Corneal Innervation Research at a Crossroads: A Tool-Driven Roadmap for the Future  <br><i>Consensus by the National Eye Institute, U01-Funded Consortium on Ocular Surface Innervation</i>

SSRN Electronic Journal · 2026-01-01

Machine learning for discovery of clinical pain biomarkers following spinal cord injury

Experimental Neurology · 2026-01-14 · 2 citations

Chronic pain is highly prevalent in patients with spinal cord injury (SCI) and further degrades the quality of life in individuals already struggling with somatic, motor, and autonomic deficits. The management of SCI pain is challenging, mainly due to the lack of reliable, FDA-approved diagnostics, effective therapies, and incomplete understanding of the underlying mechanisms. These limitations have led to increased efforts dedicated to the identification of objective pain biomarkers. However, the FDA has yet to approve a physiologically relevant biomarker for the assessment of pain in populations with SCI. Given the multidimensional nature of pain, there is increasing recognition that composite biomarkers are needed. In this paper, we review several candidate pain signatures and discuss how the inclusion of multi-modal features such as self-reported questionnaires and behavioural measures should also be considered in the identification of comprehensive biomarkers of SCI pain. Since multi-modal, large-scale data presents a particular computational challenge, we further argue that AI and ML approaches enable novel combinatorial designs of SCI pain biomarkers. The advantages of AI and ML methods, which continue to evolve at a rapid pace, include computational efficiency, discovery of latent or embedded patterns in complex data architectures, personalized diagnostics, and minimization of potential bias. We also caution against over-reliance on physiological or neural imaging features that ignore the demographic, motivational, emotional, cognitive and cultural dimensions of pain, while advocating for AI/ML models with improved interpretability.

A Foundation Model for Sleep-Based Risk Stratification and Clinical Outcomes

Research Square · 2025-04-10 · 2 citations

Status and Opportunities of Machine Learning Applications in Obstructive Sleep Apnea: A Narrative Review

medRxiv · 2025-03-01 · 3 citations

Background: Obstructive sleep apnea (OSA) is a prevalent and potentially severe sleep disorder characterized by repeated interruptions in breathing during sleep. Machine learning models have been increasingly applied in various aspects of OSA research, including diagnosis, treatment optimization, and developing biomarkers for endotypes and disease mechanisms. Objective: This narrative review evaluates the application of machine learning in OSA research, focusing on model performance, dataset characteristics, demographic representation, and validation strategies. We aim to identify trends and gaps to guide future research and improve clinical decision-making that leverages machine learning. Methods: This narrative review examines data extracted from 254 scientific publications published in the PubMed database between January 2018 and March 2023. Studies were categorized by machine learning applications, models, tasks, validation metrics, data sources, and demographics. Results: Our analysis revealed that most machine learning applications focused on OSA classification and diagnosis, utilizing various data sources such as polysomnography, electrocardiogram data, and wearable devices. We also found that study cohorts were predominantly overweight males, with an underrepresentation of women, younger obese adults, individuals over 60 years old, and diverse racial groups. Many studies had small sample sizes and limited use of robust model validation. Conclusion: Our findings highlight the need for more inclusive research approaches, starting with adequate data collection in terms of sample size and bias mitigation for better generalizability of machine learning models in OSA research. Addressing these demographic gaps and methodological opportunities is critical for ensuring more robust and equitable applications of artificial intelligence in healthcare.

Spinal cord stimulation using time-dynamic pulses achieves longer reversal of allodynia compared to tonic pulses in a rat model of neuropathic pain

Frontiers in Pain Research · 2025-04-09

Background: Spinal cord stimulation (SCS) utilizing time-dynamic pulses (TDPs) is an emergent field of neuromodulation that continuously and automatically modulates pulse parameters. We previously demonstrated that TDPs delivered for 60 min at paresthesia-free or minimal paresthesia amplitudes significantly reversed allodynia in a rat model of neuropathic pain. Because the anti-allodynic effect was observed to persist post-stimulation, we hypothesized that the anti-nociceptive effects of TDPs may persist longer than those of tonic stimulation. Methods: We extended SCS stimulation period up to 90 min and investigated the temporal dynamics of SCS-induced analgesia through PWT analysis of the aggregated data from both cohorts. Results: Both TDPs and tonic stimulation reversed paw withdrawal thresholds (PWT) to near pre-neuropathic levels within 30 min. Most TDPs exhibited significantly slower ramp-up slope (analgesia 'wash-in' rates) as compared to tonic stimulation. All TDPs showed slower wind-down slopes (analgesia 'wash-out' rates) compared to tonic, with pulse width modulation reaching significance. Extending SCS from 60 to 90 min revealed that all TDPs maintained analgesic efficacy longer than tonic stimulation, which showed significant decrease at both 75 and 90 min. Discussion: Although TDPs and tonic stimulation comparably mitigated allodynia, TDPs exhibited slower rate of wash-out, suggesting longer-lasting analgesic effects and potentially different mechanisms of action.

How quantum computing can enhance biomarker discovery

Patterns · 2025-04-29 · 19 citations

Biomarkers play a central role in medicine's gradual progress toward proactive, personalized precision diagnostics and interventions. However, finding biomarkers that provide very early indicators of a change in health status, for example, for multifactorial diseases, has been challenging. The discovery of such biomarkers stands to benefit significantly from advanced information processing and means to detect complex correlations, which quantum computing offers. In this perspective, quantum algorithms, particularly in machine learning, are mapped to key applications in biomarker discovery. The opportunities and challenges associated with the algorithms and applications are discussed. The analysis is structured according to different data types-multidimensional, time series, and erroneous data-and covers key data modalities in healthcare-electronic health records, omics, and medical images. An outlook is provided concerning open research challenges.

0410 Cardiovascular Disease Incidence Differs by Patient Cluster Based on Raw Polysomnography Data Foundational Model

SLEEP · 2025-05-01

Abstract Introduction Traditional sleep measures are not consistently associated with incident cardiovascular events in observational studies. We hypothesized deep learning approaches could uniquely identify distinct patient groups with varying cardiovascular disease (CVD) incidence, independent of traditional measures. Methods We used artificial intelligence to analyze N=10,000 polysomnograms completed at Cleveland Clinic (1/2012-12/2022) enriched with underrepresented populations. We created a time-series foundational model guided by sleep macro-architecture and respiratory annotations, generating embeddings to cluster patients into risk groups using k-means. Follow-up time was from polysomnogram until death or last follow-up. Propensity scores were estimated using multinomial logistic regression on age, sex, body mass index (BMI), and years of available data, then inverse probability weighted, stabilized, and trimmed at 1st and 99th percentiles to minimize bias introduced by extreme values. Cox proportional hazards regression examined risk groups as predictors adjusted for age, sex, BMI, and apnea hypopnea index (AHI) with comorbidities included per outcome. Patients with baseline disease were excluded. Results The cohort [age 50.4±24.7 years, 50% male, 44% White, 34% Black, 5.3% Asian, 14.8% multiracial] had follow-up of 4.4[2.0-7.5] years. A 5-cluster solution provided the best stratification: Risk Group 1 (RG1) was reference. RG5 (highest AHI and arousal index, lowest mean and minimum SaO2 and total sleep time(TST)) had the highest CVD incidence (major adverse cardiovascular events(MACE): HR=1.60, 95%CI=1.13-2.28, myocardial infarction: HR=1.75, 95%CI=1.07-2.88, atrial fibrillation: HR=1.83, 95%CI=1.15-2.93) and all-cause mortality (HR=2.76, 95%CI=1.98-3.84). RG4 (highest %time SaO2< 90%) had elevated incidence to a lesser degree than RG5. RG3 (intermediate) had the highest stroke (HR=1.38, 95%CI=1.13-1.69) and ischemic heart disease (HR=2.01, 95%CI=1.37-2.94) incidence, second highest MACE incidence (HR=1.34, 95%CI=1.12-1.60) and lowest all-cause mortality (HR=1.47, 95%CI=1.14-1.90). RG2 (lowest AHI and %time SaO2< 90%, highest mean and minimum SaO2 and TST) had the lowest CVD incidence but not all-cause mortality (HR=1.47, 95%CI=1.14-1.90). Conclusion We created a large language model of raw polysomnogram data which identified groups that differed in CVD incidence after propensity score analysis. Risk groups were associated with adverse cardiovascular outcomes, thus supporting enhanced utility of a novel deep learning model over traditional approaches for CVD risk stratification. Future directions include validation with external data sources. Support (if any) IBM Discovery Accelerator, AIM Award

1143 Innovations in Time Series Data Integration in a Polysomnographic Foundational Model Informing Risk Groups

SLEEP · 2025-05-01

Abstract Introduction Despite the abundance of polysomnography (PSG) data, the limited summary metrics used in existing approaches may not provide the most informative insights for clinical decision-making. We hypothesized that a new data-driven clustering method using the entire multimodal raw PSG data could enable a precise risk stratification approach. We leveraged a new clinical data set to facilitate this data-driven approach and create a novel Foundation Model. Methods We utilized 10,000 PSGs conducted at the Cleveland (1/2012-12/2022) and custom artificial intelligence techniques that incorporate time-series data, to develop a Foundation Model from raw PSG data. We optimized this new model to classify sleep stages, respiratory events, and oxygen desaturations. Resulting embeddings were used to cluster patients into distinct risk groups with a k-means algorithm. Baseline characteristics were compared between risk groups using chi-square tests for categorical variables and Welch’s ANOVA for continuous variables. Means and standard deviations are reported. Results Optimal stratification of embeddings was achieved with five clusters. Resulting risk groups (RG) had a graded increase in age, male predominance, cardiovascular risk factors, and sleep-disordered breathing(SDB) severity from RG1(n=3,357) to RG5(n=363). Males were more prevalent in RG4(n=1,144; 60.8%) and RG5(66.4%) and least in RG2(n=1,877; 37.6%). Body mass index was lowest in RG2(32.7±9.1kg/m2) and highest in RG5(35.4±10.4kg/m2). RG4 and RG5 were the oldest(58.6±16.1years) and RG2 the youngest(44.0±15.0years). RG1 had intermediate SDB with apnea hypopnea index (AHI:12.4±12.4), total sleep time (TST:328±61.1min), and risk (hypertension:59.8%,diabetes II:32.7%), and lowest cognitive impairment(14.1%). RG2 had the mildest SDB (AHI:5.4±6.4), longest TST (342±71.6min), and lowest cardiovascular risk (hypertension:47.5%,diabetes II:24.7%). RG3(n=2,867) had intermediate SDB and risk. RG4 had more abnormal PSG measures (AHI:22.7±13.2, TST:201±73.9min), more risk (hypertension:75.6%,diabetes II:65.2%), and highest major adverse cardiovascular events (43.6%) but lowest migraine (11.5%). RG5 had the most severe SDB (AHI:37.3±39.0), lowest TST (98.4±89.0min), and high risk (hypertension:77.4%,diabetes II:44.6%). All p-values were < 0.001. Conclusion We created a Polysomnographic Foundational Model to stratify patients into risk groups characterized by different comorbidities not completely explained by traditional measures. RG4 and RG5 exhibited more severe SDB and unique clinical characteristics that prompt future investigation and warrant more attention from healthcare providers. Support (if any) IBM Discovery Accelerator, AIM Award, NIH 1R21HL170206-01

Status and opportunities of machine learning applications in obstructive sleep apnea: A narrative review

Computational and Structural Biotechnology Journal · 2025-01-01 · 5 citations

Background: Obstructive sleep apnea (OSA) is a prevalent and potentially severe sleep disorder characterized by repeated interruptions in breathing during sleep. Machine learning models have been increasingly applied in various aspects of OSA research, including diagnosis, treatment optimization, and developing biomarkers for endotypes and disease mechanisms. Objective: This narrative review evaluates the application of machine learning in OSA research, focusing on model performance, dataset characteristics, demographic representation, and validation strategies. We aim to identify trends and gaps to guide future research and improve clinical decision-making that leverages machine learning. Methods: This narrative review examines data extracted from 254 scientific publications published in the PubMed database between January 2018 and March 2023. Studies were categorized by machine learning applications, models, tasks, validation metrics, data sources, and demographics. Results: Our analysis revealed that most machine learning applications focused on OSA classification and diagnosis, utilizing various data sources such as polysomnography, electrocardiogram data, and wearable devices. We also found that study cohorts were predominantly overweight males, with an underrepresentation of women, younger obese adults, individuals over 60 years old, and diverse racial groups. Many studies had small sample sizes and limited use of robust model validation. Conclusion: Our findings highlight the need for more inclusive research approaches, starting with adequate data collection in terms of sample size and bias mitigation for better generalizability of machine learning models in OSA research. Addressing these demographic gaps and methodological opportunities is critical for ensuring more robust and equitable applications of artificial intelligence in healthcare.

Pain Assessment in Osteoarthritis: Present Practices and Future Prospects Including the Use of Biomarkers and Wearable Technologies, and AI‐Driven Personalized Medicine

Journal of Orthopaedic Research® · 2025-04-09 · 4 citations

Osteoarthritis (OA) is a highly prevalent chronic joint disorder affecting ~600 million individuals worldwide and is characterized by complex pain mechanisms that significantly impair patient quality of life. Challenges exist in accurately assessing and measuring pain in OA due to variations in pain perception among individuals and the heterogeneous nature of the disease. Conventional pain assessment methods, such as patient-reported outcome measures and clinical evaluations, often fail to fully capture the heterogeneity of pain experiences among individuals with OA. This review will summarize and evaluate current methods of pain assessment in OA and highlight future directions for standardized pain assessment. We discuss the role of animal models in enhancing our understanding of OA pain pathophysiology and highlight the necessity of translational research to advance pain assessment strategies. Key challenges explored include identifying phenotypes of pain susceptibility, integrating biomarkers into clinical practice, and adopting personalized pain management approaches through the incorporation of multi-modal data and multilevel analysis. We underscore the imperative for continued innovation in pain assessment and management to improve outcomes for patients with OA.