Predicting dementia from spontaneous speech using large language models

Generative artificial intelligence in healthcare: A scoping review on benefits, challenges and applications

BACKGROUND

Generative artificial intelligence (GAI) is revolutionizing healthcare with solutions for complex challenges, enhancing diagnosis, treatment, and care through new data and insights. However, its integration raises questions about applications, benefits, and challenges. Our study explores these aspects, offering an overview of GAI's applications and future prospects in healthcare.

METHODS

This scoping review searched Web of Science, PubMed, and Scopus . The selection of studies involved screening titles, reviewing abstracts, and examining full texts, adhering to the PRISMA-ScR guidelines throughout the process.

RESULTS

From 1406 articles across three databases, 109 met inclusion criteria after screening and deduplication. Nine GAI models were utilized in healthcare, with ChatGPT (n = 102, 74 %), Google Bard (Gemini) (n = 16, 11 %), and Microsoft Bing AI (n = 10, 7 %) being the most frequently employed. A total of 24 different applications of GAI in healthcare were identified, with the most common being "offering insights and information on health conditions through answering questions" (n = 41) and "diagnosis and prediction of diseases" (n = 17). In total, 606 benefits and challenges were identified, which were condensed to 48 benefits and 61 challenges after consolidation. The predominant benefits included "Providing rapid access to information and valuable insights" and "Improving prediction and diagnosis accuracy", while the primary challenges comprised "generating inaccurate or fictional content", "unknown source of information and fake references for texts", and "lower accuracy in answering questions".

CONCLUSION

This scoping review identified the applications, benefits, and challenges of GAI in healthcare. This synthesis offers a crucial overview of GAI's potential to revolutionize healthcare, emphasizing the imperative to address its limitations.

Progress on early diagnosing Alzheimer's disease

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that affects both cognition and non-cognition functions. The disease follows a continuum, starting with preclinical stages, progressing to mild cognitive and behavioral impairment, ultimately leading to dementia. Early detection of AD is crucial for better diagnosis and more effective treatment. However, the current AD diagnostic tests of biomarkers using cerebrospinal fluid and/or brain imaging are invasive or expensive, and mostly are still not able to detect early disease state. Consequently, there is an urgent need to develop new diagnostic techniques with higher sensitivity and specificity during the preclinical stages of AD. Various non-cognitive manifestations, including behavioral abnormalities, sleep disturbances, sensory dysfunctions, and physical changes, have been observed in the preclinical AD stage before occurrence of notable cognitive decline. Recent research advances have identified several biofluid biomarkers as early indicators of AD. This review focuses on these non-cognitive changes and newly discovered biomarkers in AD, specifically addressing the preclinical stages of the disease. Furthermore, it is of importance to explore the potential for developing a predictive system or network to forecast disease onset and progression at the early stage of AD.

Recommendations for designing conversational companion robots with older adults through foundation models

Companion robots are aimed to mitigate loneliness and social isolation among older adults by providing social and emotional support in their everyday lives. However, older adults' expectations of conversational companionship might substantially differ from what current technologies can achieve, as well as from other age groups like young adults. Thus, it is crucial to involve older adults in the development of conversational companion robots to ensure that these devices align with their unique expectations and experiences. The recent advancement in foundation models, such as large language models, has taken a significant stride toward fulfilling those expectations, in contrast to the prior literature that relied on humans controlling robots (i.e., Wizard of Oz) or limited rule-based architectures that are not feasible to apply in the daily lives of older adults. Consequently, we conducted a participatory design (co-design) study with 28 older adults, demonstrating a companion robot using a large language model (LLM), and design scenarios that represent situations from everyday life. The thematic analysis of the discussions around these scenarios shows that older adults expect a conversational companion robot to engage in conversation actively in isolation and passively in social settings, remember previous conversations and personalize, protect privacy and provide control over learned data, give information and daily reminders, foster social skills and connections, and express empathy and emotions. Based on these findings, this article provides actionable recommendations for designing conversational companion robots for older adults with foundation models, such as LLMs and vision-language models, which can also be applied to conversational robots in other domains.

A Systematic Review of ChatGPT and Other Conversational Large Language Models in Healthcare

Background

The launch of the Chat Generative Pre-trained Transformer (ChatGPT) in November 2022 has attracted public attention and academic interest to large language models (LLMs), facilitating the emergence of many other innovative LLMs. These LLMs have been applied in various fields, including healthcare. Numerous studies have since been conducted regarding how to employ state-of-the-art LLMs in health-related scenarios to assist patients, doctors, and public health administrators.

Objective

This review aims to summarize the applications and concerns of applying conversational LLMs in healthcare and provide an agenda for future research on LLMs in healthcare.

Methods

We utilized PubMed, ACM, and IEEE digital libraries as primary sources for this review. We followed the guidance of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRIMSA) to screen and select peer-reviewed research articles that (1) were related to both healthcare applications and conversational LLMs and (2) were published before September 1st, 2023, the date when we started paper collection and screening. We investigated these papers and classified them according to their applications and concerns.

Results

Our search initially identified 820 papers according to targeted keywords, out of which 65 papers met our criteria and were included in the review. The most popular conversational LLM was ChatGPT from OpenAI (60), followed by Bard from Google (1), Large Language Model Meta AI (LLaMA) from Meta (1), and other LLMs (5). These papers were classified into four categories in terms of their applications: 1) summarization, 2) medical knowledge inquiry, 3) prediction, and 4) administration, and four categories of concerns: 1) reliability, 2) bias, 3) privacy, and 4) public acceptability. There are 49 (75%) research papers using LLMs for summarization and/or medical knowledge inquiry, and 58 (89%) research papers expressing concerns about reliability and/or bias. We found that conversational LLMs exhibit promising results in summarization and providing medical knowledge to patients with a relatively high accuracy. However, conversational LLMs like ChatGPT are not able to provide reliable answers to complex health-related tasks that require specialized domain expertise. Additionally, no experiments in our reviewed papers have been conducted to thoughtfully examine how conversational LLMs lead to bias or privacy issues in healthcare research.

Conclusions

Future studies should focus on improving the reliability of LLM applications in complex health-related tasks, as well as investigating the mechanisms of how LLM applications brought bias and privacy issues. Considering the vast accessibility of LLMs, legal, social, and technical efforts are all needed to address concerns about LLMs to promote, improve, and regularize the application of LLMs in healthcare.

A novel speech analysis algorithm to detect cognitive impairment in a Spanish population

Objective

Early detection of cognitive impairment in the elderly is crucial for diagnosis and appropriate care. Brief, cost-effective cognitive screening instruments are needed to help identify individuals who require further evaluation. This study presents preliminary data on a new screening technology using automated voice recording analysis software in a Spanish population.

Method

Data were collected from 174 Spanish-speaking individuals clinically diagnosed as cognitively normal (CN, n = 87) or impaired (mild cognitive impairment [MCI], n = 63; all-cause dementia, n = 24). Participants were recorded performing four common language tasks (Animal fluency, alternating fluency [sports and fruits], phonemic "F" fluency, and Cookie Theft Description). Recordings were processed via text-transcription and digital-signal processing techniques to capture neuropsychological variables and audio characteristics. A training sample of 122 subjects with similar demographics across groups was used to develop an algorithm to detect cognitive impairment. Speech and task features were used to develop five independent machine learning (ML) models to compute scores between 0 and 1, and a final algorithm was constructed using repeated cross-validation. A socio-demographically balanced subset of 52 participants was used to test the algorithm. Analysis of covariance (ANCOVA), covarying for demographic characteristics, was used to predict logistically-transformed algorithm scores.

Results

Mean logit algorithm scores were significantly different across groups in the testing sample (p < 0.01). Comparisons of CN with impaired (MCI + dementia) and MCI groups using the final algorithm resulted in an AUC of 0.93/0.90, with overall accuracy of 88.4%/87.5%, sensitivity of 87.5/83.3, and specificity of 89.2/89.2, respectively.

Conclusion

Findings provide initial support for the utility of this automated speech analysis algorithm as a screening tool for cognitive impairment in Spanish speakers. Additional study is needed to validate this technology in larger and more diverse clinical populations.

An Entity Extraction Pipeline for Medical Text Records Using Large Language Models: Analytical Study

BACKGROUND

The study of disease progression relies on clinical data, including text data, and extracting valuable features from text data has been a research hot spot. With the rise of large language models (LLMs), semantic-based extraction pipelines are gaining acceptance in clinical research. However, the security and feature hallucination issues of LLMs require further attention.

OBJECTIVE

This study aimed to introduce a novel modular LLM pipeline, which could semantically extract features from textual patient admission records.

METHODS

The pipeline was designed to process a systematic succession of concept extraction, aggregation, question generation, corpus extraction, and question-and-answer scale extraction, which was tested via 2 low-parameter LLMs: Qwen-14B-Chat (QWEN) and Baichuan2-13B-Chat (BAICHUAN). A data set of 25,709 pregnancy cases from the People's Hospital of Guangxi Zhuang Autonomous Region, China, was used for evaluation with the help of a local expert's annotation. The pipeline was evaluated with the metrics of accuracy and precision, null ratio, and time consumption. Additionally, we evaluated its performance via a quantified version of Qwen-14B-Chat on a consumer-grade GPU.

RESULTS

The pipeline demonstrates a high level of precision in feature extraction, as evidenced by the accuracy and precision results of Qwen-14B-Chat (95.52% and 92.93%, respectively) and Baichuan2-13B-Chat (95.86% and 90.08%, respectively). Furthermore, the pipeline exhibited low null ratios and variable time consumption. The INT4-quantified version of QWEN delivered an enhanced performance with 97.28% accuracy and a 0% null ratio.

CONCLUSIONS

The pipeline exhibited consistent performance across different LLMs and efficiently extracted clinical features from textual data. It also showed reliable performance on consumer-grade hardware. This approach offers a viable and effective solution for mining clinical research data from textual records.

The Optimization of a Natural Language Processing Approach for the Automatic Detection of Alzheimer's Disease Using GPT Embeddings

The development of noninvasive and cost-effective methods of detecting Alzheimer's disease (AD) is essential for its early prevention and mitigation. We optimize the detection of AD using natural language processing (NLP) of spontaneous speech through the use of audio enhancement techniques and novel transcription methodologies. Specifically, we utilized Boll Spectral Subtraction to improve audio fidelity and created transcriptions using state-of-the-art AI services-locally-based Wav2Vec and Whisper, alongside cloud-based IBM Cloud and Rev AI-evaluating their performance against traditional manual transcription methods. Support Vector Machine (SVM) classifiers were then trained and tested using GPT-based embeddings of transcriptions. Our findings revealed that AI-based transcriptions largely outperformed traditional manual ones, with Wav2Vec (enhanced audio) achieving the best accuracy and F-1 score (0.99 for both metrics) for locally-based systems and Rev AI (standard audio) performing the best for cloud-based systems (0.96 for both metrics). Furthermore, this study revealed the detrimental effects of interviewer speech on model performance in addition to the minimal effect of audio enhancement. Based on our findings, current AI transcription and NLP technologies are highly effective at accurately detecting AD with available data but struggle to classify probable AD and mild cognitive impairment (MCI), a prodromal stage of AD, due to a lack of training data, laying the groundwork for the future implementation of an automatic AD detection system.

Investigating the Impact of Prompt Engineering on the Performance of Large Language Models for Standardizing Obstetric Diagnosis Text: Comparative Study

BACKGROUND

The accumulation of vast electronic medical records (EMRs) through medical informatization creates significant research value, particularly in obstetrics. Diagnostic standardization across different health care institutions and regions is vital for medical data analysis. Large language models (LLMs) have been extensively used for various medical tasks. Prompt engineering is key to use LLMs effectively.

OBJECTIVE

This study aims to evaluate and compare the performance of LLMs with various prompt engineering techniques on the task of standardizing obstetric diagnostic terminology using real-world obstetric data.

METHODS

The paper describes a 4-step approach used for mapping diagnoses in electronic medical records to the International Classification of Diseases, 10th revision, observation domain. First, similarity measures were used for mapping the diagnoses. Second, candidate mapping terms were collected based on similarity scores above a threshold, to be used as the training data set. For generating optimal mapping terms, we used two LLMs (ChatGLM2 and Qwen-14B-Chat [QWEN]) for zero-shot learning in step 3. Finally, a performance comparison was conducted by using 3 pretrained bidirectional encoder representations from transformers (BERTs), including BERT, whole word masking BERT, and momentum contrastive learning with BERT (MC-BERT), for unsupervised optimal mapping term generation in the fourth step.

RESULTS

LLMs and BERT demonstrated comparable performance at their respective optimal levels. LLMs showed clear advantages in terms of performance and efficiency in unsupervised settings. Interestingly, the performance of the LLMs varied significantly across different prompt engineering setups. For instance, when applying the self-consistency approach in QWEN, the F1-score improved by 5%, with precision increasing by 7.9%, outperforming the zero-shot method. Likewise, ChatGLM2 delivered similar rates of accurately generated responses. During the analysis, the BERT series served as a comparative model with comparable results. Among the 3 models, MC-BERT demonstrated the highest level of performance. However, the differences among the versions of BERT in this study were relatively insignificant.

CONCLUSIONS

After applying LLMs to standardize diagnoses and designing 4 different prompts, we compared the results to those generated by the BERT model. Our findings indicate that QWEN prompts largely outperformed the other prompts, with precision comparable to that of the BERT model. These results demonstrate the potential of unsupervised approaches in improving the efficiency of aligning diagnostic terms in daily research and uncovering hidden information values in patient data.

Large Language Models and Healthcare Alliance: Potential and Challenges of Two Representative Use Cases

Large language models (LLMS) emerge as the most promising Natural Language Processing approach for clinical practice acceleration (i.e., diagnosis, prevention and treatment procedures). Similarly, intelligent conversational systems that leverage LLMS have disruptively become the future of therapy in the era of ChatGPT. Accordingly, this research addresses the application of LLMS in healthcare, paying particular attention to two relevant use cases: cognitive decline and depression, more specifically, postpartum depression. In the end, the most promising opportunities they represent (e.g., clinical tasks augmentation, personalized healthcare, etc.) and related concerns (e.g., data privacy and quality, fairness, etc.) are discussed to contribute to the global debate on their integration in the sanitary system.

Redefining Virtual Assistants in Health Care: The Future With Large Language Models

This editorial explores the evolving and transformative role of large language models (LLMs) in enhancing the capabilities of virtual assistants (VAs) in the health care domain, highlighting recent research on the performance of VAs and LLMs in health care information sharing. Focusing on recent research, this editorial unveils the marked improvement in the accuracy and clinical relevance of responses from LLMs, such as GPT-4, compared to current VAs, especially in addressing complex health care inquiries, like those related to postpartum depression. The improved accuracy and clinical relevance with LLMs mark a paradigm shift in digital health tools and VAs. Furthermore, such LLM applications have the potential to dynamically adapt and be integrated into existing VA platforms, offering cost-effective, scalable, and inclusive solutions. These suggest a significant increase in the applicable range of VA applications, as well as the increased value, risk, and impact in health care, moving toward more personalized digital health ecosystems. However, alongside these advancements, it is necessary to develop and adhere to ethical guidelines, regulatory frameworks, governance principles, and privacy and safety measures. We need a robust interdisciplinary collaboration to navigate the complexities of safely and effectively integrating LLMs into health care applications, ensuring that these emerging technologies align with the diverse needs and ethical considerations of the health care domain.

Prediction of multiclass surgical outcomes in glaucoma using multimodal deep learning based on free-text operative notes and structured EHR data

OBJECTIVE

Surgical outcome prediction is challenging but necessary for postoperative management. Current machine learning models utilize pre- and post-op data, excluding intraoperative information in surgical notes. Current models also usually predict binary outcomes even when surgeries have multiple outcomes that require different postoperative management. This study addresses these gaps by incorporating intraoperative information into multimodal models for multiclass glaucoma surgery outcome prediction.

MATERIALS AND METHODS

We developed and evaluated multimodal deep learning models for multiclass glaucoma trabeculectomy surgery outcomes using both structured EHR data and free-text operative notes. We compare those to baseline models that use structured EHR data exclusively, or neural network models that leverage only operative notes.

RESULTS

The multimodal neural network had the highest performance with a macro AUROC of 0.750 and F1 score of 0.583. It outperformed the baseline machine learning model with structured EHR data alone (macro AUROC of 0.712 and F1 score of 0.486). Additionally, the multimodal model achieved the highest recall (0.692) for hypotony surgical failure, while the surgical success group had the highest precision (0.884) and F1 score (0.775).

DISCUSSION

This study shows that operative notes are an important source of predictive information. The multimodal predictive model combining perioperative notes and structured pre- and post-op EHR data outperformed other models. Multiclass surgical outcome prediction can provide valuable insights for clinical decision-making.

CONCLUSIONS

Our results show the potential of deep learning models to enhance clinical decision-making for postoperative management. They can be applied to other specialties to improve surgical outcome predictions.

The Optimization of a Natural Language Processing Approach for the Automatic Detection of Alzheimer's Disease Using GPT Embeddings

As the impact of Alzheimer's disease (AD) is projected to grow in the coming decades as the world's population ages, the development of noninvasive and cost-effective methods of detecting AD is essential for the early prevention and mitigation of the progressive disease, alleviating its expected global impact. This study analyzes audio processing techniques and transcription methodologies to optimize the detection of AD through the natural language processing (NLP) of spontaneous speech. We enhanced audio fidelity using Boll Spectral Subtraction and evaluated the transcription accuracy of state-of-the-art AI services-locally-based Wav2Vec and Whisper, alongside cloud-based IBM Cloud and Rev AI-against traditional manual transcription methods. The choice between local and cloud-based solutions hinges on a trade-off between privacy, ongoing costs, and computational requirements. Leveraging OpenAI's GPT for word embeddings, we enhanced the training of Support Vector Machine (SVM) classifiers, which were crucial in analyzing transcripts and refining detection accuracy. Our findings reveal that AI-driven transcriptions significantly outperform manual counterparts when classifying AD and Control samples, with Wav2Vec using enhanced audio exhibiting the highest accuracy and F-1 scores (0.99 for both metrics) for locally based systems and Rev AI using unenhanced audio leading cloud-based methods with comparable precision (0.96 for both metrics). The study also uncovers the detrimental effect of including interviewer speech in recordings on model performance, advocating for the exclusion of such interactions to improve data quality for AD classification algorithms. Our comprehensive evaluation demonstrates that AI transcription (both Cloud and Local) and NLP technologies in their current forms can classify AD, as well as probable AD and mild cognitive impairment (MCI), a prodromal stage of AD, accurately but suffer from a lack of available training data. The insights garnered from this research lay the groundwork for future advancements in the noninvasive monitoring and early detection of cognitive impairments through linguistic analysis.

What Does ChatGPT Know About Dementia? A Comparative Analysis of Information Quality

The quality of information about dementia retrieved using ChatGPT is unknown. Content was evaluated for length, readability, and quality using the QUEST, a validated tool, and compared against online material from three North American organizations. Both sources of information avoided conflicts of interest, supported the patient-physician relationship, and used a balanced tone. Official bodies but not ChatGPT referenced identifiable research and pointed to local resources. Users of ChatGPT are likely to encounter accurate but shallow information about dementia. Recommendations are made for information creators and providers who counsel patients around digital health practices.

Clinical classification of memory and cognitive impairment with multimodal digital biomarkers

INTRODUCTION

Early detection of Alzheimer's disease and cognitive impairment is critical to improving the healthcare trajectories of aging adults, enabling early intervention and potential prevention of decline.

METHODS

To evaluate multi-modal feature sets for assessing memory and cognitive impairment, feature selection and subsequent logistic regressions were used to identify the most salient features in classifying Rey Auditory Verbal Learning Test-determined memory impairment.

RESULTS

Multimodal models incorporating graphomotor, memory, and speech and voice features provided the stronger classification performance (area under the curve = 0.83; sensitivity = 0.81, specificity = 0.80). Multimodal models were superior to all other single modality and demographics models.

DISCUSSION

The current research contributes to the prevailing multimodal profile of those with cognitive impairment, suggesting that it is associated with slower speech with a particular effect on the duration, frequency, and percentage of pauses compared to normal healthy speech.

Large Language Models in Neurology Research and Future Practice

Recent advancements in generative artificial intelligence, particularly using large language models (LLMs), are gaining increased public attention. We provide a perspective on the potential of LLMs to analyze enormous amounts of data from medical records and gain insights on specific topics in neurology. In addition, we explore use cases for LLMs, such as early diagnosis, supporting patient and caregivers, and acting as an assistant for clinicians. We point to the potential ethical and technical challenges raised by LLMs, such as concerns about privacy and data security, potential biases in the data for model training, and the need for careful validation of results. Researchers must consider these challenges and take steps to address them to ensure that their work is conducted in a safe and responsible manner. Despite these challenges, LLMs offer promising opportunities for improving care and treatment of various neurologic disorders.

Two Directions for Clinical Data Generation with Large Language Models: Data-to-Label and Label-to-Data

Large language models (LLMs) can generate natural language texts for various domains and tasks, but their potential for clinical text mining, a domain with scarce, sensitive, and imbalanced medical data, is under-explored. We investigate whether LLMs can augment clinical data for detecting Alzheimer's Disease (AD)-related signs and symptoms from electronic health records (EHRs), a challenging task that requires high expertise. We create a novel pragmatic taxonomy for AD sign and symptom progression based on expert knowledge and generated three datasets: (1) a gold dataset annotated by human experts on longitudinal EHRs of AD patients; (2) a silver dataset created by the data-to-label method, which labels sentences from a public EHR collection with AD-related signs and symptoms; and (3) a bronze dataset created by the label-to-data method which generates sentences with AD-related signs and symptoms based on the label definition. We train a system to detect AD-related signs and symptoms from EHRs. We find that the silver and bronze datasets improves the system performance, outperforming the system using only the gold dataset. This shows that LLMs can generate synthetic clinical data for a complex task by incorporating expert knowledge, and our label-to-data method can produce datasets that are free of sensitive information, while maintaining acceptable quality.

Leveraging GPT-4 for food effect summarization to enhance product-specific guidance development via iterative prompting

Food effect summarization from New Drug Application (NDA) is an essential component of product-specific guidance (PSG) development and assessment, which provides the basis of recommendations for fasting and fed bioequivalence studies to guide the pharmaceutical industry for developing generic drug products. However, manual summarization of food effect from extensive drug application review documents is time-consuming. Therefore, there is a need to develop automated methods to generate food effect summary. Recent advances in natural language processing (NLP), particularly large language models (LLMs) such as ChatGPT and GPT-4, have demonstrated great potential in improving the effectiveness of automated text summarization, but its ability with regard to the accuracy in summarizing food effect for PSG assessment remains unclear. In this study, we introduce a simple yet effective approach,iterative prompting, which allows one to interact with ChatGPT or GPT-4 more effectively and efficiently through multi-turn interaction. Specifically, we propose a three-turn iterative prompting approach to food effect summarization in which the keyword-focused and length-controlled prompts are respectively provided in consecutive turns to refine the quality of the generated summary. We conduct a series of extensive evaluations, ranging from automated metrics to FDA professionals and even evaluation by GPT-4, on 100 NDA review documents selected over the past five years. We observe that the summary quality is progressively improved throughout the iterative prompting process. Moreover, we find that GPT-4 performs better than ChatGPT, as evaluated by FDA professionals (43% vs. 12%) and GPT-4 (64% vs. 35%). Importantly, all the FDA professionals unanimously rated that 85% of the summaries generated by GPT-4 are factually consistent with the golden reference summary, a finding further supported by GPT-4 rating of 72% consistency. Taken together, these results strongly suggest a great potential for GPT-4 to draft food effect summaries that could be reviewed by FDA professionals, thereby improving the efficiency of the PSG assessment cycle and promoting generic drug product development.

Artificial intelligence for dementia research methods optimization

Artificial intelligence (AI) and machine learning (ML) approaches are increasingly being used in dementia research. However, several methodological challenges exist that may limit the insights we can obtain from high-dimensional data and our ability to translate these findings into improved patient outcomes. To improve reproducibility and replicability, researchers should make their well-documented code and modeling pipelines openly available. Data should also be shared where appropriate. To enhance the acceptability of models and AI-enabled systems to users, researchers should prioritize interpretable methods that provide insights into how decisions are generated. Models should be developed using multiple, diverse datasets to improve robustness, generalizability, and reduce potentially harmful bias. To improve clarity and reproducibility, researchers should adhere to reporting guidelines that are co-produced with multiple stakeholders. If these methodological challenges are overcome, AI and ML hold enormous promise for changing the landscape of dementia research and care. HIGHLIGHTS: Machine learning (ML) can improve diagnosis, prevention, and management of dementia. Inadequate reporting of ML procedures affects reproduction/replication of results. ML models built on unrepresentative datasets do not generalize to new datasets. Obligatory metrics for certain model structures and use cases have not been defined. Interpretability and trust in ML predictions are barriers to clinical translation.

Toward Personalized Medicine Approaches for Parkinson Disease Using Digital Technologies

Parkinson disease (PD) is a complex neurodegenerative disorder that afflicts over 10 million people worldwide, resulting in debilitating motor and cognitive impairment. In the United States alone (with approximately 1 million cases), the economic burden for treating and caring for persons with PD exceeds US $50 billion and myriad therapeutic approaches are under development, including both symptomatic- and disease-modifying agents. The challenges presented in addressing PD are compounded by observations that numerous, statistically distinct patient phenotypes present with a wide variety of motor and nonmotor symptomatic profiles, varying responses to current standard-of-care symptom-alleviating medications (L-DOPA and dopaminergic agonists), and different disease trajectories. The existence of these differing phenotypes highlights the opportunities in personalized approaches to symptom management and disease control. The prodromal period of PD can span across several decades, allowing the potential to leverage the unique array of composite symptoms presented to trigger early interventions. This may be especially beneficial as disease progression in PD (alongside Alzheimer disease and Huntington disease) may be influenced by biological processes such as oxidative stress, offering the potential for individual lifestyle factors to be tailored to delay disease onset. In this viewpoint, we offer potential scenarios where emerging diagnostic and monitoring strategies might be tailored to the individual patient under the tenets of P4 medicine (predict, prevent, personalize, and participate). These approaches may be especially relevant as the causative factors and biochemical pathways responsible for the observed neurodegeneration in patients with PD remain areas of fluid debate. The numerous observational patient cohorts established globally offer an excellent opportunity to test and refine approaches to detect, characterize, control, modify the course, and ultimately stop progression of this debilitating disease. Such approaches may also help development of parallel interventive strategies in other diseases such as Alzheimer disease and Huntington disease, which share common traits and etiologies with PD. In this overview, we highlight near-term opportunities to apply P4 medicine principles for patients with PD and introduce the concept of composite orthogonal patient monitoring.

Automated Classification of Cognitive Decline and Probable Alzheimer's Dementia Across Multiple Speech and Language Domains

BACKGROUND

Decline in language has emerged as a new potential biomarker for the early detection of Alzheimer's disease (AD). It remains unclear how sensitive language measures are across different tasks, language domains, and languages, and to what extent changes can be reliably detected in early stages such as subjective cognitive decline (SCD) and mild cognitive impairment (MCI).

METHOD

Using a scene construction task for speech elicitation in a new Spanish/Catalan speaking cohort (N = 119), we automatically extracted features across seven domains, three acoustic (spectral, cepstral, and voice quality), one prosodic, and three from text (morpholexical, semantic, and syntactic). They were forwarded to a random forest classifier to evaluate the discriminability of participants with probable AD dementia, amnestic and nonamnestic MCI, SCD, and cognitively healthy controls. Repeated-measures analyses of variance and paired-samples Wilcoxon signed-ranks test were used to assess whether and how performance differs significantly across groups and linguistic domains.

RESULTS

The performance scores of the machine learning classifier were generally satisfactorily high, with the highest scores over .9. Model performance was significantly different for linguistic domains (p < .001), and speech versus text (p = .043), with speech features outperforming textual features, and voice quality performing best. High diagnostic classification accuracies were seen even within both cognitively healthy (controls vs. SCD) and MCI (amnestic and nonamnestic) groups.

CONCLUSION

Speech-based machine learning is powerful in detecting cognitive decline and probable AD dementia across a range of different feature domains, though important differences exist between these domains as well.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.23699733.

AD-BERT: Using pre-trained language model to predict the progression from mild cognitive impairment to Alzheimer's disease

OBJECTIVE

We develop a deep learning framework based on the pre-trained Bidirectional Encoder Representations from Transformers (BERT) model using unstructured clinical notes from electronic health records (EHRs) to predict the risk of disease progression from Mild Cognitive Impairment (MCI) to Alzheimer's Disease (AD).

METHODS

We identified 3657 patients diagnosed with MCI together with their progress notes from Northwestern Medicine Enterprise Data Warehouse (NMEDW) between 2000 and 2020. The progress notes no later than the first MCI diagnosis were used for the prediction. We first preprocessed the notes by deidentification, cleaning and splitting into sections, and then pre-trained a BERT model for AD (named AD-BERT) based on the publicly available Bio+Clinical BERT on the preprocessed notes. All sections of a patient were embedded into a vector representation by AD-BERT and then combined by global MaxPooling and a fully connected network to compute the probability of MCI-to-AD progression. For validation, we conducted a similar set of experiments on 2563 MCI patients identified at Weill Cornell Medicine (WCM) during the same timeframe.

RESULTS

Compared with the 7 baseline models, the AD-BERT model achieved the best performance on both datasets, with Area Under receiver operating characteristic Curve (AUC) of 0.849 and F1 score of 0.440 on NMEDW dataset, and AUC of 0.883 and F1 score of 0.680 on WCM dataset.

CONCLUSION

The use of EHRs for AD-related research is promising, and AD-BERT shows superior predictive performance in modeling MCI-to-AD progression prediction. Our study demonstrates the utility of pre-trained language models and clinical notes in predicting MCI-to-AD progression, which could have important implications for improving early detection and intervention for AD.

Explainable Alzheimer's Disease Detection Using Linguistic Features from Automatic Speech Recognition

INTRODUCTION

Alzheimer's disease (AD) is the most prevalent type of dementia and can cause abnormal cognitive function and progressive loss of essential life skills. Early screening is thus necessary for the prevention and intervention of AD. Speech dysfunction is an early onset symptom of AD patients. Recent studies have demonstrated the promise of automated acoustic assessment using acoustic or linguistic features extracted from speech. However, most previous studies have relied on manual transcription of text to extract linguistic features, which weakens the efficiency of automated assessment. The present study thus investigates the effectiveness of automatic speech recognition (ASR) in building an end-to-end automated speech analysis model for AD detection.

METHODS

We implemented three publicly available ASR engines and compared the classification performance using the ADReSS-IS2020 dataset. Besides, the SHapley Additive exPlanations algorithm was then used to identify critical features that contributed most to model performance.

RESULTS

Three automatic transcription tools obtained mean word error rate texts of 32%, 43%, and 40%, respectively. These automated texts achieved similar or even better results than manual texts in model performance for detecting dementia, achieving classification accuracies of 89.58%, 83.33%, and 81.25%, respectively.

CONCLUSION

Our best model, using ensemble learning, is comparable to the state-of-the-art manual transcription-based methods, suggesting the possibility of an end-to-end medical assistance system for AD detection with ASR engines. Moreover, the critical linguistic features might provide insight into further studies on the mechanism of AD.

Artificial Intelligence for Dementia Research Methods Optimization

INTRODUCTION

Machine learning (ML) has been extremely successful in identifying key features from high-dimensional datasets and executing complicated tasks with human expert levels of accuracy or greater.

METHODS

We summarize and critically evaluate current applications of ML in dementia research and highlight directions for future research.

RESULTS

We present an overview of ML algorithms most frequently used in dementia research and highlight future opportunities for the use of ML in clinical practice, experimental medicine, and clinical trials. We discuss issues of reproducibility, replicability and interpretability and how these impact the clinical applicability of dementia research. Finally, we give examples of how state-of-the-art methods, such as transfer learning, multi-task learning, and reinforcement learning, may be applied to overcome these issues and aid the translation of research to clinical practice in the future.

DISCUSSION

ML-based models hold great promise to advance our understanding of the underlying causes and pathological mechanisms of dementia.

Diagnostic Accuracy of Differential-Diagnosis Lists Generated by Generative Pretrained Transformer 3 Chatbot for Clinical Vignettes with Common Chief Complaints: A Pilot Study

The diagnostic accuracy of differential diagnoses generated by artificial intelligence (AI) chatbots, including the generative pretrained transformer 3 (GPT-3) chatbot (ChatGPT-3) is unknown. This study evaluated the accuracy of differential-diagnosis lists generated by ChatGPT-3 for clinical vignettes with common chief complaints. General internal medicine physicians created clinical cases, correct diagnoses, and five differential diagnoses for ten common chief complaints. The rate of correct diagnosis by ChatGPT-3 within the ten differential-diagnosis lists was 28/30 (93.3%). The rate of correct diagnosis by physicians was still superior to that by ChatGPT-3 within the five differential-diagnosis lists (98.3% vs. 83.3%, p = 0.03). The rate of correct diagnosis by physicians was also superior to that by ChatGPT-3 in the top diagnosis (53.3% vs. 93.3%, p < 0.001). The rate of consistent differential diagnoses among physicians within the ten differential-diagnosis lists generated by ChatGPT-3 was 62/88 (70.5%). In summary, this study demonstrates the high diagnostic accuracy of differential-diagnosis lists generated by ChatGPT-3 for clinical cases with common chief complaints. This suggests that AI chatbots such as ChatGPT-3 can generate a well-differentiated diagnosis list for common chief complaints. However, the order of these lists can be improved in the future.

Artificial Intelligence-Enabled End-To-End Detection and Assessment of Alzheimer's Disease Using Voice

There is currently no simple, widely available screening method for Alzheimer's disease (AD), partly because the diagnosis of AD is complex and typically involves expensive and sometimes invasive tests not commonly available outside highly specialized clinical settings. Here, we developed an artificial intelligence (AI)-powered end-to-end system to detect AD and predict its severity directly from voice recordings. At the core of our system is the pre-trained data2vec model, the first high-performance self-supervised algorithm that works for speech, vision, and text. Our model was internally evaluated on the ADReSSo (Alzheimer's Dementia Recognition through Spontaneous Speech only) dataset containing voice recordings of subjects describing the Cookie Theft picture, and externally validated on a test dataset from DementiaBank. The AI model can detect AD with average area under the curve (AUC) of 0.846 and 0.835 on held-out and external test set, respectively. The model was well-calibrated (Hosmer-Lemeshow goodness-of-fit p-value = 0.9616). Moreover, the model can reliably predict the subject's cognitive testing score solely based on raw voice recordings. Our study demonstrates the feasibility of using the AI-powered end-to-end model for early AD diagnosis and severity prediction directly based on voice, showing its potential for screening Alzheimer's disease in a community setting.