Hasil untuk "Computer applications to medicine. Medical informatics"

Peipei Zhou, Zheng Dong, Insup Lee et al.

This report summarizes the discussions and recommendations from the NSF Workshop on Algorithm-Hardware Co-design for Medical Applications, held on September 26-27, 2024, in Pittsburgh, PA. The workshop assembled an interdisciplinary cohort of researchers, clinicians, and industry leaders to examine foundational challenges and develop a strategic roadmap for algorithm-hardware co-design in medical computing. The workshop focuses on four thematic areas: (1) teleoperations, telehealth, and surgical operations; (2) wearable and implantable medicine, including implantable living pharmacies; (3) home ICU, hospital systems, and elderly care; and (4) medical sensing, imaging, and reconstruction. This report calls for a fundamental shift in how next-generation medical technologies are conceived, designed, validated, and translated into practice. The report recommends that NSF sustain investment in shared standardized data infrastructures and compute infrastructures, develop clinic workflow-aware systems and human-AI collaboration frameworks, promote scalable validation ecosystems grounded in objective, continuous measures, and physics-informed, and enable safe, accountable, and resilient platforms, including virtual-physical healthcare ecosystems, to de-risk translational pathways. The workshop information can be found on the website: https://sites.google.com/view/nsfworkshop.

Dohyun Chun, Hae Woon Jung, Jongho Kang et al.

Objective We developed a precise, reliable artificial intelligence (AI) model for predicting the future height of children and adolescents based on anthropometric and body composition data. Materials and Methods We used an extensive longitudinal dataset from a large-scale Korean cohort study, which included 588,546 measurements from 96,485 children and adolescents aged 7–18. We developed a prediction model using the light gradient boosting method and integrated anthropometric and body composition metrics along with their standard deviation scores (SDSs) and velocity parameters. Model performance was assessed through root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). We employed Shapley additive explanations (SHAP) for model interpretability. Results The model accurately predicted future heights. For males, the average RMSE, MAE, and MAPE were 2.51 cm, 1.74 cm, and 1.14%, respectively, with female prediction results showing comparable accuracy (2.28 cm, 1.68 cm, and 1.13%, respectively). Shapley additive explanations analysis revealed that the SDS of height, height velocity, and soft lean mass velocity were key predictors of future height. The model created personalized growth curves through estimation of individual-specific height trajectories, comparison with actual measurements, and identification of key variables using local SHAP values. Conclusion Our model produces accurate and personalized growth curves, incorporating explainable AI techniques for enhanced clinical understanding. This method advances pediatric growth assessment and provides robust clinical decision support. Despite limitations including the absence of handwrist radiography comparison and Korean population specificity, our approach demonstrates significant potential for early identification of growth disorders and optimization of growth outcomes.

Joanne K. Garrett, Lewis R. Elliott, Rebecca Lovell et al.

Abstract Many studies linking spatial environmental exposures to health outcomes rely on small statistical geography units, such as Lower-layer Super Output Areas (LSOAs), to estimate exposure. However, these units commonly vary in size, particularly between urban and rural areas, leading to potential exposure misclassification. This study proposes a new method for better capturing environmental exposure at the small statistical geography unit level. Using the Living England Habitat Map as an example, we combined LSOA and postcode-level data to account for varying area sizes and mitigate edge effects. We compared our method with the typical approach, which calculates an average at the small geography unit level. Overall, our proposed method resulted in lower exposure to non-built-up areas compared to averaging across entire LSOAs, whereas exposure to built-up areas was higher by 8–10%. However, these patterns varied based on region, urban/rural classification, land cover type, and LSOA size class. We suggest that this proposed method offers a more consistent approach to estimating neighbourhood exposure to nature.

Pak-Hei Yeung, Jayroop Ramesh, Pengfei Lyu et al.

This paper explores the transfer of knowledge from general vision models pretrained on 2D natural images to improve 3D medical image segmentation. We focus on the semi-supervised setting, where only a few labeled 3D medical images are available, along with a large set of unlabeled images. To tackle this, we propose a model-agnostic framework that progressively distills knowledge from a 2D pretrained model to a 3D segmentation model trained from scratch. Our approach, M&N, involves iterative co-training of the two models using pseudo-masks generated by each other, along with our proposed learning rate guided sampling that adaptively adjusts the proportion of labeled and unlabeled data in each training batch to align with the models' prediction accuracy and stability, minimizing the adverse effect caused by inaccurate pseudo-masks. Extensive experiments on multiple publicly available datasets demonstrate that M&N achieves state-of-the-art performance, outperforming thirteen existing semi-supervised segmentation approaches under all different settings. Importantly, ablation studies show that M&N remains model-agnostic, allowing seamless integration with different architectures. This ensures its adaptability as more advanced models emerge. The code is available at https://github.com/pakheiyeung/M-N.

Orhan Demirci, Bulent Yilmaz

This paper introduces Tree-NET, a novel framework for medical image segmentation that leverages bottleneck feature supervision to enhance both segmentation accuracy and computational efficiency. While previous studies have employed bottleneck feature supervision, their applications have largely been limited to the training phase, offering no computational benefits during training or evaluation. To the best of our knowledge, this study is the first to propose a framework that incorporates two additional training phases for segmentation models, utilizing bottleneck features at both input and output stages. This approach significantly improves computational performance by reducing input and output dimensions with a negligible addition to parameter count, without compromising accuracy. Tree-NET features a three-layer architecture comprising Encoder-Net and Decoder-Net, which are autoencoders designed to compress input and label data, respectively, and Bridge-Net, a segmentation framework that supervises the bottleneck features. By focusing on dense, compressed representations, Tree-NET enhances operational efficiency and can be seamlessly integrated into existing segmentation models without altering their internal structures or increasing model size. We evaluate Tree-NET on two critical segmentation tasks -- skin lesion and polyp segmentation -- using various backbone models, including U-NET variants and Polyp-PVT. Experimental results demonstrate that Tree-NET reduces FLOPs by a factor of 4 to 13 and decreases memory usage, while achieving comparable or superior accuracy compared to the original architectures. These findings underscore Tree-NET's potential as a robust and efficient solution for medical image segmentation.

Satoshi Nishioka, Satoshi Watabe, Yuki Yanagisawa et al.

BackgroundEarly detection of adverse events and their management are crucial to improving anticancer treatment outcomes, and listening to patients’ subjective opinions (patients’ voices) can make a major contribution to improving safety management. Recent progress in deep learning technologies has enabled various new approaches for the evaluation of safety-related events based on patient-generated text data, but few studies have focused on the improvement of real-time safety monitoring for individual patients. In addition, no study has yet been performed to validate deep learning models for screening patients’ narratives for clinically important adverse event signals that require medical intervention. In our previous work, novel deep learning models have been developed to detect adverse event signals for hand-foot syndrome or adverse events limiting patients’ daily lives from the authored narratives of patients with cancer, aiming ultimately to use them as safety monitoring support tools for individual patients. ObjectiveThis study was designed to evaluate whether our deep learning models can screen clinically important adverse event signals that require intervention by health care professionals. The applicability of our deep learning models to data on patients’ concerns at pharmacies was also assessed. MethodsPharmaceutical care records at community pharmacies were used for the evaluation of our deep learning models. The records followed the SOAP format, consisting of subjective (S), objective (O), assessment (A), and plan (P) columns. Because of the unique combination of patients’ concerns in the S column and the professional records of the pharmacists, this was considered a suitable data for the present purpose. Our deep learning models were applied to the S records of patients with cancer, and the extracted adverse event signals were assessed in relation to medical actions and prescribed drugs. ResultsFrom 30,784 S records of 2479 patients with at least 1 prescription of anticancer drugs, our deep learning models extracted true adverse event signals with more than 80% accuracy for both hand-foot syndrome (n=152, 91%) and adverse events limiting patients’ daily lives (n=157, 80.1%). The deep learning models were also able to screen adverse event signals that require medical intervention by health care providers. The extracted adverse event signals could reflect the side effects of anticancer drugs used by the patients based on analysis of prescribed anticancer drugs. “Pain or numbness” (n=57, 36.3%), “fever” (n=46, 29.3%), and “nausea” (n=40, 25.5%) were common symptoms out of the true adverse event signals identified by the model for adverse events limiting patients’ daily lives. ConclusionsOur deep learning models were able to screen clinically important adverse event signals that require intervention for symptoms. It was also confirmed that these deep learning models could be applied to patients’ subjective information recorded in pharmaceutical care records accumulated during pharmacists’ daily work.

Amirreza Hashemi, Sayantan Dutta, Bertrand Georgeot et al.

Background noise in many fields such as medical imaging poses significant challenges for accurate diagnosis, prompting the development of denoising algorithms. Traditional methodologies, however, often struggle to address the complexities of noisy environments in high dimensional imaging systems. This paper introduces a novel quantum-inspired approach for image denoising, drawing upon principles of quantum and condensed matter physics. Our approach views medical images as amorphous structures akin to those found in condensed matter physics and we propose an algorithm that incorporates the concept of mode resolved localization directly into the denoising process. Notably, unlike previous studies that considered localization as a hindrance, our approach considers quantum localization as a fundamental component of image reconstruction which is used to differentiate between noisy and non-noisy modes based on diffusivity and localization measurements. This perspective eliminates the need for hyperparameter tuning, making the proposed method a standalone algorithm which can be implemented with minimal manual intervention and can perform automatic filtering of noise regardless of noise level. Through numerical validation, we showcase the effectiveness of our approach in addressing noise-related challenges in imaging and especially medical imaging, underscoring its relevance for possible quantum computing applications.

Zhuo Liu, Chenhui Yao, Hang Yu et al.

Abstract Recently, deep reinforcement learning has achieved great success by integrating deep learning models into reinforcement learning algorithms in various applications such as computer games and robots. Specially, it is promising for computer-aided diagnosis and treatment to combine deep reinforcement learning with medical big data generated and collected from medical Internet of Things. In this paper, we focus on the potential of the deep reinforcement learning for lung cancer detection as many people are suffering from the lung tumor and about 1.8 million patients died from lung cancer in 2018. Early detection and diagnosis of lung tumor can significantly improve the treatment effect and prolong survival. In this work, we present several representative deep reinforcement learning models that are potential to use for lung cancer detection. Furthermore, we summarize the common types of lung cancer and the main characteristics of each type. Finally, we point out the open challenges and possible future research directions of applying deep reinforcement learning to lung cancer detection, which is expected to promote the evolution of smart medicine with medical Internet of Things.

Jeffrey A. Kline, Brian Reed, Alex Frost et al.

Abstract Background For surveillance of episodic illness, the emergency department (ED) represents one of the largest interfaces for generalizable data about segments of the US public experiencing a need for unscheduled care. This protocol manuscript describes the development and operation of a national network linking symptom, clinical, laboratory and disposition data that provides a public database dedicated to the surveillance of acute respiratory infections (ARIs) in EDs. Methods The Respiratory Virus Laboratory Emergency Department Network Surveillance (RESP-LENS) network includes 26 academic investigators, from 24 sites, with 91 hospitals, and the Centers for Disease Control and Prevention (CDC) to survey viral infections. All data originate from electronic medical records (EMRs) accessed by structured query language (SQL) coding. Each Tuesday, data are imported into the standard data form for ARI visits that occurred the prior week (termed the index file); outcomes at 30 days and ED volume are also recorded. Up to 325 data fields can be populated for each case. Data are transferred from sites into an encrypted Google Cloud Platform, then programmatically checked for compliance, parsed, and aggregated into a central database housed on a second cloud platform prior to transfer to CDC. Results As of August, 2023, the network has reported data on over 870,000 ARI cases selected from approximately 5.2 million ED encounters. Post-contracting challenges to network execution have included local shifts in testing policies and platforms, delays in ICD-10 coding to detect ARI cases, and site-level personnel turnover. The network is addressing these challenges and is poised to begin streaming weekly data for dissemination. Conclusions The RESP-LENS network provides a weekly updated database that is a public health resource to survey the epidemiology, viral causes, and outcomes of ED patients with acute respiratory infections.

Md Mostafijur Rahman, Radu Marculescu

In recent years, medical image segmentation has become an important application in the field of computer-aided diagnosis. In this paper, we are the first to propose a new graph convolution-based decoder namely, Cascaded Graph Convolutional Attention Decoder (G-CASCADE), for 2D medical image segmentation. G-CASCADE progressively refines multi-stage feature maps generated by hierarchical transformer encoders with an efficient graph convolution block. The encoder utilizes the self-attention mechanism to capture long-range dependencies, while the decoder refines the feature maps preserving long-range information due to the global receptive fields of the graph convolution block. Rigorous evaluations of our decoder with multiple transformer encoders on five medical image segmentation tasks (i.e., Abdomen organs, Cardiac organs, Polyp lesions, Skin lesions, and Retinal vessels) show that our model outperforms other state-of-the-art (SOTA) methods. We also demonstrate that our decoder achieves better DICE scores than the SOTA CASCADE decoder with 80.8% fewer parameters and 82.3% fewer FLOPs. Our decoder can easily be used with other hierarchical encoders for general-purpose semantic and medical image segmentation tasks.

K. Koshino, R. Werner, M. Pomper et al.

Recent years have witnessed a rapidly expanding use of artificial intelligence and machine learning in medical imaging. Generative adversarial networks (GANs) are techniques to synthesize images based on artificial neural networks and deep learning. In addition to the flexibility and versatility inherent in deep learning on which the GANs are based, the potential problem-solving ability of the GANs has attracted attention and is being vigorously studied in the medical and molecular imaging fields. Here this narrative review provides a comprehensive overview for GANs and discuss their usefulness in medical and molecular imaging on the following topics: (I) data augmentation to increase training data for AI-based computer-aided diagnosis as a solution for the data-hungry nature of such training sets; (II) modality conversion to complement the shortcomings of a single modality that reflects certain physical measurement principles, such as from magnetic resonance (MR) to computed tomography (CT) images or vice versa; (III) de-noising to realize less injection and/or radiation dose for nuclear medicine and CT; (IV) image reconstruction for shortening MR acquisition time while maintaining high image quality; (V) super-resolution to produce a high-resolution image from low-resolution one; (VI) domain adaptation which utilizes knowledge such as supervised labels and annotations from a source domain to the target domain with no or insufficient knowledge; and (VII) image generation with disease severity and radiogenomics. GANs are promising tools for medical and molecular imaging. The progress of model architectures and their applications should continue to be noteworthy.

Thomas Van Laethem, Priyanka Kumari, Philippe Hubert et al.

There is a rising interest in the modeling and predicting of chromatographic retention. The progress towards more complex and comprehensive models emphasized the need for broad reliable datasets. The present dataset comprises small pharmaceutical compounds selected to cover a wide range in terms of physicochemical properties that are known to impact the retention in reversed-phase liquid chromatography. Moreover, this dataset was analyzed at five pH with two gradient slopes. It provides a reliable dataset with a diversity of conditions and compounds to support the building of new models. To enhance the robustness of the dataset, the compounds were injected individually, and each sequence of injections included a quality control sample. This unambiguous detection of each compound as well as a systematic analysis of a quality control sample ensured the quality of the reported retention times. Moreover, three different liquid chromatographic systems were used to increase the robustness of the dataset.

Helen McTaggart-Cowan, Madeleine T. King, Richard Norman et al.

Abstract Introduction Utility instruments are used to assess patients’ health-related quality of life for cost-utility analysis (CUA). However, for cancer patients, the dimensions of generic utility instruments may not capture all the information relevant to the impact of cancer. Cancer-specific utilities provide a useful alternative. Under the auspices of the Multi-Attribute Utility in Cancer Consortium, a cancer-specific utility algorithm was derived from the FACT-G. The new FACT-8D contains eight dimensions: pain, fatigue, nausea, sleep, work, support from family/friends, sadness, and worry health will get worse. The aim of the study was to obtain a Canadian value set for the FACT-8D. Methods A discrete choice experiment was administered to a Canadian general population online panel, quota sampled by age, sex, and province/territory of residence. Respondents provided responses to 16 choice sets. Each choice set consisted of two health states described by the FACT-8D dimensions plus an attribute representing survival duration. Sample weights were applied and the responses were analyzed using conditional logistic regression, parameterized to fit the quality-adjusted life year framework. The results were converted into utility weights by evaluating the marginal rate of substitution between each level of each FACT-8D dimension with respect to duration. Results 2228 individuals were recruited. The analysis dataset included n = 1582 individuals, who completed at least one choice set; of which, n = 1501 completed all choice sets. After constraining to ensure monotonicity in the utility function, the largest decrements were for the highest levels of pain (− 0.38), nausea (− 0.30), and problems doing work (− 0.23). The decrements of the remaining dimensions ranged from − 0.08 to − 0.18 for their highest levels. The utility of the worst possible health state was defined as − 0.65, considerably worse than dead. Conclusions The largest impacts on utility included three generic dimensions (i.e., pain, support, and work) and nausea, a symptom caused by cancer (e.g., brain tumours, gastrointestinal tumours, malignant bowel obstruction) and by common treatments (e.g., chemotherapy, radiotherapy, opioid analgesics). This may make the FACT-8D more informative for CUA evaluating in many cancer contexts, an assertion that must now be tested empirically in head-to-head comparisons with generic utility measures.

Ryo Omagari, Mayuko Yagishita, Fujio Shiraishi et al.

The human thyroid receptor (hTR)-antagonist activities of 691 compounds were evaluated using a yeast two-hybrid assay with Saccharomyces cerevisiae Y190 introduced hTRα and coactivator. In parallel, those YTOX tests were conducted to evaluate whether those compounds affected either antagonism or toxicity. This is the first report that focuses on the hTR-antagonist activity of many chemical compounds suspected to be endocrine disruptor. In this study, 46 compounds exhibited antagonist activity at 50% of the maximum activity (IC × 50) within 11–9940 nM. In particular, 10,10-Oxybisphenoxarsine, triphenyltin fluoride, triphenyltin hydroxide, and chlorothalonil had strong hTR-antagonist activities. This knowledge gained from the present study will boost chemical regulation strategies for human and wildlife health.

Ming Huang, Jungwei Fan, Julie Prigge et al.

BackgroundPatient-clinician secure messaging is an important function in patient portals and enables patients and clinicians to communicate on a wide spectrum of issues in a timely manner. With its growing adoption and patient engagement, it is time to comprehensively study the secure messages and user behaviors in order to improve patient-centered care. ObjectiveThe aim of this paper was to analyze the secure messages sent by patients and clinicians in a large multispecialty health system at Mayo Clinic, Rochester. MethodsWe performed message-based, sender-based, and thread-based analyses of more than 5 million secure messages between 2010 and 2017. We summarized the message volumes, patient and clinician population sizes, message counts per patient or clinician, as well as the trends of message volumes and user counts over the years. In addition, we calculated the time distribution of clinician-sent messages to understand their workloads at different times of a day. We also analyzed the time delay in clinician responses to patient messages to assess their communication efficiency and the back-and-forth rounds to estimate the communication complexity. ResultsDuring 2010-2017, the patient portal at Mayo Clinic, Rochester experienced a significant growth in terms of the count of patient users and the total number of secure messages sent by patients and clinicians. Three clinician categories, namely “physician—primary care,” “registered nurse—specialty,” and “physician—specialty,” bore the majority of message volume increase. The patient portal also demonstrated growing trends in message counts per patient and clinician. The “nurse practitioner or physician assistant—primary care” and “physician—primary care” categories had the heaviest per-clinician workload each year. Most messages by the clinicians were sent from 7 AM to 5 PM during a day. Yet, between 5 PM and 7 PM, the physicians sent 7.0% (95,785/1,377,006) of their daily messages, and the nurse practitioner or physician assistant sent 5.4% (22,121/408,526) of their daily messages. The clinicians replied to 72.2% (1,272,069/1,761,739) patient messages within 1 day and 90.6% (1,595,702/1,761,739) within 3 days. In 95.1% (1,499,316/1,576,205) of the message threads, the patients communicated with their clinicians back and forth for no more than 4 rounds. ConclusionsOur study found steady increases in patient adoption of the secure messaging system and the average workload per clinician over 8 years. However, most clinicians responded timely to meet the patients’ needs. Our study also revealed differential patient-clinician communication patterns across different practice roles and care settings. These findings suggest opportunities for care teams to optimize messaging tasks and to balance the workload for optimal efficiency.

Shanliang Zhong, Jifeng Feng

Abstract Background Some circular RNAs (circRNAs) can be translated into functional peptides by small open reading frames (ORFs) in a cap-independent manner. Internal ribosomal entry site (IRES) and N6-methyladenosine (m6A) were reported to drive translation of circRNAs. Experimental methods confirming the presence of IRES and m6A site are time consuming and labor intensive. Lacking computational tools to predict ORFs, IRESs and m6A sites for circRNAs makes it harder. Results In this report, we present circPrimer 2.0, a Java based software for annotating circRNAs and predicting ORFs, IRESs, and m6A sites of circRNAs. circPrimer 2.0 has a graphical and a command-line interface that enables the tool to be embed into an analysis pipeline. Conclusions circprimer 2.0 is an easy-to-use software for annotating circRNAs and predicting translation potential of circRNAs, and freely available at www.bio-inf.cn .

Caner Ozer, Arda Guler, Aysel Turkvatan Cansever et al.

To maintain a standard in a medical imaging study, images should have necessary image quality for potential diagnostic use. Although CNN-based approaches are used to assess the image quality, their performance can still be improved in terms of accuracy. In this work, we approach this problem by using Swin Transformer, which improves the poor-quality image classification performance that causes the degradation in medical image quality. We test our approach on Foreign Object Classification problem on Chest X-Rays (Object-CXR) and Left Ventricular Outflow Tract Classification problem on Cardiac MRI with a four-chamber view (LVOT). While we obtain a classification accuracy of 87.1% and 95.48% on the Object-CXR and LVOT datasets, our experimental results suggest that the use of Swin Transformer improves the Object-CXR classification performance while obtaining a comparable performance for the LVOT dataset. To the best of our knowledge, our study is the first vision transformer application for medical image quality assessment.

Weina Jin, Xiaoxiao Li, Mostafa Fatehi et al.

Explainable artificial intelligence (XAI) is essential for enabling clinical users to get informed decision support from AI and comply with evidence-based medical practice. Applying XAI in clinical settings requires proper evaluation criteria to ensure the explanation technique is both technically sound and clinically useful, but specific support is lacking to achieve this goal. To bridge the research gap, we propose the Clinical XAI Guidelines that consist of five criteria a clinical XAI needs to be optimized for. The guidelines recommend choosing an explanation form based on Guideline 1 (G1) Understandability and G2 Clinical relevance. For the chosen explanation form, its specific XAI technique should be optimized for G3 Truthfulness, G4 Informative plausibility, and G5 Computational efficiency. Following the guidelines, we conducted a systematic evaluation on a novel problem of multi-modal medical image explanation with two clinical tasks, and proposed new evaluation metrics accordingly. Sixteen commonly-used heatmap XAI techniques were evaluated and found to be insufficient for clinical use due to their failure in G3 and G4. Our evaluation demonstrated the use of Clinical XAI Guidelines to support the design and evaluation of clinically viable XAI.