An electronic medical record system with treatment recommendations based on patient similarity

Language inference-based learning for Low-Resource Chinese clinical named entity recognition using language model

Electronic health records (EHRs) have been widely used and are gradually replacing paper records. Therefore, extracting valuable information from EHRs has become the focus and hotspot of current research. Clinical named entity recognition (CNER) is an important task in information extraction. Most current research methods used standard supervised learning approaches to fine-tune pre-trained language models (PLMs), which require a large amount of annotated data for model training. However, in realistic medical scenarios, annotated data are scarce, especially in the healthcare field. The process of annotating data in real clinical settings is time-consuming and labour-intensive. In this paper, a language inference-based learning method (LANGIL) is proposed to study clinical NER tasks with limited annotated samples, i.e., in low-resource clinical scenarios. A method based on prompt learning is designed to reformulate the entity recognition task into a language inference-based task. Differing from the standard fine-tuning method, the approach introduced in this paper does not design the additional network layers that train from scratch. This alleviates the gap between pre-training tasks and downstream tasks, allowing the comprehension capabilities of PLMs to be leveraged under the condition of limited training samples. The experiments on four Chinese clinical named entity recognition datasets showed that LANGIL achieves significant improvements in F1-score compared to the former method.

Double chain system for online and offline medical data sharing private and consortium blockchain: A system design study

With the informatization development and digital construction in the healthcare industry, electronic medical records and Internet medicine facilitate people's medical treatment. However, the current data storage method has the risk of data loss, leakage, and tampering, and can't support extensive and secure sharing of medical data. To realize effective and secure medical data storage and sharing among offline medical institutions and Internet medicine platforms, this study used a combined private blockchain and consortium blockchain to design a medical blockchain double-chain system (MBDS). This system can store encrypted medical data in distributed storage mode and systematically integrate the medical data of patients in offline medical institutions and Internet medicine platforms, to achieve equality, credibility, and data sharing among participating nodes. The MBDS system constructed in this study incorporated Internet medicine care services into the current healthcare system and provided new solutions and practical guidance for the future development of collaborative medical care. This study helped to solve the problems of medical data interconnection and resource sharing, improve the efficiency and effect of disease diagnosis, alleviate the contradiction between doctors and patients, and facilitate personal health management. This study has substantial theoretical and practical implications for the research and application of medical data storage and sharing.

El-Kassabi H, Serhani MA, Oulhaj A, Khalil K. A Novel Patient Similarity Network (PSN) Framework Based on Multi-Model Deep Learning for Precision Medicine

Precision medicine can be defined as the comparison of a new patient with existing patients that have similar characteristics and can be referred to as patient similarity. Several deep learning models have been used to build and apply patient similarity networks (PSNs). However, the challenges related to data heterogeneity and dimensionality make it difficult to use a single model to reduce data dimensionality and capture the features of diverse data types. In this paper, we propose a multi-model PSN that considers heterogeneous static and dynamic data. The combination of deep learning models and PSN allows ample clinical evidence and information extraction against which similar patients can be compared. We use the bidirectional encoder representations from transformers (BERT) to analyze the contextual data and generate word embedding, where semantic features are captured using a convolutional neural network (CNN). Dynamic data are analyzed using a long-short-term-memory (LSTM)-based autoencoder, which reduces data dimensionality and preserves the temporal features of the data. We propose a data fusion approach combining temporal and clinical narrative data to estimate patient similarity. The experiments we conducted proved that our model provides a higher classification accuracy in determining various patient health outcomes when compared with other traditional classification algorithms.

Interaction Time with Electronic Health Records: A Systematic Review

BACKGROUND

The amount of time that health care clinicians (physicians and nurses) spend interacting with the electronic health record is not well understood.

OBJECTIVE

This study aimed to evaluate the time that health care providers spend interacting with electronic health records (EHR).

METHODS

Data are retrieved from Ovid MEDLINE(R) and Epub Ahead of Print, In-Process and Other Non-Indexed Citations and Daily, (Ovid) Embase, CINAHL, and SCOPUS.

STUDY ELIGIBILITY CRITERIA

Peer-reviewed studies that describe the use of EHR and include measurement of time either in hours, minutes, or in the percentage of a clinician's workday. Papers were written in English and published between 1990 and 2021.

PARTICIPANTS

All physicians and nurses involved in inpatient and outpatient settings.

STUDY APPRAISAL AND SYNTHESIS METHODS

A narrative synthesis of the results, providing summaries of interaction time with EHR. The studies were rated according to Quality Assessment Tool for Studies with Diverse Designs.

RESULTS

Out of 5,133 de-duplicated references identified through database searching, 18 met inclusion criteria. Most were time-motion studies (50%) that followed by logged-based analysis (44%). Most were conducted in the United States (94%) and examined a clinician workflow in the inpatient settings (83%). The average time was nearly 37% of time of their workday by physicians in both inpatient and outpatient settings and 22% of the workday by nurses in inpatient settings. The studies showed methodological heterogeneity.

CONCLUSION

This systematic review evaluates the time that health care providers spend interacting with EHR. Interaction time with EHR varies depending on clinicians' roles and clinical settings, computer systems, and users' experience. The average time spent by physicians on EHR exceeded one-third of their workday. The finding is a possible indicator that the EHR has room for usability, functionality improvement, and workflow optimization.

Patient Representation From Structured Electronic Medical Records Based on Embedding Technique: Development and Validation Study

Background

The secondary use of structured electronic medical record (sEMR) data has become a challenge due to the diversity, sparsity, and high dimensionality of the data representation. Constructing an effective representation for sEMR data is becoming more and more crucial for subsequent data applications.

Objective

We aimed to apply the embedding technique used in the natural language processing domain for the sEMR data representation and to explore the feasibility and superiority of the embedding-based feature and patient representations in clinical application.

Methods

The entire training corpus consisted of records of 104,752 hospitalized patients with 13,757 medical concepts of disease diagnoses, physical examinations and procedures, laboratory tests, medications, etc. Each medical concept was embedded into a 200-dimensional real number vector using the Skip-gram algorithm with some adaptive changes from shuffling the medical concepts in a record 20 times. The average of vectors for all medical concepts in a patient record represented the patient. For embedding-based feature representation evaluation, we used the cosine similarities among the medical concept vectors to capture the latent clinical associations among the medical concepts. We further conducted a clustering analysis on stroke patients to evaluate and compare the embedding-based patient representations. The Hopkins statistic, Silhouette index (SI), and Davies-Bouldin index were used for the unsupervised evaluation, and the precision, recall, and F1 score were used for the supervised evaluation.

Results

The dimension of patient representation was reduced from 13,757 to 200 using the embedding-based representation. The average cosine similarity of the selected disease (subarachnoid hemorrhage) and its 15 clinically relevant medical concepts was 0.973. Stroke patients were clustered into two clusters with the highest SI (0.852). Clustering analyses conducted on patients with the embedding representations showed higher applicability (Hopkins statistic 0.931), higher aggregation (SI 0.862), and lower dispersion (Davies-Bouldin index 0.551) than those conducted on patients with reference representation methods. The clustering solutions for patients with the embedding-based representation achieved the highest F1 scores of 0.944 and 0.717 for two clusters.

Conclusions

The feature-level embedding-based representations can reflect the potential clinical associations among medical concepts effectively. The patient-level embedding-based representation is easy to use as continuous input to standard machine learning algorithms and can bring performance improvements. It is expected that the embedding-based representation will be helpful in a wide range of secondary uses of sEMR data.

Using Domain Knowledge and Data-Driven Insights for Patient Similarity Analytics

Patient similarity analytics has emerged as an essential tool to identify cohorts of patients who have similar clinical characteristics to some specific patient of interest. In this study, we propose a patient similarity measure called D3K that incorporates domain knowledge and data-driven insights. Using the electronic health records (EHRs) of 169,434 patients with either diabetes, hypertension or dyslipidaemia (DHL), we construct patient feature vectors containing demographics, vital signs, laboratory test results, and prescribed medications. We discretize the variables of interest into various bins based on domain knowledge and make the patient similarity computation to be aligned with clinical guidelines. Key findings from this study are: (1) D3K outperforms baseline approaches in all seven sub-cohorts; (2) our domain knowledge-based binning strategy outperformed the traditional percentile-based binning in all seven sub-cohorts; (3) there is substantial agreement between D3K and physicians (κ = 0.746), indicating that D3K can be applied to facilitate shared decision making. This is the first study to use patient similarity analytics on a cardiometabolic syndrome-related dataset sourced from medical institutions in Singapore. We consider patient similarity among patient cohorts with the same medical conditions to develop localized models for personalized decision support to improve the outcomes of a target patient.

Security challenges and solutions using healthcare cloud computing

Cloud computing is among the most beneficial solutions to digital problems. Security is one of the focal issues in cloud computing technology, and this study aims at investigating security issues of cloud computing and their probable solutions. A systematic review was performed using Scopus, Pubmed, Science Direct, and Web of Science databases. Once the title and abstract were evaluated, the quality of studies was assessed in order to choose the most relevant according to exclusion and inclusion criteria. Then, the full texts of studies selected were read thoroughly to extract the necessary results. According to the review, data security, availability, and integrity, as well as information confidentiality and network security, were the major challenges in cloud security. Further, data encryption, authentication, and classification, besides application programming interfaces (API), were security solutions to cloud infrastructure. Data encryption could be applied to store and retrieve data from the cloud in order to provide secure communication. Besides, several central challenges, which make the cloud security engineering process problematic, have been considered in this study.

Instance-Based Learning Following Physician Reasoning for Assistance during Medical Consultation

This article presents an automatic system for modeling clinical knowledge to follow a physician’s reasoning in medical consultation. Instance-based learning is applied to provide suggestions when recording electronic medical records. The system was validated on a real case study involving advanced medical students. The proposed system is accurate and efficient: 2.5× more efficient than a baseline empirical tool for suggestions and two orders of magnitude faster than a Bayesian learning method, when processing a testbed of 250 clinical case types. The research provides a framework to implement a real-time system to assist physicians during medical consultations.

Use of artificial intelligence in infectious diseases

Infectious diseases are caused by microorganisms belonging to the class of bacteria, viruses, fungi, or parasites. These pathogens are transmitted, directly or indirectly, and can lead to epidemics or even pandemics. The resulting infection may lead to mild-to-severe symptoms such as life-threatening fever or diarrhea. Infectious diseases may be asymptomatic in some individuals but may lead to disastrous effects in others. Despite the advances in medicine, infectious diseases are a leading cause of death worldwide, especially in low-income countries. With the advent of mathematical tools, scientists are now able to better predict epidemics, understand the specificity of each pathogen, and identify potential targets for drug development. Artificial intelligence and its components have been widely publicized for their ability to better diagnose certain types of cancer from imaging data. This chapter aims at identifying potential applications of machine learning in the field of infectious diseases. We are deliberately focusing on key aspects of infection: diagnosis, transmission, response to treatment, and resistance. We are proposing the use of extreme values as an avenue of interest for future developments in the field of infectious diseases. This chapter covers a series of applications selectively chosen to showcase how artificial intelligence is moving the field of infectious disease further and how it helps institutions to better tackles them, especially in low-income countries.

Evaluating global and local sequence alignment methods for comparing patient medical records

Background

Sequence alignment is a way of arranging sequences (e.g., DNA, RNA, protein, natural language, financial data, or medical events) to identify the relatedness between two or more sequences and regions of similarity. For Electronic Health Records (EHR) data, sequence alignment helps to identify patients of similar disease trajectory for more relevant and precise prognosis, diagnosis and treatment of patients.

Methods

We tested two cutting-edge global sequence alignment methods, namely dynamic time warping (DTW) and Needleman-Wunsch algorithm (NWA), together with their local modifications, DTW for Local alignment (DTWL) and Smith-Waterman algorithm (SWA), for aligning patient medical records. We also used 4 sets of synthetic patient medical records generated from a large real-world EHR database as gold standard data, to objectively evaluate these sequence alignment algorithms.

Results

For global sequence alignments, 47 out of 80 DTW alignments and 11 out of 80 NWA alignments had superior similarity scores than reference alignments while the rest 33 DTW alignments and 69 NWA alignments had the same similarity scores as reference alignments. Forty-six out of 80 DTW alignments had better similarity scores than NWA alignments with the rest 34 cases having the equal similarity scores from both algorithms. For local sequence alignments, 70 out of 80 DTWL alignments and 68 out of 80 SWA alignments had larger coverage and higher similarity scores than reference alignments while the rest DTWL alignments and SWA alignments received the same coverage and similarity scores as reference alignments. Six out of 80 DTWL alignments showed larger coverage and higher similarity scores than SWA alignments. Thirty DTWL alignments had the equal coverage but better similarity scores than SWA. DTWL and SWA received the equal coverage and similarity scores for the rest 44 cases.

Conclusions

DTW, NWA, DTWL and SWA outperformed the reference alignments. DTW (or DTWL) seems to align better than NWA (or SWA) by inserting new daily events and identifying more similarities between patient medical records. The evaluation results could provide valuable information on the strengths and weakness of these sequence alignment methods for future development of sequence alignment methods and patient similarity-based studies.

SuperOrder: Provider order recommendation system for outpatient clinics

This study aims at developing SuperOrder, an order recommendation system for outpatient clinics. Using the electronic health record data available at midnight, SuperOrder predicts the order contents for each upcoming appointment on a daily basis. A two-level prediction framework is proposed. At the base-level, the predictions are produced by aggregating three machine learning methods. The meta-level predictions are generated by integrating the base-level predictions with the order co-occurrence network. We used the retrospective data between 1 April 2014 and 31 March 2015 in pulmonary clinics from five hospital sites within a large rural health care facility in Pennsylvania to test the feasibility. With a decrease of 6 per cent in the precision, the improvement of the recall at the meta-level is approximately 20 per cent from the base-level. This demonstrates that the proposed order co-occurrence network helps in increasing the performance of order predictions. The implementation will bring a more effective and efficient way to place outpatient orders.

Blockchain-Based Medical Records Secure Storage and Medical Service Framework

Accurate and complete medical data are one valuable asset for patients. Privacy protection and the secure storage of medical data are crucial issues during medical services. Secure storage and making full use of personal medical records has always been a concern for the general population. The emergence of blockchain technology brings a new idea to solve this problem. As a hash chain with the characteristics of decentralization, verifiability and immutability, blockchain technology can be used to securely store personal medical data. In this paper, we design a storage scheme to manage personal medical data based on blockchain and cloud storage. Furthermore, a service framework for sharing medical records is described. In addition, the characteristics of the medical blockchain are presented and analyzed through a comparison with traditional systems. The proposed storage and sharing scheme does not depend on any third-party and no single party has absolute power to affect the processing.

Best-practice care pathway for improving management of mastitis and breast abscess

BACKGROUND

Surgical subspecialization has resulted in mastitis and breast abscesses being managed with unnecessary admission to hospital, prolonged inpatient stay, variable antibiotic prescribing, incision and drainage rather than percutaneous aspiration, and loss to specialist follow-up. The objective was to evaluate a best-practice algorithm with the aim of improving management of mastitis and breast abscesses across a multisite NHS Trust. The focus was on uniformity of antibiotic prescribing, ultrasound assessment, admission rates, length of hospital stay, intervention by aspiration or incision and drainage, and specialist follow-up.

METHODS

Management was initially evaluated in a retrospective cohort (phase I) and subsequently compared with that in two prospective cohorts after introduction of a breast abscess and mastitis pathway. One prospective cohort was analysed immediately after introduction of the pathway (phase II), and the second was used to assess the sustainability of the quality improvements (phase III). The overall impact of the pathway was assessed by comparing data from phase I with combined data from phases II and III; results from phases II and III were compared to judge sustainability.

RESULTS

Fifty-three patients were included in phase I, 61 in phase II and 80 in phase III. The management pathway and referral pro forma improved compliance with antibiotic guidelines from 34 per cent to 58·2 per cent overall (phases II and III) after implementation (P = 0·003). The improvement was maintained between phases II and III (54 and 61 per cent respectively; P = 0·684). Ultrasound assessment increased from 38 to 77·3 per cent overall (P < 0·001), in a sustained manner (75 and 79 per cent in phases II and III respectively; P = 0·894). Reductions in rates of incision and drainage (from 8 to 0·7 per cent overall; P = 0·007) were maintained (0 per cent in phase II versus 1 per cent in phase III; P = 0·381). Specialist follow-up improved consistently from 43 to 95·7 per cent overall (P < 0·001), 92 per cent in phase II and 99 per cent in phase III (P = 0·120). Rates of hospital admission and median length of stay were not significantly reduced after implementation of the pathway.

CONCLUSION

A standardized approach to mastitis and breast abscess reduced undesirable practice variation, with sustained improvements in process and patient outcomes.

Patient representation learning and interpretable evaluation using clinical notes

We have three contributions in this work: 1. We explore the utility of a stacked denoising autoencoder and a paragraph vector model to learn task-independent dense patient representations directly from clinical notes. To analyze if these representations are transferable across tasks, we evaluate them in multiple supervised setups to predict patient mortality, primary diagnostic and procedural category, and gender. We compare their performance with sparse representations obtained from a bag-of-words model. We observe that the learned generalized representations significantly outperform the sparse representations when we have few positive instances to learn from, and there is an absence of strong lexical features. 2. We compare the model performance of the feature set constructed from a bag of words to that obtained from medical concepts. In the latter case, concepts represent problems, treatments, and tests. We find that concept identification does not improve the classification performance. 3. We propose novel techniques to facilitate model interpretability. To understand and interpret the representations, we explore the best encoded features within the patient representations obtained from the autoencoder model. Further, we calculate feature sensitivity across two networks to identify the most significant input features for different classification tasks when we use these pretrained representations as the supervised input. We successfully extract the most influential features for the pipeline using this technique.

An ontology-based approach to patient follow-up assessment for continuous and personalized chronic disease management

OBJECTIVE

Chronic diseases are complex and persistent clinical conditions that require close collaboration among patients and health care providers in the implementation of long-term and integrated care programs. However, current solutions focus partially on intensive interventions at hospitals rather than on continuous and personalized chronic disease management. This study aims to fill this gap by providing computerized clinical decision support during follow-up assessments of chronically ill patients at home.

METHODS

We proposed an ontology-based framework to integrate patient data, medical domain knowledge, and patient assessment criteria for chronic disease patient follow-up assessments. A clinical decision support system was developed to implement this framework for automatic selection and adaptation of standard assessment protocols to suit patient personal conditions. We evaluated our method in the case study of type 2 diabetic patient follow-up assessments.

RESULTS

The proposed framework was instantiated using real data from 115,477 follow-up assessment records of 36,162 type 2 diabetic patients. Standard evaluation criteria were automatically selected and adapted to the particularities of each patient. Assessment results were generated as a general typing of patient overall condition and detailed scoring for each criterion, providing important indicators to the case manager about possible inappropriate judgments, in addition to raising patient awareness of their disease control outcomes. Using historical data as the gold standard, our system achieved a rate of accuracy of 99.93% and completeness of 95.00%.

CONCLUSIONS

This study contributes to improving the accessibility, efficiency and quality of current patient follow-up services. It also provides a generic approach to knowledge sharing and reuse for patient-centered chronic disease management.

Dense Annotation of Free-Text Critical Care Discharge Summaries from an Indian Hospital and Associated Performance of a Clinical NLP Annotator

Electronic Health Record (EHR) use in India is generally poor, and structured clinical information is mostly lacking. This work is the first attempt aimed at evaluating unstructured text mining for extracting relevant clinical information from Indian clinical records. We annotated a corpus of 250 discharge summaries from an Intensive Care Unit (ICU) in India, with markups for diseases, procedures, and lab parameters, their attributes, as well as key demographic information and administrative variables such as patient outcomes. In this process, we have constructed guidelines for an annotation scheme useful to clinicians in the Indian context. We evaluated the performance of an NLP engine, Cocoa, on a cohort of these Indian clinical records. We have produced an annotated corpus of roughly 90 thousand words, which to our knowledge is the first tagged clinical corpus from India. Cocoa was evaluated on a test corpus of 50 documents. The overlap F-scores across the major categories, namely disease/symptoms, procedures, laboratory parameters and outcomes, are 0.856, 0.834, 0.961 and 0.872 respectively. These results are competitive with results from recent shared tasks based on US records. The annotated corpus and associated results from the Cocoa engine indicate that unstructured text mining is a viable method for cohort analysis in the Indian clinical context, where structured EHR records are largely absent.