Identifying Acute Low Back Pain Episodes in Primary Care Practice From Clinical Notes: Observational Study

The Role and Applications of Artificial Intelligence in the Treatment of Chronic Pain

PURPOSE OF REVIEW

This review aims to explore the interface between artificial intelligence (AI) and chronic pain, seeking to identify areas of focus for enhancing current treatments and yielding novel therapies.

RECENT FINDINGS

In the United States, the prevalence of chronic pain is estimated to be upwards of 40%. Its impact extends to increased healthcare costs, reduced economic productivity, and strain on healthcare resources. Addressing this condition is particularly challenging due to its complexity and the significant variability in how patients respond to treatment. Current options often struggle to provide long-term relief, with their benefits rarely outweighing the risks, such as dependency or other side effects. Currently, AI has impacted four key areas of chronic pain treatment and research: (1) predicting outcomes based on clinical information; (2) extracting features from text, specifically clinical notes; (3) modeling 'omic data to identify meaningful patient subgroups with potential for personalized treatments and improved understanding of disease processes; and (4) disentangling complex neuronal signals responsible for pain, which current therapies attempt to modulate. As AI advances, leveraging state-of-the-art architectures will be essential for improving chronic pain treatment. Current efforts aim to extract meaningful representations from complex data, paving the way for personalized medicine. The identification of unique patient subgroups should reveal targets for tailored chronic pain treatments. Moreover, enhancing current treatment approaches is achievable by gaining a more profound understanding of patient physiology and responses. This can be realized by leveraging AI on the increasing volume of data linked to chronic pain.

Comparison of Machine Learning Methods in the Study of Cancer Survivors' Return to Work: An Example of Breast Cancer Survivors with Work-Related Factors in the CONSTANCES Cohort

PURPOSE

Machine learning (ML) methods showed a higher accuracy in identifying individuals without cancer who were unable to return to work (RTW) compared to the classical methods (e.g. logistic regression models). We therefore aim to discuss the value of these methods in relation to RTW for cancer survivors.

METHODS

Breast cancer (BC) survivors who were working at diagnosis within the CONSTANCES cohort were included in the study. RTW was assessed five years after the BC diagnosis (early retirement was considered as non-RTW). Age and occupation at diagnosis, and physical occupational job exposures assessed using the Job Exposure Matrix, JEM-CONSTANCES, were evaluated as predictors of RTW five years after BC diagnosis. The following four ML methods were used: (i) k-nearest neighbors; (ii) random forest; (iii) neural network; and (iv) elastic net.

RESULTS

The training sample included 683 BC survivors (RTW: 85.7%), and the test sample 171 (RTW: 85.4%). The elastic net method had the best results despite low sensitivity (accuracy = 76.6%; sensitivity = 31.7%; specificity = 90.8%), and the random forest model was the most accurate (= 79.5%) but also the least sensitive (= 14.3%).

CONCLUSION

This study takes a first step towards opening up new possibilities for identifying the occupational determinants of cancer survivors' RTW. Further work, including a larger sample size, and more predictor variables, is now needed.

Identifying Risk Factors Associated With Lower Back Pain in Electronic Medical Record Free Text: Deep Learning Approach Using Clinical Note Annotations

Background

Lower back pain is a common weakening condition that affects a large population. It is a leading cause of disability and lost productivity, and the associated medical costs and lost wages place a substantial burden on individuals and society. Recent advances in artificial intelligence and natural language processing have opened new opportunities for the identification and management of risk factors for lower back pain. In this paper, we propose and train a deep learning model on a data set of clinical notes that have been annotated with relevant risk factors, and we evaluate the model's performance in identifying risk factors in new clinical notes.

Objective

The primary objective is to develop a novel deep learning approach to detect risk factors for underlying disease in patients presenting with lower back pain in clinical encounter notes. The secondary objective is to propose solutions to potential challenges of using deep learning and natural language processing techniques for identifying risk factors in electronic medical record free text and make practical recommendations for future research in this area.

Methods

We manually annotated clinical notes for the presence of six risk factors for severe underlying disease in patients presenting with lower back pain. Data were highly imbalanced, with only 12% (n=296) of the annotated notes having at least one risk factor. To address imbalanced data, a combination of semantic textual similarity and regular expressions was used to further capture notes for annotation. Further analysis was conducted to study the impact of downsampling, binary formulation of multi-label classification, and unsupervised pretraining on classification performance.

Results

Of 2749 labeled clinical notes, 347 exhibited at least one risk factor, while 2402 exhibited none. The initial analysis shows that downsampling the training set to equalize the ratio of clinical notes with and without risk factors improved the macro-area under the receiver operating characteristic curve (AUROC) by 2%. The Bidirectional Encoder Representations from Transformers (BERT) model improved the macro-AUROC by 15% over the traditional machine learning baseline. In experiment 2, the proposed BERT-convolutional neural network (CNN) model for longer texts improved (4% macro-AUROC) over the BERT baseline, and the multitask models are more stable for minority classes. In experiment 3, domain adaptation of BERTCNN using masked language modeling improved the macro-AUROC by 2%.

Conclusions

Primary care clinical notes are likely to require manipulation to perform meaningful free-text analysis. The application of BERT models for multi-label classification on downsampled annotated clinical notes is useful in detecting risk factors suggesting an indication for imaging for patients with lower back pain.

Artificial Intelligence for Automatic Pain Assessment: Research Methods and Perspectives

Although proper pain evaluation is mandatory for establishing the appropriate therapy, self-reported pain level assessment has several limitations. Data-driven artificial intelligence (AI) methods can be employed for research on automatic pain assessment (APA). The goal is the development of objective, standardized, and generalizable instruments useful for pain assessment in different clinical contexts. The purpose of this article is to discuss the state of the art of research and perspectives on APA applications in both research and clinical scenarios. Principles of AI functioning will be addressed. For narrative purposes, AI-based methods are grouped into behavioral-based approaches and neurophysiology-based pain detection methods. Since pain is generally accompanied by spontaneous facial behaviors, several approaches for APA are based on image classification and feature extraction. Language features through natural language strategies, body postures, and respiratory-derived elements are other investigated behavioral-based approaches. Neurophysiology-based pain detection is obtained through electroencephalography, electromyography, electrodermal activity, and other biosignals. Recent approaches involve multimode strategies by combining behaviors with neurophysiological findings. Concerning methods, early studies were conducted by machine learning algorithms such as support vector machine, decision tree, and random forest classifiers. More recently, artificial neural networks such as convolutional and recurrent neural network algorithms are implemented, even in combination. Collaboration programs involving clinicians and computer scientists must be aimed at structuring and processing robust datasets that can be used in various settings, from acute to different chronic pain conditions. Finally, it is crucial to apply the concepts of explainability and ethics when examining AI applications for pain research and management.

Natural language processing in low back pain and spine diseases: A systematic review

Natural Language Processing (NLP) is a discipline at the intersection between Computer Science (CS), Artificial Intelligence (AI), and Linguistics that leverages unstructured human-interpretable (natural) language text. In recent years, it gained momentum also in health-related applications and research. Although preliminary, studies concerning Low Back Pain (LBP) and other related spine disorders with relevant applications of NLP methodologies have been reported in the literature over the last few years. It motivated us to systematically review the literature comprised of two major public databases, PubMed and Scopus. To do so, we first formulated our research question following the PICO guidelines. Then, we followed a PRISMA-like protocol by performing a search query including terminologies of both technical (e.g., natural language and computational linguistics) and clinical (e.g., lumbar and spine surgery) domains. We collected 221 non-duplicated studies, 16 of which were eligible for our analysis. In this work, we present these studies divided into sub-categories, from both tasks and exploited models’ points of view. Furthermore, we report a detailed description of techniques used to extract and process textual features and the several evaluation metrics used to assess the performance of the NLP models. However, what is clear from our analysis is that additional studies on larger datasets are needed to better define the role of NLP in the care of patients with spinal disorders.

A comparative study on deep learning models for text classification of unstructured medical notes with various levels of class imbalance

BACKGROUND

Discharge medical notes written by physicians contain important information about the health condition of patients. Many deep learning algorithms have been successfully applied to extract important information from unstructured medical notes data that can entail subsequent actionable results in the medical domain. This study aims to explore the model performance of various deep learning algorithms in text classification tasks on medical notes with respect to different disease class imbalance scenarios.

METHODS

In this study, we employed seven artificial intelligence models, a CNN (Convolutional Neural Network), a Transformer encoder, a pretrained BERT (Bidirectional Encoder Representations from Transformers), and four typical sequence neural networks models, namely, RNN (Recurrent Neural Network), GRU (Gated Recurrent Unit), LSTM (Long Short-Term Memory), and Bi-LSTM (Bi-directional Long Short-Term Memory) to classify the presence or absence of 16 disease conditions from patients' discharge summary notes. We analyzed this question as a composition of 16 binary separate classification problems. The model performance of the seven models on each of the 16 datasets with various levels of imbalance between classes were compared in terms of AUC-ROC (Area Under the Curve of the Receiver Operating Characteristic), AUC-PR (Area Under the Curve of Precision and Recall), F1 Score, and Balanced Accuracy as well as the training time. The model performances were also compared in combination with different word embedding approaches (GloVe, BioWordVec, and no pre-trained word embeddings).

RESULTS

The analyses of these 16 binary classification problems showed that the Transformer encoder model performs the best in nearly all scenarios. In addition, when the disease prevalence is close to or greater than 50%, the Convolutional Neural Network model achieved a comparable performance to the Transformer encoder, and its training time was 17.6% shorter than the second fastest model, 91.3% shorter than the Transformer encoder, and 94.7% shorter than the pre-trained BERT-Base model. The BioWordVec embeddings slightly improved the performance of the Bi-LSTM model in most disease prevalence scenarios, while the CNN model performed better without pre-trained word embeddings. In addition, the training time was significantly reduced with the GloVe embeddings for all models.

CONCLUSIONS

For classification tasks on medical notes, Transformer encoders are the best choice if the computation resource is not an issue. Otherwise, when the classes are relatively balanced, CNNs are a leading candidate because of their competitive performance and computational efficiency.

Artificial Intelligence and Computer Aided Diagnosis in Chronic Low Back Pain: A Systematic Review

Low Back Pain (LBP) is currently the first cause of disability in the world, with a significant socioeconomic burden. Diagnosis and treatment of LBP often involve a multidisciplinary, individualized approach consisting of several outcome measures and imaging data along with emerging technologies. The increased amount of data generated in this process has led to the development of methods related to artificial intelligence (AI), and to computer-aided diagnosis (CAD) in particular, which aim to assist and improve the diagnosis and treatment of LBP. In this manuscript, we have systematically reviewed the available literature on the use of CAD in the diagnosis and treatment of chronic LBP. A systematic research of PubMed, Scopus, and Web of Science electronic databases was performed. The search strategy was set as the combinations of the following keywords: “Artificial Intelligence”, “Machine Learning”, “Deep Learning”, “Neural Network”, “Computer Aided Diagnosis”, “Low Back Pain”, “Lumbar”, “Intervertebral Disc Degeneration”, “Spine Surgery”, etc. The search returned a total of 1536 articles. After duplication removal and evaluation of the abstracts, 1386 were excluded, whereas 93 papers were excluded after full-text examination, taking the number of eligible articles to 57. The main applications of CAD in LBP included classification and regression. Classification is used to identify or categorize a disease, whereas regression is used to produce a numerical output as a quantitative evaluation of some measure. The best performing systems were developed to diagnose degenerative changes of the spine from imaging data, with average accuracy rates >80%. However, notable outcomes were also reported for CAD tools executing different tasks including analysis of clinical, biomechanical, electrophysiological, and functional imaging data. Further studies are needed to better define the role of CAD in LBP care.

A Comparison of Natural Language Processing Methods for the Classification of Lumbar Spine Imaging Findings Related to Lower Back Pain

RATIONALE AND OBJECTIVES

The use of natural language processing (NLP) in radiology provides an opportunity to assist clinicians with phenotyping patients. However, the performance and generalizability of NLP across healthcare systems is uncertain. We assessed the performance within and generalizability across four healthcare systems of different NLP representational methods, coupled with elastic-net logistic regression to classify lower back pain-related findings from lumbar spine imaging reports.

MATERIALS AND METHODS

We used a dataset of 871 X-ray and magnetic resonance imaging reports sampled from a prospective study across four healthcare systems between October 2013 and September 2016. We annotated each report for 26 findings potentially related to lower back pain. Our framework applied four different NLP methods to convert text into feature sets (representations). For each representation, our framework used an elastic-net logistic regression model for each finding (i.e., 26 binary or "one-vs.-rest" classification models). For performance evaluation, we split data into training (80%, 697/871) and testing (20%, 174/871). In the training set, we used cross validation to identify the optimal hyperparameter value and then retrained on the full training set. We then assessed performance based on area under the curve (AUC) for the test set. We repeated this process 25 times with each repeat using a different random train/test split of the data, so that we could estimate 95% confidence intervals, and assess significant difference in performance between representations. For generalizability evaluation, we trained models on data from three healthcare systems with cross validation and then tested on the fourth. We repeated this process for each system, then calculated mean and standard deviation (SD) of AUC across the systems.

RESULTS

For individual representations, n-grams had the best average performance across all 26 findings (AUC: 0.960). For generalizability, document embeddings had the most consistent average performance across systems (SD: 0.010). Out of these 26 findings, we considered eight as potentially clinically important (any stenosis, central stenosis, lateral stenosis, foraminal stenosis, disc extrusion, nerve root displacement compression, endplate edema, and listhesis grade 2) since they have a relatively greater association with a history of lower back pain compared to the remaining 18 classes. We found a similar pattern for these eight in which n-grams and document embeddings had the best average performance (AUC: 0.954) and generalizability (SD: 0.007), respectively.

CONCLUSION

Based on performance assessment, we found that n-grams is the preferred method if classifier development and deployment occur at the same system. However, for deployment at multiple systems outside of the development system, or potentially if physician behavior changes within a system, one should consider document embeddings since embeddings appear to have the most consistent performance across systems.

Improving appropriate imaging for non-specific low back pain

Non-specific low back pain (LBP) is a common condition in the USA, with approximately 80% of adults who will have LBP at some point during their life and roughly 30% of the adult population suffering from LBP at any given time. Although LBP is the most common cause of disability in the USA, it often has no identifiable anatomic or physiologic cause. Many patients seeking care for non-specific LBP receive X-rays and other imaging studies. However, for most acute LBP patients, symptoms resolve within 4 weeks and the use of routine imaging may result in unnecessary radiation exposure and add unnecessary costs and wasted time for patients without contributing to patient outcomes. The specific aim of the quality improvement (QI) project was to determine the effect of a multicomponent intervention to enhance the appropriate imaging utilisation for acute LBP to ≥90%. During the first 6 months of the QI project, 191 patients with LBP were seen. Of those patients, 156 (81.7%) received appropriate imaging over the 6-month intervention period, missing our targeted goal. Furthermore, this rate declined to baseline values after termination of the intervention, suggesting the need for additional prompts to sustain the initial intervention effect. Following a health system-wide deployment of practice-based alerts and quality score cards, the appropriate utilisation rate increased again and quickly to the target rate of 90%. To reduce variability in our clinical practice and to sustain an appropriate utilisation rate will require continued work. Health systems must find efficient methods to reduce LBP imaging and increase appropriate management of non-specific LBP in primary care. Increasing concordance with imaging guidelines can lessen harm associated with unnecessary radiation exposure and result in significant cost savings.

Modern Clinical Text Mining: A Guide and Review

Electronic health records (EHRs) are becoming a vital source of data for healthcare quality improvement, research, and operations. However, much of the most valuable information contained in EHRs remains buried in unstructured text. The field of clinical text mining has advanced rapidly in recent years, transitioning from rule-based approaches to machine learning and, more recently, deep learning. With new methods come new challenges, however, especially for those new to the field. This review provides an overview of clinical text mining for those who are encountering it for the first time (e.g., physician researchers, operational analytics teams, machine learning scientists from other domains). While not a comprehensive survey, this review describes the state of the art, with a particular focus on new tasks and methods developed over the past few years. It also identifies key barriers between these remarkable technical advances and the practical realities of implementation in health systems and in industry.