RESUMO
This study addresses challenges related to privacy issues in utilizing medical data, particularly the protection of personal information. To overcome this obstacle, the research focuses on data synthesis using real-world time-series generative adversarial networks (RTSGAN). A total of 53,005 data were synthesized using the dataset of 15,799 patients with colorectal cancer. The results of the quantitative evaluation of the synthetic data's quality are as follows: the Hellinger distance ranged from 0 to 0.25; the train on synthetic, test on real (TSTR) and train on real, test on synthetic (TRTS) results showed an average area under the curve of 0.99 and 0.98; a propensity mean squared error was 0.223. The synthetic and real data were similar in the qualitative methods including t-SNE and histogram analyses. The application of synthetic data in predicting five-year survival in colorectal cancer patients demonstrates comparable performance to models based on real data. This study employs distance to closest records and membership inference test to assess potential privacy exposure, revealing minimal risk. This study demonstrated that it is feasible to synthesize medical data, including time-series data, using the RTSGAN, and the synthetic data can be evaluated to accurately reflect the characteristics of real data through quantitative and qualitative methods as well as by utilizing real-world artificial intelligence models.
Assuntos
Neoplasias Colorretais , Humanos , Redes Neurais de ComputaçãoRESUMO
The undergraduate degree program in medical data science aims to train future data scientists with a medical lens to tackle healthcare challenges using a data-driven approach. The program is a collaborative effort within the Berlin University Alliance, addressing the lack of healthcare-focused data science education in Berlin and Germany. The curriculum covers mathematics, informatics, medical informatics, and medicine, featuring diverse didactic formats. Graduates will be equipped to lead data science and digital transformation projects in healthcare.
Assuntos
Currículo , Ciência de Dados , Informática Médica , Ciência de Dados/educação , Informática Médica/educação , Alemanha , Educação de Graduação em Medicina , HumanosRESUMO
This study addresses the missing data problem in the large-scale medical dataset MIMIC-IV, especially in situations where intubation-extubation events are paired. We employed a strategy involving patient scenario works that checked the temporal order and logical links of intubation/extubation data, and seven reconstruction rules for handling missing values. Through this, we reduced the overall loss rate from 36.89% (3321 records) to 13.37% (1204 records) and achieved a 37.26% data increase (+2117 records) compared to before reconstruction(6582).
Assuntos
Registros Eletrônicos de Saúde , Humanos , Intubação IntratraquealRESUMO
In recent years, there has been a rapid growth in the use of AI in the clinical domain. In order to keep pace with this development, a framework should be created in which clinical AI models can be easily trained, managed and applied. In our study, we propose a clinical AI platform that supports the development cycle and application of clinical AI models. We consider not only the development of an isolated clinical AI platform, but also its integration into clinical IT. This includes the consideration of so-called medical data integration centers. We evaluate our approach with the aid of a clinical AI use case to demonstrate the functionality of our clinical AI platform.
Assuntos
Inteligência Artificial , Registros Eletrônicos de Saúde , Integração de Sistemas , Humanos , Informática MédicaRESUMO
Data quality deficiencies significantly limit the applicability of real-world data in data-driven medical research. In this study, using an oncological use case, we report and discuss common quality deficiencies in real-world medical datasets, such as missing data, class imbalances, and timeliness issues. We compiled a multi-departmental real-world dataset comprising 13861 cancer cases diagnosed at University Hospital Cologne and examined data quality throughout the data integration process.
Assuntos
Confiabilidade dos Dados , Neoplasias , Humanos , Neoplasias/terapia , Oncologia , Alemanha , Registros Eletrônicos de SaúdeRESUMO
Data sharing spaces for medical data are necessary to facilitate research. To make medical data available for research, a mechanism is preferable that not only provides data a researcher has legal access to, but also contributes to the investigation of their specific research hypothesis. We propose a three-party two-stage search algorithm initiated by a researcher on centrally stored but technically and organizationally separated data. The search seeks to minimize the risk of reidentification of patients and to enable data minimization. In the first stage, we only access data IDs of patients meeting the cohort criteria. In the second stage, the actual data is downloaded if the set of matching patients satisfies the minimum cohort size. Our approach is privacy preserving, as only the researcher is able to connect medical and demographic data, while no single malicious party can get data access. We thereby hope to pave the way for a privacy-aware health data sharing space as currently proposed by the EU.
Assuntos
Algoritmos , Segurança Computacional , Confidencialidade , Humanos , Disseminação de Informação , Registros Eletrônicos de Saúde , Privacidade , Armazenamento e Recuperação da InformaçãoRESUMO
The sharing of personal health data is highly regulated due to privacy and security concerns. An alternative to sharing personal data is to share synthetic data, because ideally it should be impossible to reconstruct real personal data from synthetic data, which is called privacy. At the same time, the structure of the synthetic data should be as similar as possible to the structure of the real data to ensure that conclusions drawn from the synthetic data are also valid for the real data, which is called fidelity. Typically, there is a tradeoff between fidelity and privacy for synthetic health data. We study the fidelity and privacy of cancer data synthesized using generative machine learning approaches. To generate synthetic cancer data, we use variational autoencoders (VAEs), generative adversarial networks (GANs), and denoising diffusion probabilistic models (DDPMs). The tabular cancer registry data studied have nine categorical variables from breast cancer patients. We find that DDPMs generate synthetic cancer data with higher fidelity; that is, the structure of the synthetic data is more similar to the real cancer data than the data generated by VAEs and GANs. At the same time, synthetic cancer data from DDPMs pose a greater privacy risk because the data are more likely to reveal information from real patients than synthetic data from VAEs and GANs.
Assuntos
Sistema de Registros , Humanos , Confidencialidade , Aprendizado de Máquina , Segurança Computacional , Neoplasias , Neoplasias da Mama , Feminino , PrivacidadeRESUMO
Objective: This retrospective study aims to investigate the treatment of tic disorder (TD) in Dongfang Hospital affiliated with Beijing University of Chinese Medicine, explore its underlying mechanism, and provide valuable insights for future research and clinical management of TD. Methods: The electronic medical records of children with TD, from 2015 to 2021, were extracted from the information system of Dongfang Hospital affiliated with Beijing University of Chinese Medicine. The clinical characteristics of TD, utilization patterns of Chinese herbal medicine and synthetic drugs in prescriptions, as well as their pharmacological effects, were statistically described and categorized. In addition, association rules and network pharmacology were employed to identify core prescriptions (CPs) and elucidate their microscopic molecular mechanisms in treating TD. Results: The age range of the children was from 6 to 11â years, with a higher proportion of male participants than female ones. The average duration of treatment was 6â weeks. Regimen Z for the treatment of TD can be summarized as follows: Chinese herbal medicine [Saposhnikoviae Radix (FangFeng), Puerariae Lobatae Radix (GeGen), Uncariae Ramulus cum Uncis (GouTeng), Acori Tatarinowii Rhizoma (ShiChangPu), Chuanxiong Rhizoma (ChuanXiong)] and vitamins [lysine, inosite, and vitamin B12 oral solution] form the basic treatment, combined with immunomodulators, antibiotics, electrolyte-balancing agents, and antiallergic agents. CPs primarily exerted their effects through the modulation of gene expression (transcription), the immune system, and signal transduction pathways, with interleukin-4 and interleukin-13 pathways being particularly crucial. Among the lysine synthetic drugs used, inosite and vitamin B12 oral solution were the most frequently prescribed. Conclusion: The regimen Z drug treatment holds significant importance in the field, as it exerts its therapeutic effects through a multitude of pathways and intricate interventions. Chinese herbal medicine primarily regulates immune system-related pathways, while synthetic drugs predominantly consist of vitamins.
RESUMO
INTRODUCTION: In German and international research networks different approaches concerning patient consent are applied. So far it is time-consuming to find out to what extent data from these networks can be used for a specific research project. To make the contents of the consents queryable, we aimed for a permission-based approach (Opt-In) that can map both the permission and the withdrawal of consent contents as well as make it queryable beyond project boundaries. MATERIALS AND METHODS: The current state of research was analysed in terms of approach and reusability. Selected process models for defining consent policies were abstracted in a next step. On this basis, a standardised semantic terminology for the description of consent policies was developed and initially agreed with experts. In a final step, the resulting code was evaluated with regards to different aspects of applicability. RESULTS: A first and extendable version for a Semantic Consent Code (SCC) based on 3-axis (CLASS, ACTION, PURPOSE) was developed, consolidated und published. The added value achieved by the SCC was illustrated using the example of real consents from large national research associations (Medical Informatics Initiative and NUM NAPKON/NUKLEUS). The applicability of the SCC was successfully evaluated in terms of the manual semantic mapping of consents by briefly trained personnel and the automated interpretability of consent policies according to the SCC (and vice versa). In addition, a concept for the use of the SCC to simplify consent queries in heterogeneous research scenarios was presented. CONCLUSIONS: The Semantic Consent Code has already successfully undergone initial evaluations. As the published 3-axis code SCC is an essential preliminary work to standardising initially diverse consent texts and contents and can iteratively be extended in multiple ways in terms of content and technical additions. It should be extended in cooperation with the potential user community.
Assuntos
Pesquisa Biomédica , Documentação , Consentimento Livre e Esclarecido , Semântica , Consentimento Livre e Esclarecido/normas , Humanos , Pesquisa Biomédica/normas , Documentação/normas , AlemanhaRESUMO
BACKGROUND: The use of machine learning in medical diagnosis and treatment has grown significantly in recent years with the development of computer-aided diagnosis systems, often based on annotated medical radiology images. However, the lack of large annotated image datasets remains a major obstacle, as the annotation process is time-consuming and costly. This study aims to overcome this challenge by proposing an automated method for annotating a large database of medical radiology images based on their semantic similarity. RESULTS: An automated, unsupervised approach is used to create a large annotated dataset of medical radiology images originating from the Clinical Hospital Centre Rijeka, Croatia. The pipeline is built by data-mining three different types of medical data: images, DICOM metadata and narrative diagnoses. The optimal feature extractors are then integrated into a multimodal representation, which is then clustered to create an automated pipeline for labelling a precursor dataset of 1,337,926 medical images into 50 clusters of visually similar images. The quality of the clusters is assessed by examining their homogeneity and mutual information, taking into account the anatomical region and modality representation. CONCLUSIONS: The results indicate that fusing the embeddings of all three data sources together provides the best results for the task of unsupervised clustering of large-scale medical data and leads to the most concise clusters. Hence, this work marks the initial step towards building a much larger and more fine-grained annotated dataset of medical radiology images.
RESUMO
Background: Artificial intelligence (AI) has led to the development of various opportunities during the COVID-19 pandemic. An abundant number of applications have surfaced responding to the pandemic, while some other applications were futile. Objectives: The present study aimed to assess the perception and opportunities of AI used during the COVID-19 pandemic and to explore the perception of medical data analysts about the inclusion of AI in medical education. Material and Methods: This study adopted a mixed-method research design conducted among medical doctors for the quantitative part while including medical data analysts for the qualitative interview. Results: The study reveals that nearly 64.8% of professionals were working in high COVID-19 patient-load settings and had significantly more acceptance of AI tools compared to others (P < 0.05). The learning barrier like engaging in new skills and working under a non-medical hierarchy led to dissatisfaction among medical data analysts. There was widespread recognition of their work after the COVID-19 pandemic. Conclusion: Notwithstanding that the majority of professionals are aware that public health emergency creates a significant strain on doctors, the majority still have to work in extremely high case load setting to demand solutions. AI applications are still not being integrated into medicine as fast as technology has been advancing. Sensitization workshops can be conducted among specialists to develop interest which will encourage them to identify problem statements in their fields, and along with AI experts, they can create AI-enabled algorithms to address the problems. A lack of educational opportunities about AI in formal medical curriculum was identified.
RESUMO
This work proposes a Blockchain-enabled Organ Matching System (BOMS) designed to manage the process of matching, storing, and sharing information. Biological factors are incorporated into matching and the cross-matching process is implemented into the smart contracts. Privacy is guaranteed by using patient-associated blockchain addresses, without transmitting or using patient personal records in the matching process. The matching algorithm implemented as a smart contract is verifiable by any party. Clinical records, process updates, and matching results are also stored on the blockchain, providing tamper-resistance of recipient's records and the recipients' waiting queue. The system also is capable of handling cases in which there is a donor without an immediate compatible recipient. The system is implemented on the Ethereum blockchain and several scenarios were tested. The performance of the proposed system is compared to other existing organ donation systems, and ours outperformed any existing organ matching system built on blockchain. BOMS is tested to ascertain its compatibility with public, private, and consortium blockchain networks, checks for security vulnerabilities and cross-matching efficiency. The implementation codes are available online.
Assuntos
Algoritmos , Blockchain , Obtenção de Tecidos e Órgãos , Humanos , Obtenção de Tecidos e Órgãos/métodos , Doadores de Tecidos , Segurança ComputacionalRESUMO
In this work, we present CodeAR, a medical time series generative model for electronic health record (EHR) synthesis. CodeAR employs autoregressive modeling on discrete tokens obtained using a vector quantized-variational autoencoder (VQ-VAE), which addresses key challenges of accurate distribution modeling and patient privacy preservation in the medical domain. The proposed model is trained with next-token prediction instead of a regression problem for more accurate distribution modeling, where the autoregressive property of CodeAR is useful to capture the inherent causality in time series data. In addition, the compressive property of the VQ-VAE prevents CodeAR from memorizing the original training data, which ensures patient privacy. Experimental results demonstrate that CodeAR outperforms the baseline autoregressive-based and GAN-based models in terms of maximum mean discrepancy (MMD) and Train on Synthetic, Test on Real tests. Our results highlight the effectiveness of autoregressive modeling on discrete tokens, the utility of CodeAR in causal modeling, and its robustness against data memorization.
Assuntos
Registros Eletrônicos de Saúde , Humanos , Fatores de Tempo , Análise de RegressãoRESUMO
The work elucidates the importance of accurate Parkinson's disease classification within medical diagnostics and introduces a novel framework for achieving this goal. Specifically, the study focuses on enhancing disease identification accuracy utilizing boosting methods. A standout contribution of this work lies in the utilization of a light gradient boosting machine (LGBM) coupled with hyperparameter tuning through grid search optimization (GSO) on the Parkinson's disease dataset derived from speech recording signals. In addition, the Synthetic Minority Over-sampling Technique (SMOTE) has also been employed as a pre-processing technique to balance the dataset, enhancing the robustness and reliability of the analysis. This approach is a novel addition to the study and underscores its potential to enhance disease identification accuracy. The datasets employed in this work include both gender-specific and combined cases, utilizing several distinctive feature subsets including baseline, Mel-frequency cepstral coefficients (MFCC), time-frequency, wavelet transform (WT), vocal fold, and tunable-Q-factor wavelet transform (TQWT). Comparative analyses against state-of-the-art boosting methods, such as AdaBoost and XG-Boost, reveal the superior performance of our proposed approach across diverse datasets and metrics. Notably, on the male cohort dataset, our method achieves exceptional results, demonstrating an accuracy of 0.98, precision of 1.00, sensitivity of 0.97, F1-Score of 0.98, and specificity of 1.00 when utilizing all features with GSO-LGBM. In comparison to AdaBoost and XGBoost, the proposed framework utilizing LGBM demonstrates superior accuracy, achieving an average improvement of 5% in classification accuracy across all feature subsets and datasets. These findings underscore the potential of the proposed methodology to enhance disease identification accuracy and provide valuable insights for further advancements in medical diagnostics.
RESUMO
Purpose: In 2024, medical researchers in the Republic of Korea were invited to amend the health and medical data utilization guidelines (Government Publications Registration Number: 11-1352000-0052828-14). This study aimed to show the overall impact of the guideline revision, with a focus on clinical genomic data. Materials and Methods: This study amended the pseudonymization of genomic data defined in the previous version through a joint study led by the Ministry of Health and Welfare, the Korea Health Information Service, and the Korea Genome Organization. To develop the previous version, we held three conferences with four main medical research institutes and seven academic societies. We conducted two surveys targeting special genome experts in academia, industry, and institutes. Results: We found that cases of pseudonymization in the application of genome data were rare and that there was ambiguity in the terminology used in the previous version of the guidelines. Most experts (> ~90%) agreed that the 'reserved' condition should be eliminated to make genomic data available after pseudonymization. In this study, the scope of genomic data was defined as clinical next generation sequencing data, including FASTQ, BAM/SAM, VCF, and medical records. Pseudonymization targets genomic sequences and metadata, embedding specific elements, such as germline mutations, short tandem repeats, single-nucleotide polymorphisms, and identifiable data (for example, ID or environmental values). Expression data generated from multi-omics can be used without pseudonymization. Conclusion: This amendment will not only enhance the safe use of healthcare data but also promote advancements in disease prevention, diagnosis, and treatment.
RESUMO
As healthcare systems transition into an era dominated by quantum technologies, the need to fortify cybersecurity measures to protect sensitive medical data becomes increasingly imperative. This paper navigates the intricate landscape of post-quantum cryptographic approaches and emerging threats specific to the healthcare sector. Delving into encryption protocols such as lattice-based, code-based, hash-based, and multivariate polynomial cryptography, the paper addresses challenges in adoption and compatibility within healthcare systems. The exploration of potential threats posed by quantum attacks and vulnerabilities in existing encryption standards underscores the urgency of a change in basic assumptions in healthcare data security. The paper provides a detailed roadmap for implementing post-quantum cybersecurity solutions, considering the unique challenges faced by healthcare organizations, including integration issues, budget constraints, and the need for specialized training. Finally, the abstract concludes with an emphasis on the importance of timely adoption of post-quantum strategies to ensure the resilience of healthcare data in the face of evolving threats. This roadmap not only offers practical insights into securing medical data but also serves as a guide for future directions in the dynamic landscape of post-quantum healthcare cybersecurity.
RESUMO
Objective: The significance of big data is increasingly acknowledged across all sectors, including medicine. Moreover, the trend of data trading is on the rise, particularly in exchanging other data for medical data to rejuvenate the medical industry. This study aimed to discern the facilitating factors of healthcare data trade. Methods: We assessed five medical data market platforms on October, 2022, based on three criteria: (a) clarity in articulating the data for sale; (b) transparency in specifying the data costs; and (c) explicit indication that payment grants data access. This helped identify the traded medical data types. Additionally, we anonymously surveyed 43 representatives from medical device companies about their demand for medical data trading, achieving a response rate of 66%. Results: Of the medical data traded on these platforms, 93.34% was structured, while 5.66% was unstructured, indicating an imbalance. Although there was a higher demand for structured medical data, there was also interest in purchasing unstructured medical data. Conclusion: Unstructured big data are crucial for medical device development, fueling the demand for trading such data. Many stakeholders view the data market as essential and are willing to procure medical data. Consequently, medical device companies will need methods to acquire unstructured medical data for developing innovative and enhanced medical devices.
RESUMO
In order to achieve the goals of the Medical Informatics Initiative (MII), staff with skills in the field of medical informatics and data science are required. Each consortium has established training activities. Further, cross-consortium activities have emerged. This article describes the concepts, implemented programs, and experiences in the consortia. Fifty-one new professorships have been established and 10 new study programs have been created: 1 bachelor's degree and 6 consecutive and 3 part-time master's degree programs. Further, learning and training opportunities can be used by all MII partners. Certification and recognition opportunities have been created.The educational offers are aimed at target groups with a background in computer science, medicine, nursing, bioinformatics, biology, natural science, and data science. Additional qualifications for physicians in computer science and computer scientists in medicine seem to be particularly important. They can lead to higher quality in software development and better support for treatment processes by application systems.Digital learning methods were important in all consortia. They offer flexibility for cross-location and interprofessional training. This enables learning at an individual pace and an exchange between professional groups.The success of the MII depends largely on society's acceptance of the multiple use of medical data in both healthcare and research. The information required for this is provided by the MII's public relations work. There is also an enormous need in society for medical and digital literacy.
Assuntos
Currículo , Informática Médica , Humanos , Segurança Computacional/normas , Registros Eletrônicos de Saúde/normas , Alemanha , Informática Médica/educação , Competência Profissional/normasRESUMO
BACKGROUND: Medical diagnosis in practice connects to research through continuous feedback loops: Studies of diagnosed cases shape our understanding of disease, which shapes future diagnostic practice. Without accounting for an imperfect and complex diagnostic process in which some cases are more likely to be diagnosed correctly (or diagnosed at all), the feedback loop can inadvertently exacerbate future diagnostic errors and biases. FRAMEWORK: A feedback loop failure occurs if misleading evidence about disease etiology encourages systematic errors that self-perpetuate, compromising future diagnoses and patient care. This article defines scenarios for feedback loop failure in medical diagnosis. DESIGN: Through simulated cases, we characterize how disease incidence, presentation, and risk factors can be misunderstood when observational data are summarized naive to biases arising from diagnostic error. A fourth simulation extends to a progressive disease. RESULTS: When severe cases of a disease are diagnosed more readily, less severe cases go undiagnosed, increasingly leading to underestimation of the prevalence and heterogeneity of the disease presentation. Observed differences in incidence and symptoms between demographic groups may be driven by differences in risk, presentation, the diagnostic process itself, or a combination of these. We suggested how perceptions about risk factors and representativeness may drive the likelihood of diagnosis. Differing diagnosis rates between patient groups can feed back to increasingly greater diagnostic errors and disparities in the timing of diagnosis and treatment. CONCLUSIONS: A feedback loop between past data and future medical practice may seem obviously beneficial. However, under plausible scenarios, poorly implemented feedback loops can degrade care. Direct summaries from observational data based on diagnosed individuals may be misleading, especially concerning those symptoms and risk factors that influence the diagnostic process itself. HIGHLIGHTS: Current evidence about a disease can (and should) influence the diagnostic process. A feedback loop failure may occur if biased "evidence" encourages diagnostic errors, leading to future errors in the evidence base.When diagnostic accuracy varies for mild versus severe cases or between demographic groups, incorrect conclusions about disease prevalence and presentation will result without specifically accounting for such variability.Use of demographic characteristics in the diagnostic process should be done with careful justification, in particular avoiding potential cognitive biases and overcorrection.