Data Governance in the Health Industry: Investigating Data Quality Dimensions within a Big Data Context

The METRIC-framework for assessing data quality for trustworthy AI in medicine: a systematic review

The adoption of machine learning (ML) and, more specifically, deep learning (DL) applications into all major areas of our lives is underway. The development of trustworthy AI is especially important in medicine due to the large implications for patients' lives. While trustworthiness concerns various aspects including ethical, transparency and safety requirements, we focus on the importance of data quality (training/test) in DL. Since data quality dictates the behaviour of ML products, evaluating data quality will play a key part in the regulatory approval of medical ML products. We perform a systematic review following PRISMA guidelines using the databases Web of Science, PubMed and ACM Digital Library. We identify 5408 studies, out of which 120 records fulfil our eligibility criteria. From this literature, we synthesise the existing knowledge on data quality frameworks and combine it with the perspective of ML applications in medicine. As a result, we propose the METRIC-framework, a specialised data quality framework for medical training data comprising 15 awareness dimensions, along which developers of medical ML applications should investigate the content of a dataset. This knowledge helps to reduce biases as a major source of unfairness, increase robustness, facilitate interpretability and thus lays the foundation for trustworthy AI in medicine. The METRIC-framework may serve as a base for systematically assessing training datasets, establishing reference datasets, and designing test datasets which has the potential to accelerate the approval of medical ML products.

Analysis of Traditional Chinese Medicine Symptoms in Children with Spastic Cerebral Palsy: A Data Mining Study

Background

Cerebral palsy (CP) ranks as a major cause of motor disabilities in children, with spastic CP making up roughly 70-80% of all CP cases. The primary objective of our study is to identify characteristics of Traditional Chinese Medicine(TCM) symptom of spastic CP, thereby establishing correlations between the TCM symptom and the disease, providing a more scientific theoretical foundation for TCM treatments on spastic CP, enabling a deeper comprehension of clinical interventions, ultimately, improving rehabilitation outcomes in TCM treatment for spastic CP.

Methods

We conducted a data mining study on TCM symptom of spastic CP children aged 4-14 years old treated at Xi'an Encephalopathy Hospital Affiliated to Shaanxi University of Chinese Medicine, from October 2021 to March 2023. The medical records of all eligible and complete spastic CP patients were extracted, processed for data cleansing, transformed, and subsequently analyzed to discern distinctive TCM symptom. K-Means Clustering Analysis and Association Rule Analysis were used for data mining.

Results

Core symptoms identified for spastic CP encompassed "Motor Dysfunction", "Impaired Speech", "Delayed Development", "Limb Stiffness", "Rigidity in the limbs", "Intellectual Impairment", "Timidity and susceptibility to startle responses", "Muscle Wasting", and "Pale or Dull Complexion". Among the top-ranking associations of symptom, patterns emerge wherein "Motor dysfunction" intertwine with "Impaired speech", "Motor dysfunction" coexist with "Delayed development", and "Impaired speech" are accompanied by "Delayed development".

Conclusion

This study identified the core symptom of spastic CP and tentatively suggests that the clinical manifestations of spastic CP are essentially consistent with the TCM pattern "liver exuberance and spleen weakness". This finding has facilitated the preliminary establishment of correlations between TCM pattern differentiation and the disease in medicine. It is anticipated that this correlation will bring tangible benefits to a larger number of children with spastic CP.

Big Data Analytics and Processing Platform in Czech Republic Healthcare

Big data analytics (BDA) in healthcare has made a positive difference in the integration of Artificial Intelligence (AI) in advancements of analytical capabilities, while lowering the costs of medical care. The aim of this study is to improve the existing healthcare eSystem by implementing a Big Data Analytics (BDA) platform and to meet the requirements of the Czech Republic National Health Service (Tender-Id. VZ0036628, No. Z2017-035520). In addition to providing analytical capabilities on Linux platforms supporting current and near-future AI with machine-learning and data-mining algorithms, there is the need for ethical considerations mandating new ways to preserve privacy, all of which are preconditioned by the growing body of regulations and expectations. The presented BDA platform, has met all requirements (N > 100), including the healthcare industry-standard Transaction Processing Performance Council (TPC-H) decision support benchmark in compliance with the European Union (EU) and the Czech Republic legislations. Currently, the presented Proof of Concept (PoC) that has been upgraded to a production environment has unified isolated parts of Czech Republic healthcare over the past seven months. The reported PoC BDA platform, artefacts, and concepts are transferrable to healthcare systems in other countries interested in developing or upgrading their own national healthcare infrastructure in a cost-effective, secure, scalable and high-performance manner.