SCOPE: predicting future diagnoses in office visits using electronic health records.

We propose an interpretable and scalable model to predict likely diagnoses at an encounter based on past diagnoses and lab results. This model is intended to aid physicians in their interaction with the electronic health records (EHR). To accomplish this, we retrospectively collected and de-identified EHR data of 2,701,522 patients at Stanford Healthcare over a time period from January 2008 to December 2016. A population-based sample of patients comprising 524,198 individuals (44% M, 56% F) with multiple encounters with at least one frequently occurring diagnosis codes were chosen. A calibrated model was developed to predict ICD-10 diagnosis codes at an encounter based on the past diagnoses and lab results, using a binary relevance based multi-label modeling strategy. Logistic regression and random forests were tested as the base classifier, and several time windows were tested for aggregating the past diagnoses and labs. This modeling approach was compared to a recurrent neural network based deep learning method. The best model used random forest as the base classifier and integrated demographic features, diagnosis codes, and lab results. The best model was calibrated and its performance was comparable or better than existing methods in terms of various metrics, including a median AUROC of 0.904 (IQR [0.838, 0.954]) over 583 diseases. When predicting the first occurrence of a disease label for a patient, the median AUROC with the best model was 0.796 (IQR [0.737, 0.868]). Our modeling approach performed comparably as the tested deep learning method, outperforming it in terms of AUROC (p < 0.001) but underperforming in terms of AUPRC (p < 0.001). Interpreting the model showed that the model uses meaningful features and highlights many interesting associations among diagnoses and lab results. We conclude that the multi-label model performs comparably with RNN based deep learning model while offering simplicity and potentially superior interpretability. While the model was trained and validated on data obtained from a single institution, its simplicity, interpretability and performance makes it a promising candidate for deployment.

Practical Considerations for Developing Clinical Natural Language Processing Systems for Population Health Management and Measurement.

Experts have noted a concerning gap between clinical natural language processing (NLP) research and real-world applications, such as clinical decision support. To help address this gap, in this viewpoint, we enumerate a set of practical considerations for developing an NLP system to support real-world clinical needs and improve health outcomes. They include determining (1) the readiness of the data and compute resources for NLP, (2) the organizational incentives to use and maintain the NLP systems, and (3) the feasibility of implementation and continued monitoring. These considerations are intended to benefit the design of future clinical NLP projects and can be applied across a variety of settings, including large health systems or smaller clinical practices that have adopted electronic medical records in the United States and globally.

Development of a Natural Language Processing System for Extracting Rheumatoid Arthritis Outcomes From Clinical Notes Using the National Rheumatology Informatics System for Effectiveness Registry.

OBJECTIVE: To accelerate the use of outcome measures in rheumatology, we developed and evaluated a natural language processing (NLP) pipeline for extracting these measures from free-text outpatient rheumatology notes within the American College of Rheumatology's Rheumatology Informatics System for Effectiveness (RISE) registry. METHODS: We included all patients in RISE (2015-2018). The NLP pipeline extracted scores corresponding to 8 measures of rheumatoid arthritis (RA) disease activity (DA) and functional status (FS) documented in outpatient rheumatology notes. Score extraction performance was evaluated by chart review, and we assessed agreement with scores documented in structured data. We conducted an external validation of our NLP pipeline using data from rheumatology notes from an academic medical center that is not included in the RISE registry. RESULTS: We processed over 34 million notes from 854,628 patients, 158 practices, and 24 electronic health record (EHR) systems from RISE. Manual chart review revealed a sensitivity, positive predictive value (PPV), and F1 score of 95%, 87%, and 91%, respectively. Substantial agreement was observed between scores extracted from RISE notes and scores derived from structured data (κ = 0.43-0.68 among DA and 0.86-0.98 among FS measures). In the external validation, we found a sensitivity, PPV, and F1 score of 92%, 69%, and 79%, respectively. CONCLUSION: We developed an NLP pipeline to extract RA outcome measures from a national registry of notes from multiple EHR systems and found it to have good internal and external validity. This pipeline can facilitate measurement of clinical- and patient-reported outcomes for use in research and quality measurement.

Development and validation of MedDRA Tagger: a tool for extraction and structuring medical information from clinical notes.

Rapid and automated extraction of clinical information from patients' notes is a desirable though difficult task. Natural language processing (NLP) and machine learning have great potential to automate and accelerate such applications, but developing such models can require a large amount of labeled clinical text, which can be a slow and laborious process. To address this gap, we propose the MedDRA tagger, a fast annotation tool that makes use of industrial level libraries such as spaCy, biomedical ontologies and weak supervision to annotate and extract clinical concepts at scale. The tool can be used to annotate clinical text and obtain labels for training machine learning models and further refine the clinical concept extraction performance, or to extract clinical concepts for observational study purposes. To demonstrate the usability and versatility of our tool, we present three different use cases: we use the tagger to determine patients with a primary brain cancer diagnosis, we show evidence of rising mental health symptoms at the population level and our last use case shows the evolution of COVID-19 symptomatology throughout three waves between February 2020 and October 2021. The validation of our tool showed good performance on both specific annotations from our development set (F1 score 0.81) and open source annotated data set (F1 score 0.79). We successfully demonstrate the versatility of our pipeline with three different use cases. Finally, we note that the modular nature of our tool allows for a straightforward adaptation to another biomedical ontology. We also show that our tool is independent of EHR system, and as such generalizable.

Strategies to Address the Lack of Labeled Data for Supervised Machine Learning Training With Electronic Health Records: Case Study for the Extraction of Symptoms From Clinical Notes.

BACKGROUND: Automated extraction of symptoms from clinical notes is a challenging task owing to the multidimensional nature of symptom description. The availability of labeled training data is extremely limited owing to the nature of the data containing protected health information. Natural language processing and machine learning to process clinical text for such a task have great potential. However, supervised machine learning requires a great amount of labeled data to train a model, which is at the origin of the main bottleneck in model development. OBJECTIVE: The aim of this study is to address the lack of labeled data by proposing 2 alternatives to manual labeling for the generation of training labels for supervised machine learning with English clinical text. We aim to demonstrate that using lower-quality labels for training leads to good classification results. METHODS: We addressed the lack of labels with 2 strategies. The first approach took advantage of the structured part of electronic health records and used diagnosis codes (International Classification of Disease-10th revision) to derive training labels. The second approach used weak supervision and data programming principles to derive training labels. We propose to apply the developed framework to the extraction of symptom information from outpatient visit progress notes of patients with cardiovascular diseases. RESULTS: We used >500,000 notes for training our classification model with International Classification of Disease-10th revision codes as labels and >800,000 notes for training using labels derived from weak supervision. We show that the dependence between prevalence and recall becomes flat provided a sufficiently large training set is used (>500,000 documents). We further demonstrate that using weak labels for training rather than the electronic health record codes derived from the patient encounter leads to an overall improved recall score (10% improvement, on average). Finally, the external validation of our models shows excellent predictive performance and transferability, with an overall increase of 20% in the recall score. CONCLUSIONS: This work demonstrates the power of using a weak labeling pipeline to annotate and extract symptom mentions in clinical text, with the prospects to facilitate symptom information integration for a downstream clinical task such as clinical decision support.

Structuring clinical text with AI: Old versus new natural language processing techniques evaluated on eight common cardiovascular diseases.

Free-text clinical notes in electronic health records are more difficult for data mining while the structured diagnostic codes can be missing or erroneous. To improve the quality of diagnostic codes, this work extracts diagnostic codes from free-text notes: five old and new word vectorization methods were used to vectorize Stanford progress notes and predict eight ICD-10 codes of common cardiovascular diseases with logistic regression. The models showed good performance, with TF-IDF as the best vectorization model showing the highest AUROC (0.9499-0.9915) and AUPRC (0.2956-0.8072). The models also showed transferability when tested on MIMIC-III data with AUROC from 0.7952 to 0.9790 and AUPRC from 0.2353 to 0.8084. Model interpretability was shown by the important words with clinical meanings matching each disease. This study shows the feasibility of accurately extracting structured diagnostic codes, imputing missing codes, and correcting erroneous codes from free-text clinical notes for information retrieval and downstream machine-learning applications.

Fluorescence quantum yield rationalized by the magnitude of the charge transfer in π-conjugated terpyridine derivatives.

Terpyridine derivatives are of great interest due to their unique photophysical properties when used as antennas in metallic complexes. Several experimental and theoretical studies indicate strong charge-transfer character of the lowest electronic excited state, which could be exploited for predicting fluorescence quantum yields from the magnitude of the charge separation induced by electronic transitions. Focusing on substituted 4'-phenyl-2,2':6'2''-terpyridyl, we report on two measures of the charge separation obtained from high-level calculations in ground and excited states (length of the change of the dipole moment and the electron-hole distance). Our refined model confirms that the fluorescence quantum yield shows a global S-shape dependence on the magnitude of the charge separation, which can be quantified either by the change in dipole moments between the ground and excited states or by the associated charge-hole distances. This approach provides a remarkable tool for the molecular design of a fluorescent polyaromatic antenna.

Looking for the Origin of Allosteric Cooperativity in Metallopolymers.

The basic concept of allosteric cooperativity used in biology, chemistry and physics states that any change in the intermolecular host-guest interactions operating in multisite receptors can be assigned to intersite interactions. Using lanthanide metals as guests and linear multi-tridentate linear oligomers of variable lengths and geometries as hosts, this work shows that the quantitative modeling of metal loadings requires the consideration of a novel phenomenon originating from solvation processes. It stepwise modulates the intrinsic affinity of each isolated site in multisite receptors, and this without resorting to allosteric cooperativity. An easy-to-handle additive model predicts a negative power law dependence of the intrinsic affinity on the length of the linear metallopolymer. Applied to lanthanidopolymers, the latter common analysis overestimates cooperativity factors by more than two orders of magnitude.

Ion Pair-π Interactions.

We report that anion-π and cation-π interactions can occur on the same aromatic surface. Interactions of this type are referred to as ion pair-π interactions. Their existence, nature, and significance are elaborated in the context of spectral tuning, ion binding in solution, and activation of cell-penetrating peptides. The origin of spectral tuning by ion pair-π interactions is unraveled with energy-minimized excited-state structures: The solvent- and pH-independent red shift of absorption and emission of push-pull fluorophores originates from antiparallel ion pair-π attraction to their polarized excited state. In contrast, the complementary parallel ion pair-π repulsion is spectroscopically irrelevant, in part because of charge neutralization by intriguing proton and electron transfers on excited push-pull surfaces. With time-resolved fluorescence measurements, very important differences between antiparallel and parallel ion pair-π interactions are identified and quantitatively dissected from interference by aggregation and ion pair dissociation. Contributions from hydrogen bonding, proton transfer, π-π interactions, chromophore twisting, ion pairing, and self-assembly are systematically addressed and eliminated by concise structural modifications. Ion-exchange studies in solution, activation of cell-penetrating peptides in vesicles, and computational analysis all imply that the situation in the ground state is complementary to spectral tuning in the excited state; i.e., parallel rather than antiparallel ion pair-π interactions are preferred, despite repulsion from the push-pull dipole. The overall quite complete picture of ion pair-π interactions provided by these remarkably coherent yet complex results is expected to attract attention throughout the multiple disciplines of chemistry involved.

Where does the Raman optical activity of [Rh(en)3](3+) come from? Insight from a combined experimental and theoretical approach.

Backscattered Raman optical activity (ROA) spectra are measured for Δ- and Λ-tris-(ethylenediamine)rhodium(III) chloride in aqueous solution. In addition, the spectra of the four possible conformers in the Λ configuration are investigated by ab initio calculations. The Λ(δδδ) conformer is in best agreement with experimental spectra and examined in more details. The two most stable conformers according to the calculations are not compatible with the experimental ROA spectrum. Insights into the origin of observed band intensities are obtained by means of group coupling matrices. The influence of the first solvation shell is explored via an ab initio molecular dynamics simulation. Taking explicit solvent molecules into account further improves the agreement between calculation and experiment. Analysis of selected normal modes using group coupling matrices shows that solvent molecules lead to normal mode rotation and thus contribute to the ROA intensity, whereas the contribution of the Rh can be neglected.

Anion-π and cation-π interactions on the same surface.

Herein, we address the question whether anion-π and cation-π interactions can take place simultaneously on the same aromatic surface. Covalently positioned carboxylate-guanidinium pairs on the surface of 4-amino-1,8-naphthalimides are used as an example to explore push-pull chromophores as privileged platforms for such "ion pair-π" interactions. In antiparallel orientation with respect to the push-pull dipole, a bathochromic effect is observed. A red shift of 41 nm found in the least polar solvent is in good agreement with the 70 nm expected from theoretical calculations of ground and excited states. Decreasing shifts with solvent polarity, protonation, aggregation, and parallel carboxylate-guanidinium pairs imply that the intramolecular Stark effect from antiparallel ion pair-π interactions exceeds solvatochromic effects by far. Theoretical studies indicate that carboxylate-guanidinium pairs can also interact with the surfaces of π-acidic naphthalenediimides and π-basic pyrenes.

A polyaromatic terdentate binding unit with fused 5,6-membered chelates for complexing s-, p-, d-, and f-block cations.

The polyaromatic terdentate ligand 6-(azaindol-1-yl)-2,2'-bipyridine (L7) combines one 5-membered chelate ring with a fused 6-membered chelate ring. It is designed to provide complexation properties intermediate between 2,2';6',2″-terpyridine (L1) (two fused 5-membered chelate rings) and 2,6-bis(azaindol-1-yl)pyridine (L6) (two fused 6-membered chelate rings). In polar organic solvents, L7 displays remarkable affinities for the successive fixation of two small univalent cations M = H+ or Li+, leading to stable [M(m)(L7)]m+ (m = 1­2) complexes. Upon reaction with M = Mg2+ or Zn2+ cations, the large positive charge densities borne by the metals result in the successive cooperative complexation of two ligands to give [M(L7)n]n+ (n = 1­2). For small Sc3+, unavoidable traces of water favor the formation of the protonated ligand at millimolar concentrations in acetonitrile, but the use of larger Y3+ cations leads to [Y(L7)n]n+ (n = 1, 2), for which stability constants of log(ß(1,1)(Y,L7)) = 2.9(5) and log(ß(1,2)(Y,L7)) = 5.3(4) are estimated. The complexation behaviors are supported by speciations in solution, thermodynamic analyses, and solution and solid-state structures.