Multimodal Algorithms for the Classification of Circulation States During Out-of-Hospital Cardiac Arrest

Machine learning innovations in CPR: a comprehensive survey on enhanced resuscitation techniques

This survey paper explores the transformative role of Machine Learning (ML) and Artificial Intelligence (AI) in Cardiopulmonary Resuscitation (CPR), marking a paradigm shift from conventional, manually driven resuscitation practices to intelligent, data-driven interventions. It examines the evolution of CPR through the lens of predictive modeling, AI-enhanced devices, and real-time decision-making tools that collectively aim to improve resuscitation outcomes and survival rates. Unlike prior surveys that either focus solely on traditional CPR methods or offer general insights into ML applications in healthcare, this work provides a novel interdisciplinary synthesis tailored specifically to the domain of CPR. It presents a comprehensive taxonomy that classifies ML techniques into four key CPR-related tasks: rhythm analysis, outcome prediction, non-invasive blood pressure and chest compression modeling, and real-time detection of pulse and Return of Spontaneous Circulation (ROSC). The paper critically evaluates emerging ML approaches-including Reinforcement Learning (RL) and transformer-based models-while also addressing real-world implementation barriers such as model interpretability, data limitations, and deployment in high-stakes clinical settings. Furthermore, it highlights the role of eXplainable AI (XAI) in fostering clinical trust and adoption. By bridging the gap between resuscitation science and advanced ML techniques, this survey establishes a structured foundation for future research and practical innovation in ML-enhanced CPR. It offers clear insights, identifies unexplored opportunities, and sets a forward-looking research agenda identifying emerging trends and practical implementation challenges aiming to improve both the reliability and effectiveness of CPR in real-world emergencies.

A deep learning model for QRS delineation in organized rhythms during in-hospital cardiac arrest

BACKGROUND

Cardiac arrest (CA) is the sudden cessation of heart function, typically resulting in loss of consciousness and cessation of pulse and breathing. The electrocardiogram (ECG) stands as an essential tool extensively utilized by clinicians, during CA treatment. Within the ECG, the QRS complex reflects the depolarization of the ventricles, yielding valuable perspectives on cardiac health and potential irregularities. The delineation of QRS complexes is crucial for obtaining that information, but classical algorithms have not been tested with CA rhythms.

OBJECTIVE

This research aims to introduce a new deep learning-based model for accurately delineating QRS complexes in patients experiencing organized rhythms during in-hospital CA.

MATERIAL AND METHODS

Two databases have been employed, one comprising 332 episodes of in-hospital CA (151815 QRS complexes) and another consisting of 105 hemodynamically stable patients (112497 QRS complexes). The method comprises three stages: signal preprocessing for noise removal, windowing and sample classification with a U-Net model, and finally, the segmented windows are merged to complete the process.

RESULTS

The proposed method exhibited mean (standard deviation) F1 score/Sensitivity/Specificity/intersection over union values of 97.03(8.28)/ 97.69(11.38)/96.47(9.92)/79.09(15.78), and a 8.56(11.62) error for QRSon, and 25.11(25.86) for QRSoff instant delineation.

CONCLUSIONS

A precise delineator like this could support clinical practice by quantifying QRS features to enhance diagnostic accuracy and optimize treatment strategies.

Temporal Trends in End-Tidal Capnography and Outcomes in Out-of-Hospital Cardiac Arrest: A Secondary Analysis of a Randomized Clinical Trial

Importance

While widely measured, the time-varying association between exhaled end-tidal carbon dioxide (EtCO2) and out-of-hospital cardiac arrest (OHCA) outcomes is unclear.

Objective

To evaluate temporal associations between EtCO2 and return of spontaneous circulation (ROSC) in the Pragmatic Airway Resuscitation Trial (PART).

Design, Setting, and Participants

This study was a secondary analysis of a cluster randomized trial performed at multicenter emergency medical services agencies from the Resuscitation Outcomes Consortium. PART enrolled 3004 adults (aged ≥18 years) with nontraumatic OHCA from December 1, 2015, to November 4, 2017. EtCO2 was available in 1172 cases for this analysis performed in June 2023.

Interventions

PART evaluated the effect of laryngeal tube vs endotracheal intubation on 72-hour survival. Emergency medical services agencies collected continuous EtCO2 recordings using standard monitors, and this secondary analysis identified maximal EtCO2 values per ventilation and determined mean EtCO2 in 1-minute epochs using previously validated automated signal processing. All advanced airway cases with greater than 50% interpretable EtCO2 signal were included, and the slope of EtCO2 change over resuscitation was calculated.

Main Outcomes and Measures

The primary outcome was ROSC determined by prehospital or emergency department palpable pulses. EtCO2 values were compared at discrete time points using Mann-Whitney test, and temporal trends in EtCO2 were compared using Cochran-Armitage test of trend. Multivariable logistic regression was performed, adjusting for Utstein criteria and EtCO2 slope.

Results

Among 1113 patients included in the study, 694 (62.4%) were male; 285 (25.6%) were Black or African American, 592 (53.2%) were White, and 236 (21.2%) were another race; and the median (IQR) age was 64 (52-75) years. Cardiac arrest was most commonly unwitnessed (n = 579 [52.0%]), nonshockable (n = 941 [84.6%]), and nonpublic (n = 999 [89.8%]). There were 198 patients (17.8%) with ROSC and 915 (82.2%) without ROSC. Median EtCO2 values between ROSC and non-ROSC cases were significantly different at 10 minutes (39.8 [IQR, 27.1-56.4] mm Hg vs 26.1 [IQR, 14.9-39.0] mm Hg; P < .001) and 5 minutes (43.0 [IQR, 28.1-55.8] mm Hg vs 25.0 [IQR, 13.3-37.4] mm Hg; P < .001) prior to end of resuscitation. In ROSC cases, median EtCO2 increased from 30.5 (IQR, 22.4-54.2) mm HG to 43.0 (IQR, 28.1-55.8) mm Hg (P for trend < .001). In non-ROSC cases, EtCO2 declined from 30.8 (IQR, 18.2-43.8) mm Hg to 22.5 (IQR, 12.8-35.4) mm Hg (P for trend < .001). Using adjusted multivariable logistic regression with slope of EtCO2, the temporal change in EtCO2 was associated with ROSC (odds ratio, 1.45 [95% CI, 1.31-1.61]).

Conclusions and Relevance

In this secondary analysis of the PART trial, temporal increases in EtCO2 were associated with increased odds of ROSC. These results suggest value in leveraging continuous waveform capnography during OHCA resuscitation.

Trial Registration

ClinicalTrials.gov Identifier: NCT02419573.

A Random Forest Model for Pulseless Electrical Activity Prognosis Prediction During Out-of-Hospital Cardiac Arrest Using Invasive Blood Pressure

Out-of-hospital cardiac arrest (OHCA) is a major health concern, with an incidence of approximately 55 per 100,000 person-years in the United States. Pulseless electrical activity (PEA) is a cardiac rhythm observed in 20-30% of OHCA cases and it consists on a regular electrical activity presenting disassociation with cardiac mechanical contractions. Discriminating those PEA with favorable prognosis is crucial to decide pre/post resuscitation therapy. A machine learning model is proposed to assist rescuers to predict evolution of PEA. The ECG and the transthoracic impedance recorded using defibrillation pads were integrated in the model, together with the invasive blood pressure. A total of 238 PEA segments were extracted from 49 patients. A Random Forest model was trained with 25 features extracted from the three signals to discriminate between the PEA with favorable prognosis (return of spontaneous circulation). The optimal model showed median (interquartile range) values of 88.9(14.2)% for Area Under the Curve, 94.1(21.7)% for Sensitivity, 68.1(30.6)% for Specificity, 66.4(29.5)% for Positive Predictive Value, and 87.5(21.5)% for Negative Predictive Value.Clinical relevance- The study concludes that adding IBP based features to models traditionally based on ECG and TTI enhances PEA prognosis prediction during OHCA.

Extracting physiologic and clinical data from defibrillators for research purposes to improve treatment for patients in cardiac arrest

Background

A defibrillator should be connected to all patients receiving cardiopulmonary resuscitation (CPR) to allow early defibrillation. The defibrillator will collect signal data such as the electrocardiogram (ECG), thoracic impedance and end-tidal CO2, which allows for research on how patients demonstrate different responses to CPR. The aim of this review is to give an overview of methodological challenges and opportunities in using defibrillator data for research.

Methods

The successful collection of defibrillator files has several challenges. There is no scientific standard on how to store such data, which have resulted in several proprietary industrial solutions. The data needs to be exported to a software environment where signal filtering and classifications of ECG rhythms can be performed. This may be automated using different algorithms and artificial intelligence (AI). The patient can be classified being in ventricular fibrillation or -tachycardia, asystole, pulseless electrical activity or having obtained return of spontaneous circulation. How this dynamic response is time-dependent and related to covariates can be handled in several ways. These include Aalen's linear model, Weibull regression and joint models.

Conclusions

The vast amount of signal data from defibrillator represents promising opportunities for the use of AI and statistical analysis to assess patient response to CPR. This may provide an epidemiologic basis to improve resuscitation guidelines and give more individualized care. We suggest that an international working party is initiated to facilitate a discussion on how open formats for defibrillator data can be accomplished, that obligates industrial partners to further develop their current technological solutions.

Machine learning model to predict evolution of pulseless electrical activity during in-hospital cardiac arrest

Background

During pulseless electrical activity (PEA) the cardiac mechanical and electrical functions are dissociated, a phenomenon occurring in 25-42% of in-hospital cardiac arrest (IHCA) cases. Accurate evaluation of the likelihood of a PEA patient transitioning to return of spontaneous circulation (ROSC) may be vital for the successful resuscitation.

The aim

We sought to develop a model to automatically discriminate between PEA rhythms with favorable and unfavorable evolution to ROSC.

Methods

A dataset of 190 patients, 120 with ROSC, were acquired with defibrillators from different vendors in three hospitals. The ECG and the transthoracic impedance (TTI) signal were processed to compute 16 waveform features. Logistic regression models where designed integrating both automated features and characteristics annotated in the QRS to identify PEAs with better prognosis leading to ROSC. Cross validation techniques were applied, both patient-specific and stratified, to evaluate the performance of the algorithm.

Results

The best model consisted in a three feature algorithm that exhibited median (interquartile range) Area Under the Curve/Balanced accuracy/Sensitivity/Specificity of 80.3(9.9)/75.6(8.0)/ 77.4(15.2)/72.3(16.4) %, respectively.

Conclusions

Information hidden in the waveforms of the ECG and TTI signals, along with QRS complex features, can predict the progression of PEA. Automated methods as the one proposed in this study, could contribute to assist in the targeted treatment of PEA in IHCA.

Artificial intelligence in emergency medicine. A systematic literature review

Motivation and objective: Emergency medicine is becoming a popular application area for artificial intelligence methods but remains less investigated than other healthcare branches. The need for time-sensitive decision-making on the basis of high data volumes makes the use of quantitative technologies inevitable. However, the specifics of healthcare regulations impose strict requirements for such applications. Published contributions cover separate parts of emergency medicine and use disparate data and algorithms. This study aims to systematize the relevant contributions, investigate the main obstacles to artificial intelligence applications in emergency medicine, and propose directions for further studies.

METHODS

The contributions selection process was conducted with systematic electronic databases querying and filtering with respect to established exclusion criteria. Among the 380 papers gathered from IEEE Xplore, ACM Digital Library, Springer Library, ScienceDirect, and Nature databases 116 were considered to be a part of the survey. The main features of the selected papers are the focus on emergency medicine and the use of machine learning or deep learning algorithms.

FINDINGS AND DISCUSSION

The selected papers were classified into two branches: diagnostics-specific and triage-specific. The former ones are focused on either diagnosis prediction or decision support. The latter covers such applications as mortality, outcome, admission prediction, condition severity estimation, and urgent care prediction. The observed contributions are highly specialized within a single disease or medical operation and often use privately collected retrospective data, making them incomparable. These and other issues can be addressed by creating an end-to-end solution based on human-machine interaction.

CONCLUSION

Artificial intelligence applications are finding their place in emergency medicine, while most of the corresponding studies remain isolated and lack higher generalization and more sophisticated methodology, which can be a matter of forthcoming improvements.

Artificial intelligence to support out-of-hospital cardiac arrest care: A scoping review

Background

Artificial intelligence (AI) has demonstrated significant potential in supporting emergency medical services personnel during out-of-hospital cardiac arrest (OHCA) care; however, the extent of research evaluating this topic is unknown. This scoping review examines the breadth of literature on the application of AI in early OHCA care.

Methods

We conducted a search of PubMed®, Embase, and Web of Science in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews guidelines. Articles focused on non-traumatic OHCA and published prior to January 18th, 2023 were included. Studies were excluded if they did not use an AI intervention (including machine learning, deep learning, or natural language processing), or did not utilize data from the prehospital phase of care.

Results

Of 173 unique articles identified, 54 (31%) were included after screening. Of these studies, 15 (28%) were from the year 2022 and with an increasing trend annually starting in 2019. The majority were carried out by multinational collaborations (20/54, 38%) with additional studies from the United States (10/54, 19%), Korea (5/54, 10%), and Spain (3/54, 6%). Studies were classified into three major categories including ECG waveform classification and outcome prediction (24/54, 44%), early dispatch-level detection and outcome prediction (7/54, 13%), return of spontaneous circulation and survival outcome prediction (15/54, 20%), and other (9/54, 16%). All but one study had a retrospective design.

Conclusions

A small but growing body of literature exists describing the use of AI to augment early OHCA care.

Artificial intelligence and machine learning in prehospital emergency care: A scoping review

Our scoping review provides a comprehensive analysis of the landscape of artificial intelligence (AI) applications in prehospital emergency care (PEC). It contributes to the field by highlighting the most studied AI applications and identifying the most common methodological approaches across 106 included studies. The findings indicate a promising future for AI in PEC, with many unique use cases, such as prognostication, demand prediction, resource optimization, and the Internet of Things continuous monitoring systems. Comparisons with other approaches showed AI outperforming clinicians and non-AI algorithms in most cases. However, most studies were internally validated and retrospective, highlighting the need for rigorous prospective validation of AI applications before implementation in clinical settings. We identified knowledge and methodological gaps using an evidence map, offering a roadmap for future investigators. We also discussed the significance of explainable AI for establishing trust in AI systems among clinicians and facilitating real-world validation of AI models.

Impedance-Based Ventilation Detection and Signal Quality Control During Out-of-Hospital Cardiopulmonary Resuscitation

Feedback on ventilation could help improve cardiopulmonary resuscitation quality and survival from out-of-hospital cardiac arrest (OHCA). However, current technology that monitors ventilation during OHCA is very limited. Thoracic impedance (TI) is sensitive to air volume changes in the lungs, allowing ventilations to be identified, but is affected by artifacts due to chest compressions and electrode motion. This study introduces a novel algorithm to identify ventilations in TI during continuous chest compressions in OHCA. Data from 367 OHCA patients were included, and 2551 one-minute TI segments were extracted. Concurrent capnography data were used to annotate 20724 ground truth ventilations for training and evaluation. A three-step procedure was applied to each TI segment: First, bidirectional static and adaptive filters were applied to remove compression artifacts. Then, fluctuations potentially due to ventilations were located and characterized. Finally, a recurrent neural network was used to discriminate ventilations from other spurious fluctuations. A quality control stage was also developed to anticipate segments where ventilation detection could be compromised. The algorithm was trained and tested using 5-fold cross-validation, and outperformed previous solutions in the literature on the study dataset. The median (interquartile range, IQR) per-segment and per-patient F 1-scores were 89.1 (70.8-99.6) and 84.1 (69.0-93.9), respectively. The quality control stage identified most low performance segments. For the 50% of segments with highest quality scores, the median per-segment and per-patient F 1-scores were 100.0 (90.9-100.0) and 94.3 (86.5-97.8). The proposed algorithm could allow reliable, quality-conditioned feedback on ventilation in the challenging scenario of continuous manual CPR in OHCA.

Mathematical Model of Blood Circulation with Compression of the Prototype's Mechanical CPR Waveform

The waveform of chest compressions directly affects the blood circulation of patients with cardiac arrest. Currently, few pieces of research have focused on the influence of the cardiopulmonary resuscitation (CPR) device's mechanical waveform on blood circulation. This study investigates the effect of the mechanical waveform from a novel CPR prototype on blood circulation and explores the optimal compression parameters of the mechanical waveform to optimize blood circulation. A novel CPR prototype was designed and built to establish a kinetic model during compressions. The prototype's mechanical waveforms at various operating conditions were obtained for comparison with manual waveforms and the investigation of the optimal compression parameters. The novel CPR prototype can complete chest compressions quickly and stably. The cardiac output (CO), coronary perfusion pressure (CPP), and cerebral flow (CF) obtained by mechanical waveform compressions (1.22367 ± 0.00942 L/min, 30.95083 ± 0.24039 mmHg, 0.31992 ± 0.00343 L/min, respectively) were significantly better than those obtained by manual waveform compressions (1.10783 ± 0.03601 L/min, 21.39210 ± 1.42771 mmHg, 0.29598 ± 0.01344 L/min, respectively). With the compression of the prototype, the blood circulation can be optimized at the compression depth of 50 mm, approximately 0.6 duty cycle, and approximately 110 press/min, which is of guiding significance for the practical use of CPR devices to rescue patients with cardiac arrest.

Textural features reflecting local activity of the hippocampus improve the diagnosis of Alzheimer's disease and amnestic mild cognitive impairment: A radiomics study based on functional magnetic resonance imaging

Background

Textural features of the hippocampus in structural magnetic resonance imaging (sMRI) images can serve as potential diagnostic biomarkers for Alzheimer's disease (AD), while exhibiting a relatively poor discriminant performance in detecting early AD, such as amnestic mild cognitive impairment (aMCI). In contrast to sMRI, functional magnetic resonance imaging (fMRI) can identify brain functional abnormalities in the early stages of cerebral disorders. However, whether the textural features reflecting local functional activity in the hippocampus can improve the diagnostic performance for AD and aMCI remains unclear. In this study, we combined the textural features of the amplitude of low frequency fluctuation (ALFF) in the slow-5 frequency band and structural images in the hippocampus to investigate their diagnostic performance for AD and aMCI using multimodal radiomics technique.

Methods

Totally, 84 AD, 50 aMCI, and 44 normal controls (NCs) were included in the current study. After feature extraction and feature selection, the radiomics models incorporating sMRI images, ALFF values and their combinations in the bilateral hippocampus were established for the diagnosis of AD and aMCI. The effectiveness of these models was evaluated by receiver operating characteristic (ROC) analysis. The radiomics models were further validated using the external data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.

Results

The results of ROC analysis showed that the radiomics models based on structural images in the hippocampus had a better diagnostic performance for AD compared with the models using ALFF, while the ALFF-based model exhibited better discriminant performance for aMCI than the models with structural images. The radiomics models based on the combinations of structural images and ALFF were found to exhibit the highest accuracy for distinguishing AD from NCs and aMCI from NCs.

Conclusion

In this study, we found that the textural features reflecting local functional activity could improve the diagnostic performance of traditional structural models for both AD and aMCI. These findings may deepen our understanding of the pathogenesis of AD, contributing to the early diagnosis of AD.

Tidal volume measurements via transthoracic impedance waveform characteristics: The effect of age, body mass index and gender. A single centre interventional study

BACKGROUND AND AIM

Measuring tidal volumes (TV) during bag-valve ventilation is challenging in the clinical setting. The ventilation waveform amplitude of the transthoracic impedance (TTI-amplitude) correlates well with TV for an individual, but poorer between patients. We hypothesized that TV to TTI-amplitude relations could be improved when adjusted for morphometric variables like body mass index (BMI), gender or age, and that TTI-amplitude cut-offs for ventilations with adequate TV (>400ml) could be established.

MATERIALS AND METHODS

Twenty-one consenting adults (9 female, and 9 overall overweight) during positive pressure ventilation in anaesthesia before scheduled surgery were included. Seventeen ventilator modes were used (⩾ five breaths per mode) to adjust different TVs (150-800 ml), ventilation frequencies (10-30 min^-1) and insufflation times (0.5-3.5 s). TTI from the defibrillation pads was filtered to obtain ventilation TTI-amplitudes. Linear regression models were fitted between target and explanatory variables, and compared (coefficient of determination, R²).

RESULTS

The TV to TTI-amplitude slope was 1.39 Ω/l (R²=0.52), with significant differences (p<0.05) between male/female (1.04 Ω/l vs 1.84 Ω/l) and normal/overweight subjects (1.65 Ω/l vs 1.04 Ω/l). The median (interquartile range) TTI-amplitude cut-off for adequate TV was 0.51 Ω(0.14-1.20) with significant differences between males and females (0.58 Ω/0.39 Ω), and normal and overweight subjects (0.52 Ω/0.46 Ω). The TV to TTI-amplitude model improved (R²=0.66) when BMI, age and gender were included.

CONCLUSIONS

TTI-amplitude to TV relations were established and cut-offs for ventilations with adequate TV determined. Patient morphometric variables related to gender, age and BMI explain part of the variability in the measurements.

A Machine Learning Model for the Prognosis of Pulseless Electrical Activity during Out-of-Hospital Cardiac Arrest

Pulseless electrical activity (PEA) is characterized by the disassociation of the mechanical and electrical activity of the heart and appears as the initial rhythm in 20–30% of out-of-hospital cardiac arrest (OHCA) cases. Predicting whether a patient in PEA will convert to return of spontaneous circulation (ROSC) is important because different therapeutic strategies are needed depending on the type of PEA. The aim of this study was to develop a machine learning model to differentiate PEA with unfavorable (unPEA) and favorable (faPEA) evolution to ROSC. An OHCA dataset of 1921 5s PEA signal segments from defibrillator files was used, 703 faPEA segments from 107 patients with ROSC and 1218 unPEA segments from 153 patients with no ROSC. The solution consisted of a signal-processing stage of the ECG and the thoracic impedance (TI) and the extraction of the TI circulation component (ICC), which is associated with ventricular wall movement. Then, a set of 17 features was obtained from the ECG and ICC signals, and a random forest classifier was used to differentiate faPEA from unPEA. All models were trained and tested using patientwise and stratified 10-fold cross-validation partitions. The best model showed a median (interquartile range) area under the curve (AUC) of 85.7(9.8)% and a balance accuracy of 78.8(9.8)%, improving the previously available solutions at more than four points in the AUC and three points in balanced accuracy. It was demonstrated that the evolution of PEA can be predicted using the ECG and TI signals, opening the possibility of targeted PEA treatment in OHCA.