privacy and security. The integration of molecular data with other multimodal data types was recognized as a promising strategy for leveraging machine learning and AI approaches to advancing therapeutics development and improving treatment outcomes for DRD patients. Partnerships with entities such as the National Eye Institute, part of the National Institutes of Health, foundations, and industry were deemed vital for the successful implementation of these initiatives.

Accelerometer-assisted computer vision data fusion framework for structural dynamic displacement reconstruction

Accurate displacement measurement provides crucial information for evaluating the structural health condition. It is important to note that both direct and indirect displacement measurement methods have their own limitations, which can be remedied by leveraging each other’s strengths. In this study, a novel accelerometer-assisted computer vision data fusion framework is developed for accurately reconstructing structural dynamic displacement. The framework does not require a specific mathematical model or prior knowledge about the data’s characteristics, allowing it to be applied more broadly across different applications without the constraints of model assumptions. The core of this framework is to integrate successive variational mode decomposition (SVMD) and Bayesian optimization approaches to adaptively determine the weight factors of the fusion components. Importantly, an enhanced optical flow approach is presented for converting pixel movement to structural displacement from natural targets. This approach can effectively reduce the selection of mis-matched points within the ROI (Region of Interest), thereby reducing drift errors. The developed framework is verified via shaking table tests of a reinforced concrete frame structure under seismic excitation. Results indicate that the developed framework excels in accurately estimating structural dynamic displacement. Compared to single-vision displacement identification results, the proposed framework demonstrates lower peak error (< 1.65 mm) and normalized root mean square error (< 0.30). Meanwhile, the reconstructed displacement, by introducing dynamic displacement component from acceleration measurement, presents a wider frequency range than the vision-based displacement.

Bridging vision and touch: advancing robotic interaction prediction with self-supervised multimodal learning.

Predicting the consequences of the agent's actions on its environment is a pivotal challenge in robotic learning, which plays a key role in developing higher cognitive skills for intelligent robots. While current methods have predominantly relied on vision and motion data to generate the predicted videos, more comprehensive sensory perception is required for complex physical interactions such as contact-rich manipulation or highly dynamic tasks. In this work, we investigate the interdependence between vision and tactile sensation in the scenario of dynamic robotic interaction. A multi-modal fusion mechanism is introduced to the action-conditioned video prediction model to forecast future scenes, which enriches the single-modality prototype with a compressed latent representation of multiple sensory inputs. Additionally, to accomplish the interactive setting, we built a robotic interaction system that is equipped with both web cameras and vision-based tactile sensors to collect the dataset of vision-tactile sequences and the corresponding robot action data. Finally, through a series of qualitative and quantitative comparative study of different prediction architecture and tasks, we present insightful analysis of the cross-modality influence between vision, tactile and action, revealing the asymmetrical impact that exists between the sensations when contributing to interpreting the environment information. This opens possibilities for more adaptive and efficient robotic control in complex environments, with implications for dexterous manipulation and human-robot interaction.

Efficacy and Toxicity of [177Lu]Lu-PSMA-617 for Metastatic Castration-Resistant Prostate Cancer: Results from the U.S. Expanded-Access Program and Comparisons with Phase 3 VISION Data.

The phase 3 VISION trial demonstrated that [177Lu]Lu-PSMA-617 prolonged progression-free survival and overall survival (OS) in prostate-specific membrane antigen [PSMA]-positive metastatic castration-resistant prostate cancer (mCRPC) patients who progressed on taxane-based chemotherapy and androgen receptor-signaling inhibitors (ARSIs). The U.S. expanded-access program (EAP; NCT04825652) was opened to provide access to [177Lu]Lu-PSMA-617 for eligible patients until regulatory approval was obtained. This study aimed to evaluate the efficacy and safety profile of [177Lu]Lu-PSMA-617 within the EAP and compare the results with those from the VISION trial. Methods: Patients enrolled in the EAP at 4 institutions in the United States with available toxicity and outcome data were included. Outcome measures included OS, a prostate-specific antigen (PSA) response rate (RR) of at least 50%, and incidences of toxicity according to Common Terminology Criteria for Adverse Events version 5.0. Differences in baseline characteristics, outcome data, and toxicity between the EAP and VISION were evaluated using t testing of proportions and survival analyses. Results: In total, 117 patients with mCRPC who received [177Lu]Lu-PSMA-617 within the EAP between May 2021 and March 2022 were eligible and included in this analysis. Patients enrolled in the EAP were more heavily pretreated with ARSI (≥2 ARSI regimens: 70% vs. 46%; P < 0.001) and had worse performance status at baseline (Eastern Cooperative Oncology Group score ≥ 2: 19% vs. 7%; P < 0.001) than VISION patients. EAP and VISION patients had similar levels of grade 3 or higher anemia (18% vs. 13%; P = 0.15), thrombocytopenia (13% vs. 8%; P = 0.13), and neutropenia (3% vs. 3%; P = 0.85) and similar PSA RRs (42% vs. 46%; P = 0.50) and OS (median: 15.1 vs. 15.3 mo; P > 0.05). Conclusion: Patients with PSMA-positive mCRPC who received [177Lu]Lu-PSMA-617 within the EAP were later in their disease trajectory than VISION patients. Patients enrolled in the EAP achieved similar PSA RRs and OS and had a safety profile similar to that of the VISION trial patients.

EARLY DETECTION OF LOW VISION DETERMINANT FACTORS USING THE E-SIGALON SELF-ASSESSMENT APPLICATION

Background: Low Vision is an end-stage condition that cannot be cured. The best treatment for low vision is the early detection. However, there is still no integrated low vision data in health services, the lack of public knowledge about low vision, the low rate of case discovery, and the lack of optimal referral and treatment mechanisms for low vision are problems that must be addressed at this time. Purpose: This study aims to detect early low vision determinant factors using a self-assessment application. Methods: This study is analytical and quantitative research. Respondents for this research were from five selected populations in East Java who filled out the e-SIGALON application from September to December 2023. The respondents were suspected as having low vision if they had score of 6 or higher. The variables were analyzed using logistic binary regression, the validity and reliability test was also performed. Results: From 446 respondents, there were 237 people (53.14%) suspected of low vision but only 66 people came to referral hospital. The low vision suspect showed 12 of the 15 statistically significant questions (p&lt;0.05). The most common was sitting very close to the television/monitor (48.43%), followed by having difficulties in seeing objects in dim light (44.17%) and difficulties in doing the things they want to do (40.13%). Conclusion: The e-SIGALON application can be a solution in managing low vision. Through this application, people can easily find out and recognize whether they have low vision or not.

Early Clinical Experience of Finerenone in People with Chronic Kidney Disease and Type 2 Diabetes in Japan-A Multi-Cohort Study from the FOUNTAIN (FinerenOne mUltidatabase NeTwork for Evidence generAtIoN) Platform.

Background: In the phase 3 clinical trials FIGARO-DKD and FIDELIO-DKD, finerenone reduced the risk of cardiovascular and kidney events among people with chronic kidney disease (CKD) and type 2 diabetes (T2D). Evidence regarding finerenone use in real-world settings is limited. Methods: A retrospective cohort study (NCT06278207) using two Japanese nationwide hospital-based databases provided by Medical Data Vision (MDV) and Real World Data Co., Ltd. (RWD Co., Kyoto Japan), converted to the OMOP common data model, was conducted. Persons with CKD and T2D initiating finerenone from 1 July 2021, to 30 August 2023, were included. Baseline characteristics were described. The occurrence of hyperkalemia after finerenone initiation was assessed. Results: 1029 new users of finerenone were included (967 from MDV and 62 from RWD Co.). Mean age was 69.5 and 72.4 years with 27.3% and 27.4% being female in the MDV and RWD Co. databases, respectively. Hypertension (92 and 95%), hyperlipidemia (59 and 71%), and congestive heart failure (60 and 66%) were commonly observed comorbidities. At baseline, 80% of persons were prescribed angiotensin-converting-enzyme inhibitors or angiotensin-receptor blockers. Sodium-glucose cotransporter 2 inhibitors and glucagon-like peptide 1 receptor agonists were prescribed in 72% and 30% of the study population, respectively. The incidence proportions of hyperkalemia were 2.16 and 2.70 per 100 persons in the MDV and RWD Co. databases, respectively. There were no hospitalizations associated with hyperkalemia observed in either of the two datasets. Conclusions: For the first time, we report the largest current evidence on the clinical use of finerenone in real-world settings early after the drug authorization in Japan. This early evidence from clinical practice suggests that finerenone is used across comorbidities and comedications.

Enhancing IIoT Vision Data Transmission and Processing via Spatial-Difference-Attention-Guided Saliency Detection

Enhancing IIoT Vision Data Transmission and Processing via Spatial-Difference-Attention-Guided Saliency Detection

Navigating the AI horizon in hospitality: a novel classification and future research agenda

PurposeThis study aims to evaluate Artificial Intelligence (AI) research in the hospitality industry based on the service AI framework (mechanical-thinking-feeling) and highlight prospective avenues for future inquiry in this growing domain.Design/methodology/approachThis paper conceptualizes timely concepts supported by research spanning multiple domains.FindingsThis research introduces a novel classification for the domain of AI hospitality research. This classification encompasses prediction and pattern recognition, computer vision, NLP, behavioral research, and synthetic data generation. Based on this classification, this study identifies and elaborates upon five emerging research topics, each linked to a corresponding set of research questions. These focal points encompass the realms of interpretable AI, controllable AI, AI ethics, collaborative AI, and synthetic data generation.Originality/valueThis viewpoint provides a foundational framework and a directional compass for future research in AI within the hospitality industry. It pushes the industry forward with a balanced approach to leveraging AI to augment human potential and enrich customer experiences. Both the classification and the research agenda would contribute to the body of knowledge that will guide the industry toward a future where technology and human service coalesce to create unparalleled value for all stakeholders.

Association of vision and hearing impairment and dietary diversity among the oldest old in China: findings from the Chinese longitudinal healthy longevity survey

BackgroundThe presence of sensory impairment among older age cohorts exerts a significant impact on both individuals and society generally. Although the impact of dietary patterns on health is vital across all stages of life, there still a paucity of comprehensive research on the association between dietary variety and sensory impairments.ObjectiveTo investigate the potential relationship between dietary diversity and the prevalence of visual and hearing impairment or dual sensory impairments (visual and hearing impairment) among the oldest old population.MethodsThis is a cross-sectional study relied on data obtained from the 2018 survey conducted by the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Subjects aged 80 and older with complete vision and hearing data were included in the study. Multivariate logistic regression models were developed to examine the association between dietary components and visual and hearing impairment while controlling for age, gender, socioeconomic demographic factors, living habits, other food habits, and general health status.ResultsThe study included 10,093 participants, with an average age of 92.29 ± 7.75 years. Vision and hearing function were assessed based on the ability to distinguish the direction of the break in the circle and the requirement for hearing aids, respectively. Upon controlling for confounding variables, individuals with a greater Dietary Diversity Score (DDS, the number of food groups, range: 1–11) had a reduced likelihood of experiencing visual impairment (odds ratio [OR] = 0.944, 95% confidence interval [CI], 0.915—0.974) and dual sensory impairment (OR = 0.930, 95% CI, 0.905—0.955). In comparison to the low dietary variety group (insufficient dietary diversity, DDS < 4), the high dietary diversity group (sufficient dietary diversity, DDS ≥ 4) exhibited a decreased risk of visual impairment (OR = 0.820, 95% CI, 0.713—0.944) and dual sensory impairment (OR = 0.751, 95% CI, 0.667—0.846). However, no statistically significant correlation was observed between dietary diversity and the presence of only hearing impairment (OR = 0.924, 95% CI, 0.815—1.047) (P < 0.05).Conclusions and implicationsThe synthesis of research findings suggests that following diverse dietary patterns and healthy nutritional practices may be an effective and affordable way to prevent age-related decline in visual impairment and dual sensory impairment.

MACHINE LEARNING TECHNIQUE TO DETECT DEFECTS ON THE STEEL SURFACE

In this paper is presented a technique concerning the machine learning using computer vision data for the detection of defects on the steel surface. The early detection of defects can reduce the product damage and manufacturing cost. Moreover, it provides information that will provide data concerning quality of products and correct classification and this means that the product will not be rejected by the customer and avoid other costs. The non-contact inspection of the surface defects using the computer vision has become very popular in manufacturing industrial systems as it provides reliable and fast results. The machine learning achieved impressive results in image classification tasks, though it requires the previous learning phase to be completed. The proposed machine learning technique allows fast identification of defects and also permits learning during the process.

Drug utilization pattern of romosozumab and other osteoporosis treatments in Japan, 2019-2021.

Describe real-world treatment of osteoporosis and romosozumab treatment patterns in Japan. Data for patients initiating romosozumab or other antiosteoporotic medications between March 01, 2018, and May 31, 2022, were extracted from the Medical Data Vision (MDV) and Japan Medical Data Center (JMDC) databases. Patients were categorized into four cohorts: those who newly initiated romosozumab within the first (MDV: n = 4782; JMDC: n = 2578) or second (MDV: n = 3888; JMDC: n = 2446) year after launch and those who initiated teriparatide (TPTD; MDV: n = 14,576; JMDC: n = 8259) or non-TPTD antiosteoporotic medications within the first year of romosozumab launch (MDV: n = 352,142; JMDC: n = 185,785). Mean age, sex, baseline cardiovascular history, comorbidities, and concomitant medications were similar across cohorts. In the MDV database, fracture history was higher in the romosozumab year-1 (59.3%), year-2 (64.1%), and TPTD (65.5%) cohorts versus the non-TPTD cohort (24.4%). Similar rates were identified in the JMDC database: romosozumab year-1 (64.7%), year-2 (66.6%), TPTD (67.5%), and non-TPTD (27.8%). Vertebral fractures were most common in all cohorts. 12-month romosozumab discontinuation varied between the year-1 and year-2 cohorts in MDV (62.4% and 58.8%) and JMDC (57.1% and 52.7%), whereas mean number of injections remained consistent (MDV: 9.7 and 9.8; JMDC: 7.3 and 7.8). Romosozumab persistence was lower in year-1 versus year-2 (MDV: 37.6% and 42.9%; JMDC: 41.2% and 47.3%). Patients initiating romosozumab and TPTD had a high fracture history. Given the dual effects of promoting bone formation and suppressing resorption, improving romosozumab adherence and persistence over time may be important for antiosteoporotic therapy.

Deep learning for computer vision based activity recognition and fall detection of the elderly: a systematic review

As the proportion of elderly individuals in developed countries continues to rise globally, addressing their healthcare needs, particularly in preserving their autonomy, is of paramount concern. A growing body of research focuses on Ambient Assisted Living (AAL) systems, aimed at alleviating concerns related to the independent living of the elderly. This systematic review examines the literature pertaining to fall detection and Human Activity Recognition (HAR) for the elderly, two critical tasks for ensuring their safety when living alone. Specifically, this review emphasizes the utilization of Deep Learning (DL) approaches on computer vision data, reflecting current trends in the field. A comprehensive search yielded 2,616 works from five distinct sources, spanning the years 2019 to 2023 (inclusive). From this pool, 151 relevant works were selected for detailed analysis. The review scrutinizes the employed DL models, datasets, and hardware configurations, with particular emphasis on aspects such as privacy preservation and real-world deployment. The main contribution of this study lies in the synthesis of recent advancements in DL-based fall detection and HAR for the elderly, providing insights into the state-of-the-art techniques and identifying areas for further improvement. Given the increasing importance of AAL systems in enhancing the quality of life for the elderly, this review serves as a valuable resource for researchers, practitioners, and policymakers involved in developing and implementing such technologies.Graphical abstract

Multidimensional vision sensors for information processing.

The visual scene in the physical world integrates multidimensional information (spatial, temporal, polarization, spectrum and so on) and typically shows unstructured characteristics. Conventional image sensors cannot process this multidimensional vision data, creating a need for vision sensors that can efficiently extract features from substantial multidimensional vision data. Vision sensors are able to transform the unstructured visual scene into featured information without relying on sophisticated algorithms and complex hardware. The response characteristics of sensors can be abstracted into operators with specific functionalities, allowing for the efficient processing of perceptual information. In this Review, we delve into the hardware implementation of multidimensional vision sensors, exploring their working mechanisms and design principles. We exemplify multidimensional vision sensors built on emerging devices and silicon-based system integration. We further provide benchmarking metrics for multidimensional vision sensors and conclude with the principle of device-system co-design and co-optimization.

Secure and Efficient Federated Learning for Smart Optical Cable Monitoring Systems

This study presents a pioneering federated multi-modal data classification model tailored for smart optical cable monitoring systems. By harnessing federated learning techniques, the model ensures data privacy while achieving performance on par with centralized models. Through comprehensive experiments spanning various modalities, including vision and auditory data, our approach showcases promising outcomes, as evidenced by accuracy and precision metrics. Furthermore, comparative analyses with centralized models highlight the superior data security and reduced network strain offered by federated learning. Moreover, we delineate the design and deployment of a smart optical cable monitoring system leveraging edge computing, accentuating the pivotal role of information technology in elevating operational efficiency within the cable monitoring domain. Through meticulous analysis and simulations, our proposed system adeptly monitors environmental variables, thereby bolstering safety and efficiency in smart optical cable monitoring applications.

Enhancing fall risk assessment: instrumenting vision with deep learning during walks

BackgroundFalls are common in a range of clinical cohorts, where routine risk assessment often comprises subjective visual observation only. Typically, observational assessment involves evaluation of an individual’s gait during scripted walking protocols within a lab to identify deficits that potentially increase fall risk, but subtle deficits may not be (readily) observable. Therefore, objective approaches (e.g., inertial measurement units, IMUs) are useful for quantifying high resolution gait characteristics, enabling more informed fall risk assessment by capturing subtle deficits. However, IMU-based gait instrumentation alone is limited, failing to consider participant behaviour and details within the environment (e.g., obstacles). Video-based eye-tracking glasses may provide additional insight to fall risk, clarifying how people traverse environments based on head and eye movements. Recording head and eye movements can provide insights into how the allocation of visual attention to environmental stimuli influences successful navigation around obstacles. Yet, manual review of video data to evaluate head and eye movements is time-consuming and subjective. An automated approach is needed but none currently exists. This paper proposes a deep learning-based object detection algorithm (VARFA) to instrument vision and video data during walks, complementing instrumented gait.MethodThe approach automatically labels video data captured in a gait lab to assess visual attention and details of the environment. The proposed algorithm uses a YoloV8 model trained on with a novel lab-based dataset.ResultsVARFA achieved excellent evaluation metrics (0.93 mAP50), identifying, and localizing static objects (e.g., obstacles in the walking path) with an average accuracy of 93%. Similarly, a U-NET based track/path segmentation model achieved good metrics (IoU 0.82), suggesting that the predicted tracks (i.e., walking paths) align closely with the actual track, with an overlap of 82%. Notably, both models achieved these metrics while processing at real-time speeds, demonstrating efficiency and effectiveness for pragmatic applications.ConclusionThe instrumented approach improves the efficiency and accuracy of fall risk assessment by evaluating the visual allocation of attention (i.e., information about when and where a person is attending) during navigation, improving the breadth of instrumentation in this area. Use of VARFA to instrument vision could be used to better inform fall risk assessment by providing behaviour and context data to complement instrumented e.g., IMU data during gait tasks. That may have notable (e.g., personalized) rehabilitation implications across a wide range of clinical cohorts where poor gait and increased fall risk are common.

AI in evaluating ambulation of stroke patients: severity classification with video and functional ambulation category scale

ABSTRACT Background The evaluation of gait function and severity classification of stroke patients are important to determine the rehabilitation goal and the level of exercise. Physicians often qualitatively evaluate patients’ walking ability through visual gait analysis using naked eye, video images, or standardized assessment tools. Gait evaluation through observation relies on the doctor’s empirical judgment, potentially introducing subjective opinions. Therefore, conducting research to establish a basis for more objective judgment is crucial. Objective To verify a deep learning model that classifies gait image data of stroke patients according to Functional Ambulation Category (FAC) scale. Methods Gait vision data from 203 stroke patients and 182 healthy individuals recruited from six medical institutions were collected to train a deep learning model for classifying gait severity in stroke patients. The recorded videos were processed using OpenPose. The dataset was randomly split into 80% for training and 20% for testing. Results The deep learning model attained a training accuracy of 0.981 and test accuracy of 0.903. Area Under the Curve(AUC) values of 0.93, 0.95, and 0.96 for discriminating among the mild, moderate, and severe stroke groups, respectively. Conclusion This confirms the potential of utilizing human posture estimation based on vision data not only to develop gait parameter models but also to develop models to classify severity according to the FAC criteria used by physicians. To develop an AI-based severity classification model, a large amount and variety of data is necessary and data collected in non-standardized real environments, not in laboratories, can also be used meaningfully.

Efficacy and toxicity of [177Lu]Lu-PSMA-617 for metastatic castration-resistant prostate cancer in a real-world setting: Results from the U.S. Expanded Access Program and comparisons with phase 3 VISION data.

Efficacy and toxicity of [¹⁷⁷Lu]Lu-PSMA-617 for metastatic castration-resistant prostate cancer in a real-world setting: Results from the U.S. Expanded Access Program and comparisons with phase 3 VISION data.