Home Baidu AI Empowers the Full Lifecycle of Medical Documentation, Revitalizing Medical Record Management

Baidu AI Empowers the Full Lifecycle of Medical Documentation, Revitalizing Medical Record Management

Dec 23, 2020 17:15 CST Updated 17:15

At the “5th Cross-Strait Forum on Medical Record Management” held on December 19, Baidu Lingyi Zhihui was invited to deliver a thematic presentation, offering full-cycle intelligent services to address the “pain points” of traditional hospital medical record management and help revitalize hospital medical record management.


image.png

Technological Innovation Drives the Intelligentization of Medical Records


Currently, there are multiple "pain points" in hospital medical record quality control: incomplete completion of the front page of medical records, inconsistency between clinical principal diagnoses and the selection principles for principal diagnoses in medical records, difficulty in controlling connotation-based quality control, and overly complex data statistics. These issues collectively hinder the development of healthcare institutions' informatization systems due to deficiencies in traditional medical record quality.


However, with the advancement of machine reading comprehension technologies, it has become feasible for computers to interpret medical records. Since the release of its ERNIE 1.0 model in 2019, Baidu’s independently developed natural language understanding technology has garnered numerous international accolades. Leveraging natural language understanding technology and an evidence-based AI medical knowledge graph, Baidu enables computers to comprehend medical documents with intelligent and precise understanding, thereby facilitating more efficient and comprehensive medical record documentation and achieving quality control tasks that are virtually impossible for humans to perform manually.


Compared with traditional information systems, Baidu Smart Medical Record can understand more medical records and provide more accurate suggestions for medical record writing with the help of AI.


image.png

Empowering the Full Lifecycle of Medical Documentation


Baidu has integrated management thinking into system development, offering intelligent service plugins for the entire lifecycle, thereby driving an intelligent upgrade in the quality control of medical record content. During the process of electronic medical record (EMR) documentation by clinicians, Baidu Smart Medical Records conducts comprehensive quality control of medical record content, providing real-time alerts to physicians. This helps mitigate risks associated with EMRs, such as inconsistencies in information descriptions, conflicts in causal relationships between preceding and subsequent data, and discrepancies between diagnoses or medical orders and patient demographics.


image.png


Baidu Smart Medical Records also provides an intelligent “one-click generation” tool for the medical record front sheet. Similar to automated abstract extraction for academic papers, this tool leverages Baidu’s natural language understanding technology to automatically generate the front sheet by analyzing the entire medical record. This helps liberate physicians from routine administrative tasks, allowing them to redirect the time previously spent on manual entry toward quality control of the medical record front sheet.


In terms of homepage quality control, Baidu Smart Medical Records has established a three-tier, six-category quality control capability centered around 35 types of standard medical document forms. This system covers over 400 global quality control checkpoints and provides automated machine-driven quality control services for both process-based and final-stage scenarios. By leveraging these automated quality control services, the drawbacks of traditional sampling-based quality control can be addressed, enhancing traditional information-based review capabilities by nearly 40%.


image.png


The service also provides auxiliary functions such as compliance with industry documentation standards, adherence to coding entry specifications, and consistency checks against the clinical content of medical records.


Notably, leveraging the role of medical records as payment vouchers, Baidu also offers intelligent services such as CHS-DRG-based pre-grouping and correction recommendations, intelligent coding, and diagnosis sequencing recommendations.


image.png

Seamlessly integrate with existing electronic medical record systems


Adhering to the philosophy of openness and mutual benefit, Baidu’s smart medical record solutions can seamlessly integrate with hospitals’ existing electronic medical record (EMR) systems. This approach delivers value-added services to current health IT vendors and provides micro-solutions tailored for vendors to encourage the adoption of intelligent capabilities.


The Smart Medical Record Service supports private deployment and offers diverse service models, including API-based, componentized, integrated, and customized options. It features simple installation, user-friendly operation, and continuously expanding professional AI capabilities and documentation empowerment scenarios. Furthermore, the service fully opens its core components to partners, enabling standardized service delivery. This facilitates a unified approach to hospital informatization, allowing for the joint development of comprehensive solutions with healthcare IT vendors to collectively drive the practical implementation of intelligent medical documentation.