Pith. sign in

REVIEW 4 major objections 5 minor 159 references

Predicting Extubation Failure in Intensive Care: The Development of a Novel, End-to-End Actionable and Interpretable Prediction System

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read Uniform resampling of irregular ICU data makes LSTM and TCN models collapse to single-class predictions; stratifying features by sampling frequency and fusing their outputs restores weak discrimination (AUC-ROC around 0.6).

desk verdict Honest empirical study showing that naive resampling of irregular ICU data creates complete class bias in temporal models; useful as a methodological caution, but the outcome definition counting routine O2 as failure undermines the ground truth. read the letter →

arxiv 2412.00105 v1 pith:F47BVRGT submitted 2024-11-27 cs.LG cs.CE

classification cs.LGcs.CE MSC 68T0768T0592C50
keywords extubationfailuremechanicalventilationMIMIC-IVtemporaldeeplearningLSTMconvolutionalnetworksyntheticdataclinicalinterpretability
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This thesis argues that the main obstacle to predicting extubation failure from ICU time series is not the choice of neural architecture but the way irregularly recorded data is resampled into synthetic sequences. The author shows that when all features are uniformly interpolated to 30-minute steps, both LSTM and TCN models collapse into predicting every patient as the same class, regardless of hyperparameter tuning or added static features. Splitting features into low-, medium-, and high-frequency subsets, resampling each at a rate that respects its observed density, masking absent values, and fusing the three sub-models removes that total bias. On 4,701 MIMIC-IV patients the resulting models still discriminate only weakly, with AUC-ROC around 0.6 and F1 below 0.5. The point matters because extubation decisions are high-stakes, and the finding redirects attention from architecture toward data-synthesis and outcome labelling.

What carries the argument

The central mechanism is the frequency-stratified feature subset with a fused decision head. Dynamic features are divided into groups based on average observation frequency in the 6-hour window, resampled at rates matched to their observed density, and passed through separate LSTM or TCN branches; static features are processed by a feed-forward network; the last valid output of each branch is concatenated and fused into a sigmoid prediction. Masking tells the models to ignore time steps where no real observation existed, avoiding wholesale synthetic imputation. This machinery is what converts complete single-class bias into weak but genuine discrimination.

What would settle it

Re-label the same MIMIC-IV cohort with failure defined only as reintubation or death within 48 hours, dropping the 6-hour ventilatory-support criterion, and retrain the fused LSTM/TCN pipeline; if AUC-ROC jumps substantially above 0.6, the broad definition's mislabelled failures were the main performance ceiling, while if AUC-ROC stays near 0.6 the weak discrimination is intrinsic to the data.

Watch

Extended reading notes

Core claim

The paper's central claim is that synthetic data handling, not model architecture, was the dominant driver of the observed failures. Vanilla LSTM and TCN models trained on data uniformly resampled to 30-minute intervals showed complete bias, predicting all patients into one class; this persisted through grid search, Bayesian optimisation, static data inclusion, and architecture changes. The author then grouped dynamic features into low, medium, and high sampling-frequency subsets, resampled each subset at a bespoke rate (2 hours, 1 hour, and 30 minutes respectively), masked missing values instead of imputing them, and processed each subset through a separate LSTM or TCN whose outputs were fused. This frequency-stratified, fused design eliminated the total class bias, but final discrimination remained modest across all architectures and feature sets, with AUC-ROC around 0.6 and F1 consistently below 0.5.

Load-bearing premise

The outcome definition is load-bearing: extubation failure is labelled as reintubation or death within 48 hours, or any ventilatory support (NIV, oxygen flow, CPAP, BiPAP) within 6 hours after extubation, which yields a 32.8% failure rate, and if routine post-extubation support is not true failure then a large fraction of labels are wrong.

Editorial extensions

If this is right

  • If the paper is right, future LSTM/TCN studies on irregular ICU data should stratify features by observation frequency before declaring an architecture inadequate; uniform resampling can by itself produce total class bias.
  • The fused decision system removes complete bias but caps discrimination at AUC-ROC around 0.6, so the temporal architectures as configured here are not yet sufficient for clinical use in extubation failure prediction.
  • Static data and additional features do not reliably move AUC-ROC; parsimonious feature sets perform comparably to larger ones, echoing earlier findings that smaller feature sets can match larger ones.
  • Feature ablation shows near-zero change for most individual features, meaning single-feature interpretability offers little actionable guidance on this dataset.
  • Because the study defines failure to include any ventilatory support within 6 hours, the cohort failure rate is 32.8%, far above the commonly cited 10–20%, so comparisons with prior extubation models must account for label definition.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The frequency-stratified, mask-and-fuse recipe could be tested as a general preprocessing baseline for any irregularly sampled electronic health record time series before attributing poor performance to the model family.
  • A cleaner outcome definition (reintubation or death within 48 hours only) might raise apparent discrimination; if AUC-ROC rises markedly, much of the observed ceiling is label noise rather than missing predictive signal.
  • Because the fused design keeps frequency subsets separate until the final layer, cross-frequency interactions (for example, a low-frequency blood gas paired with a high-frequency oxygen saturation) are only available very late; an input-level multi-resolution fusion would directly test whether those interactions carry signal.
  • The near-zero ablation deltas suggest the models are insensitive to single features; permutation-based importance across entire subsets would test whether the signal lives in groups of features rather than in individuals.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper develops and evaluates an end-to-end machine learning pipeline for predicting extubation failure from 6 hours of pre-extubation data in MIMIC-IV (4,701 patients). Three feature sets, constructed by cross-referencing literature popularity with clinical availability in the WAVE study, are used to train LSTM, TCN, and LightGBM models. Initial temporal models trained on uniformly resampled data are reported to be completely biased toward one class; the authors then group features by sampling frequency, resample each subset at a bespoke rate, mask missing values, and combine three subnetworks in a fused decision architecture. The final models attain AUC-ROC around 0.6 and F1 below 0.5, with no architecture, feature set, or static-data configuration clearly superior. The paper honestly reports weak discrimination and frames its contribution as a set of preprocessing strategies that remove complete class bias and as a foundation for future work.

Significance. If the empirical claims hold, the paper makes a useful cautionary contribution: it documents how naive uniform resampling of sparsely and irregularly sampled ICU time series can induce complete class bias, and it demonstrates a frequency-stratified, masking-based alternative that restores some discriminatory signal. The study also releases code on GitHub, involves clinician input in cohort and feature decisions, and reports negative results without overclaiming clinical utility. However, the significance is tempered because the headline result is modest (AUC-ROC about 0.6), the evaluation is based on point estimates from a single split, and the outcome-label definition is nonstandard and potentially mislabels routine post-extubation oxygen support as failure. The paper's value is therefore primarily methodological and cautionary, rather than a demonstrated clinically actionable predictor.

major comments (4)
  1. [§3.1, Data annotation] The outcome definition is load-bearing and needs a sensitivity analysis. Extubation failure is defined as reintubation or death within 48 hours, or any ventilatory support including 'O2 flow' within 6 hours post-extubation. This yields a 32.8% failure rate in the final cohort, roughly double the 10–20% range the paper itself cites in §2.2. Supplemental oxygen is routinely administered after extubation as supportive care, so a large fraction of positive labels may not represent true failure. Since the central claim is that the frequency-stratified preprocessing removed complete class bias, that claim is currently about models trained and evaluated on potentially mislabeled ground truth. The authors acknowledge the lack of consensus but provide no sensitivity analysis. I request ablations of the label: (a) reintubation/mortality only, (b) exclude the O2-flow criterion, and (c) exclude all ventilatory-support-only positives. If the bias-removal result is robust to these label variations, the claim is substantially strengthened; if not, the conclusion needs to be restricted to the authors' specific label definition.
  2. [§5, Table 5.1 and §3.2, train/test split] All reported results are point estimates from a single 80/20 stratified split, with no confidence intervals, bootstrap replicates, or repeated-seed experiments. Differences such as Fused LSTM AUC-ROC 0.6567 versus Fused TCN 0.6116 on Feature Set 1, or the modest AUC-ROC changes when static data is added, may be within sampling noise. Given the paper's central comparison is between preprocessing strategies and architectures, the absence of uncertainty quantification makes it impossible to know whether the observed differences are real. Please add bootstrap confidence intervals for the test-set metrics, or repeated cross-validation with multiple seeds, and state whether architecture and static-data comparisons are statistically distinguishable.
  3. [§4.2, 'Devising a new approach' and §5, Table 5.1] The claim that frequency-stratified subsetting 'removed complete class bias' is confounded by several simultaneous changes between the vanilla and fused pipelines. In moving from the initial LSTM/TCN setup to the fused setup, the authors also change the resampling interval per subset, introduce NaN masking, replace the single network with three parallel subnetworks, remove Ventilator Mode, and drop the SpO2:FiO2 and PaO2:FiO2 engineered ratios. Any one of these changes could be responsible for the observed improvement. To support the attribution, I request an ablation that varies one factor at a time: for example, a vanilla LSTM/TCN trained on uniformly resampled data but with the same masking, or a fused architecture trained on the original 30-minute uniform resampling. Without such experiments, the paper's headline causal claim is not established.
  4. [§4.2, Low/Medium/High frequency subsets and §3.2, synthetic data stratification] The paper does not report the actual proportion of synthetic (interpolated) values within each frequency subset. For the low-frequency subset, features are sampled on average about 0.5 times per 6-hour window and are resampled to a 2-hour interval (sequence length 4); with only one real observation, most time steps must be filled by the start/end imputation logic and linear interpolation. The authors stratify the train/test split by an overall synthetic-data proportion but never state the per-subset proportions that motivated the whole approach. Please quantify synthetic fractions per subset before and after the new strategy, and show that the 'low' subset is not still dominated by synthetic patterns. This is needed to substantiate the claim that the strategy actually minimised synthetic data impact rather than merely redistributing it.
minor comments (5)
  1. [Title page / Abstract] The document is formally a thesis ('A Thesis submitted in fulfillment...'), and the abstract contains the phrase 'This thesis highlights the challenges...'. For arXiv publication in cs.LG, please reframe as a research paper and remove thesis-specific formatting artifacts.
  2. [Throughout] There are numerous typographical errors and formatting glitches, including 'Extubation F ailure' in the title, 'Confustion matrix' in §4.1, 'F eatures' in section headings, and several run-together words such as 'W edecided'. A careful proofread is needed.
  3. [§3.1, Data extraction] The cohort flow is confusing: 5,970 patients before extraction become 4,701 after extraction, but the success/failure counts of 4,059 and 1,911 correspond to 5,970. Please present a clear CONSORT-style flow diagram with numbers at each exclusion stage, including the 1,269 patients with no recorded data.
  4. [Table 3.2 and §2.2] The text in §4.1 says features such as O2 saturation are recorded 'approximately once an hour,' and Table 3.2 confirms an average frequency of 6.6 per 6-hour window, which is consistent. However, the low-frequency features PH, PaCO2, and PaO2 are described as 'once every 12 hours' while the table lists about 0.5 per window; please harmonise the wording with the table.
  5. [§4.3, Implementation] A GitHub link is provided, but the paper does not describe the repository structure, software versions, or how to reproduce the experiments. Given the emphasis on rigor and a conditional acceptance, please add a reproducibility statement with environment details and a brief description of the release.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the study is an empirical supervised-learning pipeline whose outcome labels, feature sets, and model outputs are not mutually defining.

full rationale

The paper's derivation chain is empirical: it defines a clinical outcome (Section 3.1), extracts features from the external MIMIC-IV benchmark using literature and WAVE data, trains LSTM/TCN/LightGBM models on a held-out test split, and reports AUC-ROC and F1. No step renames a fitted parameter as a prediction, and no load-bearing result is imported from the authors' prior work, since the thesis contains no self-citations to prior papers. The only debatable construction is the Section 3.1 outcome definition that counts "O2 flow" or NIV within 6 hours post-extubation as failure, which raises the failure rate to 32.8% versus the 10–20% cited in Section 2.2; however, this is a clinically motivated label definition fixed before any model is trained, so it may threaten ground-truth validity but does not make the reported performance circular. All predictive claims are evaluated on a held-out test set, and feature selection uses external clinical data and published literature rather than the model's own outputs, so there is no circular derivation to flag.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central result depends on several hand-set preprocessing thresholds, a nonstandard outcome definition, and standard ML hyperparameters fit to validation data. No new physical or conceptual entities are introduced.

free parameters (5)
  • Feature observation frequency thresholds = 0.5 for Feature Sets 1 and 2; 0.15 for Feature Set 3
    Hand-chosen cutoffs for dropping low-observed features; they determine which features enter the models and how much synthetic data is created.
  • Frequency subset boundaries = low <1, medium 1-3, high >3 observations per 6-hour window
    Hand-chosen thresholds that group features for bespoke resampling; the exact boundaries are not derived from theory or prior literature.
  • Subset resampling intervals = low: 2 hours (sequence length 4), medium: 1 hour (sequence length 7), high: 30 minutes (sequence length 13)
    Chosen with clinical input to balance data density and synthetic data; they directly set the temporal granularity for each fused model.
  • Extubation failure definition windows = 48 hours for reintubation and mortality; 6 hours for ventilatory support
    The labeling rule that creates the outcome variable. Including oxygen flow or NIV within 6 hours as failure is a nonstandard choice that inflates the failure rate to 32.8%.
  • Model hyperparameters for LSTM, TCN, and FFNN = Vary per model, selected by Bayesian optimization with 100 trials
    Hidden dimensions, layer counts, dropout, learning rate, sampling method, and loss weighting are fit to validation AUC-ROC. This is standard practice, but the reported test performance inherits the choices.
assumptions (4)
  • domain assumption MIMIC-IV chart events accurately represent the patient's true physiological state during the 6-hour window.
    All model inputs come from MIMIC-IV chartevents; if recording is biased, all downstream predictions inherit that bias. Invoked in Section 3.1, Data extraction.
  • domain assumption Linear interpolation after imputing start and end values preserves clinically meaningful temporal trajectories.
    The paper relies on this simplification to create uniform sequences; it is explicitly acknowledged in Sections 4.1 and 4.2 as a practical choice rather than a validated representation.
  • domain assumption The 6-hour pre-extubation window is the clinically appropriate period for prediction.
    Chosen with clinician input, but not empirically validated against other window lengths; it determines all input data. Section 3.1, Data extraction.
  • domain assumption WAVE dataset features and literature popularity are valid proxies for feature predictive value.
    Feature selection is based on a cross-referencing exercise with the WAVE study and literature counts; no independent feature selection is performed on the MIMIC cohort. Section 3.1, Feature selection.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Predicting Extubation Failure in Intensive Care: The Development of a Novel, End-to-End Actionable and Interpretable Prediction System." pith.science (2026). https://pith.science/paper/F47BVRGT

@misc{pith2026241200105,
  author       = {Pith},
  title        = {Pith review of: Predicting Extubation Failure in Intensive Care: The Development of a Novel, End-to-End Actionable and Interpretable Prediction System},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F47BVRGT}},
  note         = {Machine review of arXiv:2412.00105}
}
read the original abstract

Predicting extubation failure in intensive care is challenging due to complex data and the severe consequences of inaccurate predictions. Machine learning shows promise in improving clinical decision-making but often fails to account for temporal patient trajectories and model interpretability, highlighting the need for innovative solutions. This study aimed to develop an actionable, interpretable prediction system for extubation failure using temporal modelling approaches such as Long Short-Term Memory (LSTM) and Temporal Convolutional Networks (TCN). A retrospective cohort study of 4,701 mechanically ventilated patients from the MIMIC-IV database was conducted. Data from the 6 hours before extubation, including static and dynamic features, were processed through novel techniques addressing data inconsistency and synthetic data challenges. Feature selection was guided by clinical relevance and literature benchmarks. Iterative experimentation involved training LSTM, TCN, and LightGBM models. Initial results showed a strong bias toward predicting extubation success, despite advanced hyperparameter tuning and static data inclusion. Data was stratified by sampling frequency to reduce synthetic data impacts, leading to a fused decision system with improved performance. However, all architectures yielded modest predictive power (AUC-ROC ~0.6; F1 <0.5) with no clear advantage in incorporating static data or additional features. Ablation analysis indicated minimal impact of individual features on model performance. This thesis highlights the challenges of synthetic data in extubation failure prediction and introduces strategies to mitigate bias, including clinician-informed preprocessing and novel feature subsetting. While performance was limited, the study provides a foundation for future work, emphasising the need for reliable, interpretable models to optimise ICU outcomes.

Figures

Figures reproduced from arXiv: 2412.00105 by the authors.

Figure 2
Figure 2. , from Goligher [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 2.1
Figure 2.1. Interaction between patient and ventilator. When in control mode, the ventilator bears [PITH_FULL_IMAGE:figures/full_fig_p009_2_1.png] view at source ↗
Figure 2.2
Figure 2.2. LSTM Memory Cell architecture. X: input, h: hidden state, C: cell state, [PITH_FULL_IMAGE:figures/full_fig_p024_2_2.png] view at source ↗
Figures from the paper (16 more)
Figure 2.3
Figure 2.3. Figure 2.3: Exemplar TCN network for binary classification. The network employs a kernel size of [PITH_FULL_IMAGE:figures/full_fig_p025_2_3.png]
Figure 3.1
Figure 3.1. Figure 3.1: Patient cohort extraction flow implementing clinically determined inclusion and exclu [PITH_FULL_IMAGE:figures/full_fig_p029_3_1.png]
Figure 4.1
Figure 4.1. Figure 4.1: Confusion matrix for initial LSTM model with Grid Search [PITH_FULL_IMAGE:figures/full_fig_p039_4_1.png]
Figure 4.2
Figure 4.2. Figure 4.2: Confusion matrix for initial LSTM model with Bayesian Optimisation [PITH_FULL_IMAGE:figures/full_fig_p041_4_2.png]
Figure 4.3
Figure 4.3. Figure 4.3: Architecture of the LSTM-Linear model for dynamic and static data processing. Dy [PITH_FULL_IMAGE:figures/full_fig_p043_4_3.png]
Figure 4.4
Figure 4.4. Figure 4.4: Confustion matrix of LSTM with static data included [PITH_FULL_IMAGE:figures/full_fig_p044_4_4.png]
Figure 4.5
Figure 4.5. Figure 4.5: Architecture of the Temporal Convolutional Network (TCN) model. Input time se [PITH_FULL_IMAGE:figures/full_fig_p045_4_5.png]
Figure 4.6
Figure 4.6. Figure 4.6: Decision fusion. All classifiers process input data in parallel and combine their outputs [PITH_FULL_IMAGE:figures/full_fig_p050_4_6.png]
Figure 4.7
Figure 4.7. Figure 4.7: Novel architecture of the Fused LSTM/TCN model for extubation failure prediction. [PITH_FULL_IMAGE:figures/full_fig_p050_4_7.png]
Figure 4.8
Figure 4.8. Figure 4.8: Confusion matrix for Fused LSTM post novel pre-processing [PITH_FULL_IMAGE:figures/full_fig_p052_4_8.png]
Figure 4.9
Figure 4.9. Figure 4.9: Confusion matrix for Fused TCN post novel pre-processing [PITH_FULL_IMAGE:figures/full_fig_p053_4_9.png]
Figure 4.10
Figure 4.10. Figure 4.10: Architecture of the novel Fused LSTM/TCN-FFNN model for dynamic and static [PITH_FULL_IMAGE:figures/full_fig_p054_4_10.png]
Figure 5.1
Figure 5.1. Figure 5.1: ROC curves for LSTM, TCN and LightGBM models trained on Feature Set 1 [PITH_FULL_IMAGE:figures/full_fig_p059_5_1.png]
Figure 5.2
Figure 5.2. Figure 5.2: Feature ablation values for Fused LSTM on dynamic only data (A) and Fused LSTM [PITH_FULL_IMAGE:figures/full_fig_p060_5_2.png]
Figure 5.3
Figure 5.3. Figure 5.3: Feature ablation values for Fused TCN on dynamic only data (A) and Fused TCN [PITH_FULL_IMAGE:figures/full_fig_p060_5_3.png]
Figure 6.1
Figure 6.1. Figure 6.1: Novel end-to-end pipeline to predict extubation failure. The pipeline outlines steps [PITH_FULL_IMAGE:figures/full_fig_p063_6_1.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

159 extracted references · 50 canonical work pages

  1. [1]

    Global prevalence of COVID-19-induced acute respiratory distress syndrome: systematic review and meta-analysis

    Azagew AW, Beko ZW, Ferede YM, Mekonnen HS, Abate HK, and Mekonnen CK. Global prevalence of COVID-19-induced acute respiratory distress syndrome: systematic review and meta-analysis. Systematic Reviews 2023; 12:1–9. doi: 10.1186/S13643-023-02377- 0/FIGURES/5. Available from:https://systematicreviewsjournal.biomedcentral.com/ articles/10.1186/s13643-023-02377-0

  2. [2]

    The changing epidemiology of mechanical ventilation: a population-based study

    Carson SS, Cox CE, Holmes GM, Howard A, and Carey TS. The changing epidemiology of mechanical ventilation: a population-based study. Journal of Intensive Care Medicine 2006 May; 21(3):173–83. Available from: https : / / go . gale . com / ps / i . do ? p = AONE & sw = w & issn=08850666&v=2.1&it=r&id=GALE%7CA146062867&sid=googleScholar&linkaccess= fulltext%...

  3. [3]

    https://www.who.int/news-room/fact-sheets/detail/ageing- and-health

    Ageing and Health. https://www.who.int/news-room/fact-sheets/detail/ageing- and-health. Accessed: 2024-09-06. 2022 Oct

  4. [4]

    Pathophysiology and classification of respiratory failure

    Lamba TS, Sharara RS, Singh AC, and Balaan M. Pathophysiology and classification of respiratory failure. Critical Care Nursing Quarterly 2016; 39(2):85–93.doi: 10.1097/CNQ. 0000000000000102. Available from:https://journals.lww.com/ccnq/fulltext/2016/ 04000/pathophysiology_and_classification_of_respiratory.2.aspx

  5. [5]

    Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries

    Bellani G, Laffey JG, Pham T, Fan E, Brochard L, Esteban A, Gattinoni L, Haren FMV, Larsson A, McAuley DF, Ranieri M, Rubenfeld G, Thompson BT, Wrigge H, Slutsky AS, and Pesenti A. Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. JAMA 2016 Feb; 315:788–800. doi: 10...

  6. [6]

    Long-term neurocognitive function after critical illness

    Hopkins RO and Jackson JC. Long-term neurocognitive function after critical illness. Chest 2006; 130(3):869–78. doi: 10.1378/CHEST.130.3.869 . Available from:https://pubmed. ncbi.nlm.nih.gov/16963688/

  7. [7]

    Acute Lung Injury and the Acute Respiratory Distress Syn- drome: Pathophysiology and Treatment

    Matuschak GM and Lechner AJ. Acute Lung Injury and the Acute Respiratory Distress Syn- drome: Pathophysiology and Treatment. The Journal of Missouri State Medical Association 2010

  8. [8]

    Clinical challenges in mechanical ventilation

    Goligher EC, Ferguson ND, and Brochard LJ. Clinical challenges in mechanical ventilation. Lancet (London, England) 2016 Apr; 387(10030):1856–66.doi: 10.1016/S0140-6736(16) 30176-3. Available from:https://pubmed.ncbi.nlm.nih.gov/27203509/

Show all 159 references
  1. [9]

    Mechanical Ventilation: State of the Art

    Pham T, Brochard LJ, and Slutsky AS. Mechanical Ventilation: State of the Art. Mayo Clinic proceedings 2017 Sep; 92(9):1382–400. doi: 10 . 1016 / J . MAYOCP . 2017 . 05 . 004. Available from:https://pubmed.ncbi.nlm.nih.gov/28870355/

  2. [10]

    Positive Pressure Ventilation

    Potchileev I, Doroshenko M, and Mohammed AN. Positive Pressure Ventilation. 2024

  3. [11]

    Invasive and non-invasive mechanical ventilation

    Popat B and Jones AT. Invasive and non-invasive mechanical ventilation. Medicine 2012 Jun; 40(6):298–304. doi: 10.1016/j.mpmed.2012.03.010 80 BIBLIOGRAPHY

  4. [12]

    The patient needing prolonged mechanical ventilation: a nar- rative review

    Ambrosino N and Vitacca M. The patient needing prolonged mechanical ventilation: a nar- rative review. Multidisciplinary Respiratory Medicine 2018 Dec; 13(1):6. doi: 10.1186/ s40248-018-0118-7

  5. [13]

    Weaning from mechanical ventilation

    Alía I and Esteban A. Weaning from mechanical ventilation. 2000.doi: 10.1186/cc660

  6. [14]

    Weaning Patients from the Ventilator

    McConville JF and Kress JP. Weaning Patients from the Ventilator. New England Journal of Medicine 2012 Dec; 367(23):2233–9. doi: 10 . 1056 / NEJMRA1203367 / SUPPL _ FILE / NEJMRA1203367_DISCLOSURES.PDF . Available from:https://www.nejm.org/doi/full/ 10.1056/nejmra1203367

  7. [15]

    Clinical practice, decision-making, and use of clinical decision support systems in invasivemechanicalventilation:anarrativereview.BritishJournalofAnaesthesia 2024 Apr; 0(0)

    Murali M, Ni M, Karbing DS, Rees SE, Komorowski M, Marshall D, Ramnarayan P, and Patel BV. Clinical practice, decision-making, and use of clinical decision support systems in invasivemechanicalventilation:anarrativereview.BritishJournalofAnaesthesia 2024 Apr; 0(0). doi: 10.101...

  8. [16]

    The decision to extubate in the intensive care unit

    Thille AW, Richard JCM, and Brochard L. The decision to extubate in the intensive care unit. American journal of respiratory and critical care medicine 2013 Jun; 187(12):1294–

  9. [17]

    Machine learning for predicting successful extubation in patients receiving mechanical ventilation

    Igarashi Y, Ogawa K, Nishimura K, Osawa S, Ohwada H, and Yokobori S. Machine learning for predicting successful extubation in patients receiving mechanical ventilation. Frontiers in Medicine 2022 Aug; 9. doi: 10 . 3389 / fmed . 2022 . 961252. Available from: https : //www.ncbi...

  10. [18]

    Effect of Failed Extubation on the Outcome of Mechanical Ventilation

    Epstein SK, Ciubotaru RL, and Wong JB. Effect of Failed Extubation on the Outcome of Mechanical Ventilation. Chest 1997 Jul; 112(1):186–92.doi: 10.1378/chest.112.1.186

  11. [19]

    Risk factors for and prediction by caregivers of extubation failure in ICU patients: a prospective study

    Thille AW, Boissier F, Ghezala HB, Razazi K, Mekontso-Dessap A, and Brun-Buisson C. Risk factors for and prediction by caregivers of extubation failure in ICU patients: a prospective study. Critical care medicine 2015 Mar; 43(3):613–20. doi: 10 . 1097 / CCM . 0000000000000748....

  12. [20]

    Variation in extubation failure rates after neonatal congenital heart surgery across Pediatric Cardiac Critical Care Consortium hospitals

    Benneyworth BD, Mastropietro CW, Graham EM, Klugman D, Costello JM, Zhang W, and Gaies M. Variation in extubation failure rates after neonatal congenital heart surgery across Pediatric Cardiac Critical Care Consortium hospitals. The Journal of thoracic and cardiovascular surge...

  13. [21]

    Extubation Failure in Brain-injured Patients: Risk Factors and Development of a Prediction Score in a Preliminary Prospective Cohort Study

    Godet T, Chabanne R, Marin J, Kauffmann S, Futier E, Pereira B, and Constantin JM. Extubation Failure in Brain-injured Patients: Risk Factors and Development of a Prediction Score in a Preliminary Prospective Cohort Study. Anesthesiology 2017 Jan; 126(1):104–14. doi: 10.1097/A...

  14. [22]

    Ventilator Weaning and Spontaneous Breathing Trials; an Educational Review

    Zein H, Baratloo A, Negida A, and Safari S. Ventilator Weaning and Spontaneous Breathing Trials; an Educational Review. Emergency 2016; 4(2):65. Available from:/pmc/articles/ PMC4893753/%20/pmc/articles/PMC4893753/?report=abstract%20https://www.ncbi. nlm.nih.gov/pmc/articles/P...

  15. [23]

    Prolonged Mechanical Ventilation: Outcomes and Man- agement

    Huang HY, Huang CY, and Li LF. Prolonged Mechanical Ventilation: Outcomes and Man- agement. Journal of Clinical Medicine 2022 May; 11(9). doi: 10 . 3390 / JCM11092451. Available from: /pmc/articles/PMC9103623/%20/pmc/articles/PMC9103623/?report= abstract%20https://www.ncbi.nlm...

  16. [24]

    To extubate or not to extubate: Risk factors for extubation failure and deterioration with further mechanical ventilation

    Xie J, Cheng G, Zheng Z, Luo H, Oon |, and Ooi C. To extubate or not to extubate: Risk factors for extubation failure and deterioration with further mechanical ventilation. J Card Surg 2019; 34:1004–11. doi: 10 . 1111 / jocs . 14189. Available from: https : / / onlinelibrary.w...

  17. [25]

    Experience with a step-down respiratory care center at a tertiary referral medical center in Taiwan

    Su J, Lin CY, Chen PJ, Lin FJ, Chen SK, and Kuo HT. Experience with a step-down respiratory care center at a tertiary referral medical center in Taiwan. Journal of critical care 2006 Jun; 21(2):156–61. doi: 10 . 1016 / J . JCRC . 2005 . 10 . 004. Available from: https://pubmed...

  18. [26]

    The efficacy and effectiveness of machine learning for weaning in mechanically ventilated patients at the intensive care unit: a systematic review

    Kwong MT, Colopy GW, Weber AM, Ercole A, Bergmann JHM, and Bergmann BHJM. The efficacy and effectiveness of machine learning for weaning in mechanically ventilated patients at the intensive care unit: a systematic review. Bio-des. Manuf. 2019; 2:31–40.doi: 10.1007/s42242-018-0...

  19. [27]

    Personalized mechanical ventilation in acute respiratory distress syndrome

    Pelosi P, Ball L, Barbas CSV, Bellomo R, Burns KEA, Einav S, Gattinoni L, Laffey JG, Marini JJ, Myatra SN, Schultz MJ, Teboul JL, and Rocco PRM. Personalized mechanical ventilation in acute respiratory distress syndrome. Critical Care 2021 Dec; 25(1):250.doi: 10.1186/s13054-02...

  20. [28]

    Medical decision support in me- chanical ventilation employing combined model information of gas exchange and respiratory mechanics

    Kretschmer J, Riedlinger A, Schranz C, and Möller K. Medical decision support in me- chanical ventilation employing combined model information of gas exchange and respiratory mechanics. IFAC Proceedings Volumes 2014; 47(3):9919–23.doi: 10.3182/20140824-6- ZA-1003.01431

  21. [29]

    The role of computer-based clinical decision support systems to deliver protective mechanical ventilation

    Khemani RG, Hotz JC, Sward KA, and Newth CJ. The role of computer-based clinical decision support systems to deliver protective mechanical ventilation. Current opinion in critical care 2020 Feb; 26(1):73–81.doi: 10.1097/MCC.0000000000000688. Available from: https://pubmed.ncbi...

  22. [30]

    Clinical Verification of A Clinical Decision Support System for Ventilator Weaning

    Hsu JC, Chen YF, Chung WS, Tan TH, Chen T, and Chiang JY. Clinical Verification of A Clinical Decision Support System for Ventilator Weaning. BioMedical Engineering Online 2013 Dec; 12(SUPPL 1):1–14. doi: 10.1186/1475-925X-12-S1-S4/TABLES/5 . Available from: https://biomedical...

  23. [31]

    An Open-Loop, Physiologic Model–Based DecisionSupportSystemCanProvideAppropriateVentilatorSettings.CriticalCareMedicine 2018 Jul; 46(7):e642–e648

    Karbing DS, Spadaro S, Dey N, Ragazzi R, Marangoni E, Corte FD, Moro F, Lodahl D, Hansen NS, Winding R, Rees SE, and Volta CA. An Open-Loop, Physiologic Model–Based DecisionSupportSystemCanProvideAppropriateVentilatorSettings.CriticalCareMedicine 2018 Jul; 46(7):e642–e648. doi...

  24. [32]

    CDSS-RM: a clinical decision support system reference model

    Zikos D and Delellis N. CDSS-RM: a clinical decision support system reference model. BMC Medical Research Methodology 2018 Nov; 18(1). doi: 10 . 1186 / S12874 - 018 - 0587 - 6. Available from: /pmc/articles/PMC6240189/%20/pmc/articles/PMC6240189/?report= abstract%20https://www...

  25. [33]

    Barriers to implementation of a computerized decision support system for depression: an observa- tional report on lessons learned in ”real world” clinical settings

    Trivedi MH, Daly EJ, Kern JK, Grannemann BD, Sunderajan P, and Claassen CA. Barriers to implementation of a computerized decision support system for depression: an observa- tional report on lessons learned in ”real world” clinical settings. BMC Medical Informatics and Decision...

  26. [34]

    Understanding Decision Making in Critical Care

    Lighthall GK and Vazquez-Guillamet C. Understanding Decision Making in Critical Care. Clinical Medicine and Research 2015; 13:156–68.doi: 10.3121/cmr.2015.1289

  27. [35]

    A look into the nature and causes of human errors in the intensive care unit

    Donchin Y, Gopher D, Olin M, Badihi Y, Biesky MR, Sprung CL, Pizov R, and Cotev S. A look into the nature and causes of human errors in the intensive care unit. Critical Care Medicine 1995 Feb; 23(2):294–300. doi: 10.1097/00003246-199502000-00015

  28. [36]

    Documentation, Data, and Decision-Making

    Munro CL and Swamy L. Documentation, Data, and Decision-Making. American Journal of Critical Care 2024 May; 33(3):162–5. doi: 10 . 4037 / AJCC2024617. Available from: www.ajcconline.org

  29. [37]

    Ventilator Management

    Carpio ALM and Mora JI. Ventilator Management. 2024

  30. [38]

    Machine Learning for Mechanical Ventilation Control

    Suo D, Agarwal N, Xia W, Chen X, Ghai U, Yu A, Gradu P, Singh K, Zhang C, Minasyan E, LaChance J, Zajdel T, Schottdorf M, Cohen D, and Hazan E. Machine Learning for Mechanical Ventilation Control. 2022. arXiv:2102.06779 [cs.LG]. Available from:https: //arxiv.org/abs/2102.06779

  31. [39]

    Developing an explainable machine learning model to predict the mechanical ventilation duration of patients with ARDS in intensive care units

    Wang Z, Zhang L, Huang T, Yang R, Cheng H, Wang H, Yin H, and Lyu J. Developing an explainable machine learning model to predict the mechanical ventilation duration of patients with ARDS in intensive care units. Heart and Lung 2023 Mar; 58:74.doi: 10. 1016/J.HRTLNG.2022.11.005...

  32. [40]

    A novel machine learning model to predict respiratory failure and invasive mechanical ventilation in critically ill patients suffering from COVID-19

    Bendavid I, Statlender L, Shvartser L, Teppler S, Azullay R, Sapir R, and Singer P. A novel machine learning model to predict respiratory failure and invasive mechanical ventilation in critically ill patients suffering from COVID-19. Scientific Reports | 123; 12:10573.doi: 10....

  33. [41]

    Decision to extubate

    Epstein S. Decision to extubate. Intensive Care Medicine 2002 May; 28(5):535–46. doi: 10.1007/s00134-002-1268-8

  34. [42]

    Machine Learning for Prediction of Successful Extubation of Mechanical Ventilated Patients in an Intensive Care Unit: A Retrospective Observational Study

    Otaguro T, Tanaka H, Igarashi Y, Tagami T, Masuno T, Yokobori S, Matsumoto H, Ohwada H, and Yokota H. Machine Learning for Prediction of Successful Extubation of Mechanical Ventilated Patients in an Intensive Care Unit: A Retrospective Observational Study. Journal of Nippon Me...

  35. [43]

    Mechanical Ventilation to Minimize Progression of Lung Injury in Acute Respiratory Failure

    Brochard L, Slutsky A, and Pesenti A. Mechanical Ventilation to Minimize Progression of Lung Injury in Acute Respiratory Failure. American journal of respiratory and critical care medicine 2017 Feb; 195(4):438–42. doi: 10.1164/RCCM.201605-1081CP . Available from: https://pubme...

  36. [44]

    Improvement in the Prediction of Ventilator Weaning Outcomes by an Artificial Neural Network in a Medical ICU

    Kuo HJ, Chiu HW, Lee CN, Chen TT, Chang CC, and Bien MY. Improvement in the Prediction of Ventilator Weaning Outcomes by an Artificial Neural Network in a Medical ICU. Respiratory Care 2015 Nov; 60(11):1560–9.doi: 10.4187/RESPCARE.03648. Available from: https://rc.rcjournal.co...

  37. [45]

    An Artificial Neural Network Model for Predicting Successful Extubation in Intensive Care Units

    Hsieh MH, Hsieh MJ, Chen CM, Hsieh CC, Chao CM, and Lai CC. An Artificial Neural Network Model for Predicting Successful Extubation in Intensive Care Units. Journal of Clinical Medicine 2018 Aug; 7(9):240. doi: 10.3390/jcm7090240

  38. [46]

    Rapid shallow breathing index

    Karthika M, Enezi FAA, Pillai LV, and Arabi YM. Rapid shallow breathing index. Annals of Thoracic Medicine 2016 Jul; 11(3):167. doi: 10.4103/1817-1737.176876 . Available from: /pmc/articles/PMC4966218/%20/pmc/articles/PMC4966218/?report=abstract% 20https://www.ncbi.nlm.nih.gov...

  39. [47]

    The use of the area under the ROC curve in the evaluation of machine learning algorithms

    Bradley AP. The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognition 1997 Jul; 30(7):1145–59. doi: 10.1016/S0031-3203(96) 00142-2

  40. [48]

    Therapeutic Intervention Scoring System (TISS) - a method for calculating costs in the intensive care unit (ICU) and intermediate care unit (IMCU)

    Kaufmann I and Briegel J. Therapeutic Intervention Scoring System (TISS) - a method for calculating costs in the intensive care unit (ICU) and intermediate care unit (IMCU). Critical Care 2000; 4(Suppl 1):P243.doi: 10.1186/CC962. Available from:/pmc/articles/ PMC3333165/%20htt...

  41. [49]

    Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units

    Zhao QY, Wang H, Luo JC, Luo MH, Liu LP, Yu SJ, Liu K, Zhang YJ, Sun P, Tu GW, and Luo Z. Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units. Frontiers in Medicine 2021 May; 8. doi: 10.3389/fmed.2021.676343

  42. [50]

    Prediction of Extubation Failure for Intensive Care Unit Patients Using Light Gradient Boosting Machine

    Chen T, Xu J, Ying H, Chen X, Feng R, Fang X, Gao H, and Wu J. Prediction of Extubation Failure for Intensive Care Unit Patients Using Light Gradient Boosting Machine. IEEE Access 2019; 7:150960–8. doi: 10.1109/ACCESS.2019.2946980

  43. [51]

    CatBoost for big data: an interdisciplinary review

    Hancock JT and Khoshgoftaar TM. CatBoost for big data: an interdisciplinary review. Journal of Big Data 2020 Dec; 7(1):1–45.doi: 10.1186/S40537-020-00369-8/FIGURES/9. Available from: https : / / journalofbigdata . springeropen . com / articles / 10 . 1186 / s40537-020-00369-8

  44. [52]

    Interpretation of machine learning models using shapley values: application to compound potency and multi-target activity predictions

    Rodríguez-Pérez R and Bajorath J. Interpretation of machine learning models using shapley values: application to compound potency and multi-target activity predictions. Journal of Computer-Aided Molecular Design 2020 Oct; 34(10):1013–26.doi: 10.1007/s10822-020- 00314-0

  45. [53]

    LightGBM: A Highly Efficient Gradient Boosting Decision Tree

    Ke G, Meng Q, Finley T, Wang T, Chen W, Ma W, Ye Q, and Liu TY. LightGBM: A Highly Efficient Gradient Boosting Decision Tree. Advances in Neural Information Process- ing Systems 2017; 30. Available from:https://github.com/Microsoft/LightGBM

  46. [54]

    A Machine Learning decision-making tool for extubation in Intensive Care Unit patients

    Fabregat A, Magret M, Ferré JA, Vernet A, Guasch N, Rodríguez A, Gómez J, and Bodí M. A Machine Learning decision-making tool for extubation in Intensive Care Unit patients. Computer Methods and Programs in Biomedicine 2021 Mar; 200:105869.doi: 10.1016/ J.CMPB.2020.105869

  47. [55]

    Developing a machine-learning model for real-time prediction of successful extubation in mechanically ventilated patients using time-series ventilator-derived parameters

    Huang KY, Hsu YL, Chen HC, Horng MH, Chung CL, Lin CH, Xu JL, and Hou MH. Developing a machine-learning model for real-time prediction of successful extubation in mechanically ventilated patients using time-series ventilator-derived parameters. Frontiers in Medicine 2023; 10:1...

  48. [56]

    Predictors for extubation failure in COVID-19 patients using a machine learning approach

    Fleuren LM et al. Predictors for extubation failure in COVID-19 patients using a machine learning approach. Critical Care 2021 Dec; 25(1):448.doi: 10.1186/s13054-021-03864-3 84 BIBLIOGRAPHY

  49. [57]

    Scientific Reports 2024 14:1 2024 Feb; 14(1):1–30

    PotharlankaJLandBhatMN.FeatureimportancefeedbackwithDeepQprocessinensemble- based metaheuristic feature selection algorithms. Scientific Reports 2024 14:1 2024 Feb; 14(1):1–30. doi: 10.1038/s41598-024-53141-w . Available from:https://www.nature. com/articles/s41598-024-53141-w

  50. [58]

    Enforcing Interpretability and its Statistical Im- pacts: Trade-offs between Accuracy and Interpretability

    Dziugaite GK, Ben-David S, and Roy DM. Enforcing Interpretability and its Statistical Im- pacts: Trade-offs between Accuracy and Interpretability. 2020. arXiv:2010.13764 [cs.LG]. Available from:https://arxiv.org/abs/2010.13764

  51. [59]

    A Unified Approach to Interpreting Model Predictions

    Lundberg SM and Lee SI. A Unified Approach to Interpreting Model Predictions. Advances in Neural Information Processing Systems 2017 May; 2017-December:4766–75. Available from: https://arxiv.org/abs/1705.07874v2

  52. [60]

    Employing feature engineering strategies to improve the performance of machine learning algorithms on echocardiogram dataset

    Huang HN, Chen HM, Lin WW, Huang CJ, Chen YC, Wang YH, and Yang CT. Employing feature engineering strategies to improve the performance of machine learning algorithms on echocardiogram dataset. Digital Health 2023 Jan; 9. doi: 10.1177/20552076231207589. Available from:/pmc/art...

  53. [61]

    Re-Thinking Data Strategy and Integration for Artificial Intelligence: Concepts, Opportunities, and Challenges

    Aldoseri A, Al-Khalifa KN, and Hamouda AM. Re-Thinking Data Strategy and Integration for Artificial Intelligence: Concepts, Opportunities, and Challenges. Applied Sciences 2023

  54. [62]

    Minute ventilation recovery time: a predictor of extubation outcome

    Martinez A, Seymour C, and Nam M. Minute ventilation recovery time: a predictor of extubation outcome. Chest 2003 Apr; 123(4):1214–21. doi: 10.1378/CHEST.123.4.1214. Available from:https://pubmed.ncbi.nlm.nih.gov/12684314/

  55. [63]

    Available from:https://www.mdpi.com/2076-3417/13/ 12/7082

    doi: 10.3390/app13127082. Available from:https://www.mdpi.com/2076-3417/13/ 12/7082

  56. [64]

    Designing an Interpretability-Based Model to Explain the Arti- ficial Intelligence Algorithms in Healthcare

    Ennab M and McHeick H. Designing an Interpretability-Based Model to Explain the Arti- ficial Intelligence Algorithms in Healthcare. Diagnostics 2022 Jul; 12(7). doi: 10.3390/ DIAGNOSTICS12071557. Available from:/pmc/articles/PMC9319389/%20/pmc/articles/ PMC9319389/?report=abst...

  57. [65]

    Chest 2003 Jul; 124(1):262–8.doi: 10.1378/CHEST

    SminaM,SalamA,KhamieesM,GadaP,Amoateng-AdjepongY,andManthousCA.Cough peak flows and extubation outcomes. Chest 2003 Jul; 124(1):262–8.doi: 10.1378/CHEST. 124.1.262. Available from:https://pubmed.ncbi.nlm.nih.gov/12853532/

  58. [66]

    Deep Learning for Time Series Fore- casting: Advances and Open Problems

    Casolaro A, Capone V, Iannuzzo G, and Camastra F. Deep Learning for Time Series Fore- casting: Advances and Open Problems. Information 2023, Vol. 14, Page 598 2023 Nov; 14(11):598. doi: 10.3390/INFO14110598. Available from:https://www.mdpi.com/2078- 2489/14/11/598/htm%20https:...

  59. [67]

    Analyzing Patient Trajectories with Artificial Intelligence

    Allam A, Feuerriegel S, Rebhan M, and Krauthammer M. Analyzing Patient Trajectories with Artificial Intelligence. Journal of Medical Internet Research 2021 Dec; 23(12):e29812. doi: 10.2196/29812. Available from:https://www.jmir.org/2021/12/e29812

  60. [68]

    Technology investigation on time series classification and prediction

    Tong Y, Liu J, Yu L, Zhang L, Sun L, Li W, Ning X, Xu J, Qin H, and Cai Q. Technology investigation on time series classification and prediction. PeerJ Computer Science 2022 May; 8:e982. doi: 10.7717/peerj-cs.982 BIBLIOGRAPHY 85

  61. [69]

    Generalizability of Machine Learning Models: Quantitative Evaluation of Three Methodological Pitfalls

    Maleki F, Ovens K, Gupta R, Reinhold C, Spatz A, and Forghani R. Generalizability of Machine Learning Models: Quantitative Evaluation of Three Methodological Pitfalls. Radi- ology: Artificial Intelligence 2023 Jan; 5(1). doi: 10.1148/ryai.220028

  62. [70]

    Long Short-Term Memory

    Hochreiter S and Schmidhuber J. Long Short-Term Memory. Neural Computation 1997 Nov; 9(8):1735–80. doi: 10.1162/NECO.1997.9.8.1735. Available from:https://dx.doi. org/10.1162/neco.1997.9.8.1735

  63. [71]

    Recurrent neural networks and time series prediction

    Connor J and Atlas L. Recurrent neural networks and time series prediction. Proceedings. IJCNN-91-Seattle: International Joint Conference on Neural Networks 1991 :301–6.doi: 10.1109/IJCNN.1991.155194

  64. [72]

    Long Short-Term Memory Recurrent Neural Network for Automatic Speech Recognition

    Oruh J, Viriri S, and Adegun A. Long Short-Term Memory Recurrent Neural Network for Automatic Speech Recognition. IEEE Access 2022; 10:30069–79. doi: 10.1109/ACCESS. 2022.3159339

  65. [73]

    Natural language processing: state of the art, current trends and challenges

    Khurana D, Koli A, Khatter K, and Singh S. Natural language processing: state of the art, current trends and challenges. Multimedia Tools and Applications 2023 Jan; 82(3):3713–44. doi: 10.1007/S11042-022-13428-4/FIGURES/3. Available from:https://link.springer. com/article/10.1...

  66. [74]

    Stock price prediction with optimized deep LSTM network with artificial rabbits optimization algorithm

    Gülmez B. Stock price prediction with optimized deep LSTM network with artificial rabbits optimization algorithm. Expert Systems with Applications 2023 Oct; 227:120346. doi: 10.1016/J.ESWA.2023.120346

  67. [75]

    Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation

    Cho K, Merrienboer B van, Gulcehre C, Bahdanau D, Bougares F, Schwenk H, and Bengio Y. Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation. 2014. arXiv: 1406.1078 [cs.CL]. Available from:https://arxiv.org/abs/ 1406.1078

  68. [76]

    Convolutional neural networks: an overview and application in radiology

    Yamashita R, Nishio M, Kinh R, Do G, and Togashi K. Convolutional neural networks: an overview and application in radiology. Insights Imaging 2018.doi: 10.1007/s13244-018- 0639-9. Available from:https://doi.org/10.1007/s13244-018-0639-9

  69. [77]

    Temporal Convolutional Networks: A Unified Approach to Action Segmentation

    Lea C, Vidal R, Reiter A, and Hager GD. Temporal Convolutional Networks: A Unified Approach to Action Segmentation. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2016 Aug; 9915 LNCS:47–54. d...

  70. [78]

    Partial dependence through stratification

    Parr T and Wilson JD. Partial dependence through stratification. Machine Learning with Applications 2021 Dec; 6:100146. doi: 10.1016/J.MLWA.2021.100146

  71. [79]

    Temporal convolutional networks allow early prediction of events in critical care

    Catling FJR and Wolff AH. Temporal convolutional networks allow early prediction of events in critical care. Journal of the American Medical Informatics Association 2020 Mar; 27(3):355–65. doi: 10.1093/jamia/ocz205

  72. [80]

    Interpretable recurrent neural network models for dynamic prediction of the extubation failure risk in patients with invasive mechanical ventilation in the intensive care unit

    Zeng Z, Tang X, Liu Y, He Z, and Gong X. Interpretable recurrent neural network models for dynamic prediction of the extubation failure risk in patients with invasive mechanical ventilation in the intensive care unit. BioData Mining 2022 Sep; 15(1):21.doi: 10.1186/ s13040-022-...

  73. [81]

    Developing and validating a machine learning model to predict successful next-day extubation in the ICU

    Fenske SW, Peltekian A, Kang M, Markov NS, Zhu M, Grudzinski K, Bak MJ, Pawlowski A, Gupta V, Mao Y, Bratchikov S, Stoeger T, Rasmussen LV, Choudhary AN, Misharin AV, Singer BD, Budinger GS, Wunderink RG, Agrawal A, Gao CA, and NU SCRIPT Study Investigators the. Developing and...

  74. [82]

    Fallacies of last observation carried forward analyses

    Lachin JM. Fallacies of last observation carried forward analyses. Clinical trials (London, England) 2016 Apr; 13(2):161–8.doi: 10.1177/1740774515602688. Availablefrom:https: //pubmed.ncbi.nlm.nih.gov/26400875/

  75. [83]

    Effect of sequence padding on the perfor- mance of deep learning models in archaeal protein functional prediction

    Rio ALd, Martin M, Perera-Lluna A, and Saidi R. Effect of sequence padding on the perfor- mance of deep learning models in archaeal protein functional prediction. Scientific Reports 2020 Sep; 10(1):14634. doi: 10.1038/s41598-020-71450-8

  76. [84]

    Moving from bytes to bedside: a systematic review on the use of artificial intelligence in the intensive care unit

    Sande D van de, Genderen ME van, Huiskens J, Gommers D, and Bommel J van. Moving from bytes to bedside: a systematic review on the use of artificial intelligence in the intensive care unit. Intensive care medicine 2021 Jul; 47(7):750–60. doi: 10.1007/S00134-021- 06446-7. Avail...

  77. [85]

    Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

    Chung J, Gulcehre C, Cho K, and Bengio Y. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. 2014 Dec

  78. [86]

    The perceptron: A probabilistic model for information storage and organiza- tion in the brain

    Rosenblatt F. The perceptron: A probabilistic model for information storage and organiza- tion in the brain. Psychological Review 1958 Nov; 65(6):386–408.doi: 10.1037/H0042519

  79. [87]

    doi: 10.1007/BF02478259/ METRICS

    McCullochWSandPittsW.Alogicalcalculusoftheideasimmanentinnervousactivity.The Bulletin of Mathematical Biophysics 1943 Dec; 5(4):115–33. doi: 10.1007/BF02478259/ METRICS. Available from:https://link.springer.com/article/10.1007/BF02478259

  80. [88]

    Activation Functions in Artificial Neural Networks: A Systematic Overview

    Lederer J. Activation Functions in Artificial Neural Networks: A Systematic Overview. arXiv.org 2021

  81. [89]

    Deep Learning: A Comprehensive Overview on Techniques, Taxonomy, Appli- cations and Research Directions

    Sarker IH. Deep Learning: A Comprehensive Overview on Techniques, Taxonomy, Appli- cations and Research Directions. SN Computer Science 2021 Nov; 2(6):1–20. doi: 10 . 1007/S42979-021-00815-1/FIGURES/11 . Available from:https://link.springer.com/ article/10.1007/s42979-021-00815-1

  82. [90]

    Conceptual Understanding of Convolu- tional Neural Network- A Deep Learning Approach

    Indolia S, Goswami AK, Mishra SP, and Asopa P. Conceptual Understanding of Convolu- tional Neural Network- A Deep Learning Approach. Procedia Computer Science 2018 Jan; 132:679–88. doi: 10.1016/J.PROCS.2018.05.069

  83. [91]

    A review on the long short-term memory model

    Houdt GV, Mosquera C, and Nápoles G. A review on the long short-term memory model. ArtificialIntelligenceReview 2020 Dec;53(8):5929–55. doi: 10.1007/S10462-020-09838-1

  84. [92]

    Ensemble Dilated Convolutional Neural Network and Its Application in Rotating Machinery Fault Diagnosis

    Cai Y, Wang Z, Yao L, Lin T, and Zhang J. Ensemble Dilated Convolutional Neural Network and Its Application in Rotating Machinery Fault Diagnosis. Computational Intelligence and Neuroscience 2022; 2022. doi: 10.1155/2022/6316140 . Available from:/pmc/articles/ PMC9519275/%20/p...

  85. [93]

    An adversarial discriminative temporal convolutional network for EEG-based cross-domain emotion recognition

    He Z, Zhong Y, and Pan J. An adversarial discriminative temporal convolutional network for EEG-based cross-domain emotion recognition. Computers in Biology and Medicine 2022 Feb; 141:105048.doi: 10.1016/j.compbiomed.2021.105048

  86. [94]

    MIMIC-IV (version 3.0)

    Johnson A, Bulgarelli L, Pollard T, Gow B, Moody B, Horng S, Celi LA, and Mark R. MIMIC-IV (version 3.0). 2024

  87. [95]

    Temporal con- volutional networks and data rebalancing for clinical length of stay and mortality prediction

    Bednarski BP, Singh AD, Zhang W, Jones WM, Naeim A, and Ramezani R. Temporal con- volutional networks and data rebalancing for clinical length of stay and mortality prediction. Scientific Reports 2022 Dec; 12(1):21247. doi: 10.1038/s41598-022-25472-z

  88. [96]

    PhysioBank, PhysioToolkit, and PhysioNet: compo- nents of a new research resource for complex physiologic signals

    Goldberger AL, Amaral LA, Glass L, Hausdorff JM, Ivanov PC, Mark RG, Mietus JE, Moody GB, Peng CK, and Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: compo- nents of a new research resource for complex physiologic signals. Circulation 2000; 101(23). doi: 10.1161/01.CIR....

  89. [97]

    Lehman L wei, Celi LA, and Mark RG

    Johnson AE, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, Pollard TJ, Moody B, Gow B, H. Lehman L wei, Celi LA, and Mark RG. MIMIC-IV, a freely accessible electronic health record dataset. Scientific Data 2023 Dec; 10(1).doi: 10.1038/S41597-022-01899-X BIBLIOGRAPHY 87

  90. [98]

    Decision support system to evaluate ventilation in the acute respiratory distress syndrome (DeVENT study)—trial protocol

    Patel B, Mumby S, Johnson N, Falaschetti E, Hansen J, Adcock I, McAuley D, Takata M, Karbing DS, Jabaudon M, Schellengowski P, and Rees SE. Decision support system to evaluate ventilation in the acute respiratory distress syndrome (DeVENT study)—trial protocol. Trials 2022 Dec...

  91. [99]

    Rees SE and Karbing DS. Model-based advice for mechanical ventilation: From research (INVENT) to product (Beacon Caresystem).2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) . IEEE, 2015 Aug :5331–4. doi: 10.1109/EMBC.201...

  92. [100]

    A new method of classifying prognostic comorbidityinlongitudinalstudies:Developmentandvalidation.JournalofChronicDiseases 1987 Jan; 40(5):373–83

    Charlson ME, Pompei P, Ales KL, and MacKenzie C. A new method of classifying prognostic comorbidityinlongitudinalstudies:Developmentandvalidation.JournalofChronicDiseases 1987 Jan; 40(5):373–83. doi: 10.1016/0021-9681(87)90171-8

  93. [101]

    Prediction of Mortality in Incident Hemodialysis Patients: A Validation and Comparison of CHADS2, CHA2DS2, and CCI Scores

    Yang H, Chen YH, Hsieh TF, Chuang SY, and Wu MJ. Prediction of Mortality in Incident Hemodialysis Patients: A Validation and Comparison of CHADS2, CHA2DS2, and CCI Scores. PLOS ONE 2016 May; 11(5):e0154627. doi: 10.1371/journal.pone.0154627

  94. [102]

    A survey on imbalanced learning: latest re- search, applications and future directions

    Chen W, Yang K, Yu Z, Shi Y, and Chen CL. A survey on imbalanced learning: latest re- search, applications and future directions. Artificial Intelligence Review 2024 Jun; 57(6):1–

  95. [103]

    Health Services and Delivery Research 2015

    HarrisonDA,Ferrando-VivasP,ShahinJ,andRowanKM.Externalvalidationofthecurrent Intensive Care National Audit and Research Centre model in Scottish critical care units. Health Services and Delivery Research 2015. Available from:https://www.ncbi.nlm.nih. gov/books/NBK321618/

  96. [104]

    The receiver operating characteristic curve accurately assesses imbalanced datasets

    Richardson E, Trevizani R, Greenbaum JA, Carter H, Nielsen M, and Peters B. The receiver operating characteristic curve accurately assesses imbalanced datasets. Patterns 2024 Jun; 5(6):100994. doi: 10.1016/J.PATTER.2024.100994

  97. [105]

    1007/ S10462- 024- 10759 -6 /FIGURES /11

    doi: 10. 1007/ S10462- 024- 10759 -6 /FIGURES /11. Available from: https: // link. springer.com/article/10.1007/s10462-024-10759-6

  98. [106]

    Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning

    Lemaître G, Nogueira F, and Aridas CK. Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning. Journal of Machine Learning Research 2016 Sep; 18:1–5. Available from:https://arxiv.org/abs/1609.06570v1

  99. [107]

    Optuna: A Next-generation Hy- perparameter Optimization Framework

    Akiba T, Sano S, Yanase T, Ohta T, and Koyama M. Optuna: A Next-generation Hy- perparameter Optimization Framework. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 2019 Jul :2623–31.doi: 10.1145/ 3292500.3330701. Available from:htt...

  100. [108]

    Empirical Study of Overfitting in Deep Learning for Pre- dictingBreastCancerMetastasis.Cancers 2023 Apr;15(7)

    Xu C, Coen-Pirani P, and Jiang X. Empirical Study of Overfitting in Deep Learning for Pre- dictingBreastCancerMetastasis.Cancers 2023 Apr;15(7). doi: 10.3390/CANCERS15071969/ S1. Available from: /pmc / articles / PMC10093528 / %20 / pmc / articles / PMC10093528 / ?report=abstr...

  101. [109]

    Random Search for Hyper-Parameter Optimization Yoshua Bengio

    Bergstra J, Ca JB, and Ca YB. Random Search for Hyper-Parameter Optimization Yoshua Bengio. Journal of Machine Learning Research 2012; 13:281–305. Available from:http : //scikit-learn.sourceforge.net

  102. [110]

    Bayesian Optimization is Superior to Random Search for Machine Learning Hyperparameter Tuning: Analysis of the Black-Box Optimization Challenge 2020

    Turner R, Eriksson D, McCourt M, Kiili J, Laaksonen E, Xu Z, and Guyon I. Bayesian Optimization is Superior to Random Search for Machine Learning Hyperparameter Tuning: Analysis of the Black-Box Optimization Challenge 2020. Proceedings of Machine Learning Research 2021 Apr; 13...

  103. [111]

    A Taxonomy of Global Optimization Methods Based on Response Surfaces

    Jones DR. A Taxonomy of Global Optimization Methods Based on Response Surfaces. Journal of Global Optimization 2001 Dec; 21(4):345–83.doi: 10.1023/A:1012771025575

  104. [112]

    Hyperparameter Tuning for Machine Learn- ing Algorithms Used for Arabic Sentiment Analysis

    Elgeldawi E, Sayed A, Galal AR, and Zaki AM. Hyperparameter Tuning for Machine Learn- ing Algorithms Used for Arabic Sentiment Analysis. Informatics 2021 Nov; 8(4):79.doi: 10.3390/informatics8040079

  105. [113]

    Selection of relevant features and examples in machine learning

    Blum AL and Langley P. Selection of relevant features and examples in machine learning. Artificial Intelligence 1997 Dec; 97(1-2):245–71. doi: 10.1016/S0004-3702(97)00063-5

  106. [114]

    Hyperparameter optimization for machine learning models based on Bayesian optimization

    Wu J, Chen XY, Zhang H, Xiong LD, Lei H, and Deng SH. Hyperparameter optimization for machine learning models based on Bayesian optimization. Journal of Electronic Science and Technology 2019 Mar; 17(1):26–40. doi: 10.11989/JEST.1674-862X.80904120. Available from: https : / / ...

  107. [115]

    In- formative variable identifier: Expanding interpretability in feature selection

    Muñoz-Romero S, Gorostiaga A, Soguero-Ruiz C, Mora-Jiménez I, and Rojo-Álvarez JL. In- formative variable identifier: Expanding interpretability in feature selection. Pattern Recog- nition 2020 Feb; 98:107077. doi: 10.1016/J.PATCOG.2019.107077

  108. [116]

    Batch Normalization: Accelerating Deep Network Training by Re- ducing Internal Covariate Shift

    Ioffe S and Szegedy C. Batch Normalization: Accelerating Deep Network Training by Re- ducing Internal Covariate Shift. 2015 Jun. Available from:https://proceedings.mlr. press/v37/ioffe15.html

  109. [117]

    The impact of age on intensive care

    Akinosoglou K, Schinas G, Almyroudi MP, Gogos C, and Dimopoulos G. The impact of age on intensive care. Ageing Research Reviews 2023 Feb; 84:101832.doi: 10.1016/j.arr. 2022.101832

  110. [118]

    An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling

    Bai S, Kolter JZ, and Koltun V. An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling. 2018 Mar. Available from:https://arxiv. org/abs/1803.01271v2

  111. [119]

    Impact of Dataset Size on Classification Performance: An Empirical Evaluation in the Medical Domain

    Althnian A, AlSaeed D, Al-Baity H, Samha A, Dris AB, Alzakari N, Elwafa AA, and Kurdi H. Impact of Dataset Size on Classification Performance: An Empirical Evaluation in the Medical Domain. Applied Sciences 2021, Vol. 11, Page 796 2021 Jan; 11(2):796. doi: 10.3390/APP11020796....

  112. [120]

    Deep Residual Learning for Image Recognition

    He K, Zhang X, Ren S, and Sun J. Deep Residual Learning for Image Recognition. 2015. arXiv: 1512.03385 [cs.CV]. Available from:https://arxiv.org/abs/1512.03385

  113. [121]

    IFAC-PapersOnLine 2020 Jan; 53(2):1243–8

    RehmerAandKrollA.OnthevanishingandexplodinggradientprobleminGatedRecurrent Units. IFAC-PapersOnLine 2020 Jan; 53(2):1243–8. doi: 10.1016/J.IFACOL.2020.12. 1342

  114. [122]

    Glasgow Coma Scale

    Jain S and Iverson LM. Glasgow Coma Scale. StatPearls 2023 Jun. Available from:https: //www.ncbi.nlm.nih.gov/books/NBK513298/%20http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=PMC1032822

  115. [123]

    Explaining Time Series Predictions with Dynamic Masks

    Crabbe J and Schaar M. Explaining Time Series Predictions with Dynamic Masks. Interna- tional Conference on Machine Learning 2021

  116. [124]

    The Richmond Agitation–Sedation Scale

    Sessler CN, Gosnell MS, Grap MJ, Brophy GM, O’Neal PV, Keane KA, Tesoro EP, and El- swick RK. The Richmond Agitation–Sedation Scale. https://doi.org/10.1164/rccm.2107138 2012 Dec;166(10):1338–44. doi: 10.1164/RCCM.2107138.Availablefrom: www.atsjournals. org BIBLIOGRAPHY 89

  117. [125]

    Decision Combination in Multiple Classifier Systems

    Ho K, Hull JJ, and Srihari SN. Decision Combination in Multiple Classifier Systems. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 1994; 16

  118. [126]

    End-to-end pre-trained CNN-based computer-aided classification system design for chest radiographs

    Chandola Y, Virmani J, Bhadauria H, and Kumar P. End-to-end pre-trained CNN-based computer-aided classification system design for chest radiographs. Elsevier, 2021 :117–40. doi: 10.1016/B978-0-323-90184-0.00011-4

  119. [127]

    Review of Classifier Combination Methods

    Tulyakov S, Jaeger S, Govindaraju V, and Doermann D. Review of Classifier Combination Methods. Studies in Computational Intelligence 2008; 90:361–86. doi: 10.1007/978-3- 540-76280-5_14. Available from:https://link.springer.com/chapter/10.1007/978- 3-540-76280-5_14

  120. [128]

    A Comparison of Undersampling, Oversampling, and SMOTE Methods for Dealing with Imbalanced Classification in Educational Data Mining

    Wongvorachan T, He S, and Bulut O. A Comparison of Undersampling, Oversampling, and SMOTE Methods for Dealing with Imbalanced Classification in Educational Data Mining. Information 2023, Vol. 14, Page 54 2023 Jan; 14(1):54. doi: 10 . 3390 / INFO14010054. Available from:https:/...

  121. [129]

    A comprehensive review on ensemble deep learning: Opportuni- ties and challenges

    Mohammed A and Kora R. A comprehensive review on ensemble deep learning: Opportuni- ties and challenges. Journal of King Saud University - Computer and Information Sciences 2023 Feb; 35(2):757–74. doi: 10.1016/J.JKSUCI.2023.01.014

  122. [130]

    Effects of sliding window variation in the performance of acceleration-based human activity recognition using deep learning models

    Jaén-Vargas M, Leiva KMR, Fernandes F, Goncalves SB, Silva MT, Lopes DS, and Olmedo JJS. Effects of sliding window variation in the performance of acceleration-based human activity recognition using deep learning models. PeerJ Computer Science 2022; 8.doi: 10. 7717/PEERJ-CS.10...

  123. [131]

    Training LSTMS with circular-shift epochs for accurate event fore- casting in imbalanced time series

    Chen X and Gupta L. Training LSTMS with circular-shift epochs for accurate event fore- casting in imbalanced time series. Expert Systems with Applications 2024 Mar; 238:121701. doi: 10.1016/j.eswa.2023.121701

  124. [132]

    LSTM Inefficiency in Long-Term Dependencies Regression Problems

    Al-Selwi SM, Hassan MF, Abdulkadir SJ, and Muneer A. LSTM Inefficiency in Long-Term Dependencies Regression Problems. Journal of Advanced Research in Applied Sciences and Engineering Technology 2023 May; 30(3):16–31. doi: 10.37934/ARASET.30.3.1631

  125. [133]

    Learning Sequence Representations

    Bayer JS. Learning Sequence Representations. PLoS Comput Biol. 2015

  126. [134]

    Synthetic Data generation using DCGAN for im- provedtrafficsignrecognition.NeuralComputingandApplications 2021 Dec;34(24):21465–

    Dewi C, Chen RC, Liu YT, and Tai SK. Synthetic Data generation using DCGAN for im- provedtrafficsignrecognition.NeuralComputingandApplications 2021 Dec;34(24):21465–

  127. [135]

    Synthetic Data in AI: Challenges, Applications, and Ethical Implications

    Hao S, Han W, Jiang T, Li Y, Wu H, Zhong C, Zhou Z, and Tang H. Synthetic Data in AI: Challenges, Applications, and Ethical Implications. 2024 Jan

  128. [136]

    Dissecting Person Re-identification from the Viewpoint of Viewpoint

    Sun X and Zheng L. Dissecting Person Re-identification from the Viewpoint of Viewpoint. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2018 Dec; 2019-June:608–17.doi: 10.1109/CVPR.2019.00070. Available from: https://arxiv.org/abs...

  129. [137]

    A review of ensemble learning and data augmen- tation models for class imbalanced problems: Combination, implementation and evaluation

    Khan AA, Chaudhari O, and Chandra R. A review of ensemble learning and data augmen- tation models for class imbalanced problems: Combination, implementation and evaluation. Expert Systems with Applications 2024 Jun; 244:122778. doi: 10.1016/j.eswa.2023. 122778

  130. [138]

    Available from:https://dl.acm.org/doi/10

    doi: 10.1007/S00521-021-05982-Z . Available from:https://dl.acm.org/doi/10. 1007/s00521-021-05982-z 90 BIBLIOGRAPHY

  131. [139]

    SynSys: A Synthetic Data Generation System for Healthcare Appli- cations

    Dahmen J and Cook D. SynSys: A Synthetic Data Generation System for Healthcare Appli- cations. Sensors 2019, Vol. 19, Page 1181 2019 Mar; 19(5):1181.doi: 10.3390/S19051181. Available from: https://www.mdpi.com/1424-8220/19/5/1181/htm%20https://www. mdpi.com/1424-8220/19/5/1181

  132. [140]

    Generating Multi-label Discrete Patient Records using Generative Adversarial Networks

    Choi E, Biswal S, Malin B, Duke J, Stewart WF, Org S, and Sun J. Generating Multi-label Discrete Patient Records using Generative Adversarial Networks. 2017 Nov. Available from: https://proceedings.mlr.press/v68/choi17a.html

  133. [141]

    ”Why Should I Trust You?”: Explaining the Predic- tions of Any Classifier

    Ribeiro MT, Singh S, and Guestrin C. ”Why Should I Trust You?”: Explaining the Predic- tions of Any Classifier. NAACL-HLT 2016 - 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Proceed- ings of the De...

  134. [142]

    A Comparative Study of Detecting Anomalies in Time Series Data Using LSTM and TCN Models

    Gopali S, Abri F, Siami-Namini S, and Namin AS. A Comparative Study of Detecting Anomalies in Time Series Data Using LSTM and TCN Models. 2021

  135. [143]

    Language models are weak learners

    Manikandan H, Jiang Y, and Kolter JZ. Language models are weak learners. 2023. arXiv: 2306.14101 [cs.LG]. Available from:https://arxiv.org/abs/2306.14101

  136. [144]

    Explainable AI for clinical and remote health applications: a survey on tabular and time series data

    Martino FD and Delmastro F. Explainable AI for clinical and remote health applications: a survey on tabular and time series data. Artificial Intelligence Review 2023 Jun; 56(6):5261–

  137. [145]

    The eICU Collabo- rative Research Database, a freely available multi-center database for critical care research

    Pollard TJ, Johnson AE, Raffa JD, Celi LA, Mark RG, and Badawi O. The eICU Collabo- rative Research Database, a freely available multi-center database for critical care research. Scientific Data 2018 5:1 2018 Sep; 5(1):1–13. doi: 10.1038/sdata.2018.178 . Available from: https:...

  138. [146]

    Prediction of weaning from me- chanical ventilation using Convolutional Neural Networks

    Jia Y, Kaul C, Lawton T, Murray-Smith R, and Habli I. Prediction of weaning from me- chanical ventilation using Convolutional Neural Networks. Artificial Intelligence in Medicine 2021 Jul; 117. Epub 2021 May 5:102087.doi: 10.1016/j.artmed.2021.102087

  139. [147]

    Mechanical Ventilation

    Hickey SM, Sankari A, and Giwa AO. Mechanical Ventilation. StatPearls 2024 Mar. Avail- able from: https://www.ncbi.nlm.nih.gov/books/NBK539742/

  140. [148]

    COVID-19 and disparities affecting ethnic minorities

    Morales DR and Ali SN. COVID-19 and disparities affecting ethnic minorities. The Lancet 2021 May; 397(10286):1684–5. doi: 10.1016/S0140-6736(21)00949-1

  141. [149]

    Shifting machine learning for healthcare from devel- opment to deployment and from models to data

    Zhang A, Xing L, Zou J, and Wu JC. Shifting machine learning for healthcare from devel- opment to deployment and from models to data. Nature Biomedical Engineering 2022 Jul; 6(12):1330–45. doi: 10.1038/s41551-022-00898-y

  142. [150]

    Mikhno A and Ennett CM. Prediction of extubation failure for neonates with respiratory distress syndrome using the MIMIC-II clinical database.2012 Annual International Confer- ence of the IEEE Engineering in Medicine and Biology Society . IEEE, 2012 Aug :5094–7. doi: 10.1109/E...

  143. [151]

    Using Retrospective Analysis of the MIMIC-III Database and Comparison of Machine Learning Algorithms to Predict for Weaning Failure of Mechanical Ventilation

    Lee S. Using Retrospective Analysis of the MIMIC-III Database and Comparison of Machine Learning Algorithms to Predict for Weaning Failure of Mechanical Ventilation. Thesis Su- pervisor: Joong ho, Chang. Master’s Thesis. Seoul, South Korea: Seoul School of Integrated Science a...

  144. [152]

    A Reinforcement Learn- ing Approach to Weaning of Mechanical Ventilation in Intensive Care Units

    Prasad N, Cheng LF, Chivers C, Draugelis M, and Engelhardt BE. A Reinforcement Learn- ing Approach to Weaning of Mechanical Ventilation in Intensive Care Units. Uncertainty in Artificial Intelligence - Proceedings of the 33rd Conference, UAI 2017 2017 Apr. Available from: http...

  145. [153]

    Linear Discriminant Analysis

    Liu ZP. Linear Discriminant Analysis. Encyclopedia of Systems Biology 2013 :1132–3. doi: 10.1007/978-1-4419-9863-7_395 . Available from:https://link.springer.com/ referenceworkentry/10.1007/978-1-4419-9863-7_395 BIBLIOGRAPHY 91

  146. [154]

    Kim J, Kim YK, Kim H, Jung H, Koh S, Kim Y, Yoon D, Yi H, and Kim HJ. Machine Learning Algorithms Predict Successful Weaning From Mechanical Ventilation Before In- tubation: Retrospective Analysis From the Medical Information Mart for Intensive Care IV Database. JMIR Formative...

  147. [157]

    Explainable machine learning approach to predict extu- bation in critically ill ventilated patients: a retrospective study in central Taiwan

    Pai KC, Su SA, Chan MC, et al. Explainable machine learning approach to predict extu- bation in critically ill ventilated patients: a retrospective study in central Taiwan. BMC Anesthesiology 2022; 22. doi: 10 . 1186 / s12871 - 022 - 01888 - y. Available from: https : //doi.or...

  148. [158]

    Development and validation of a machine-learning model for prediction of hypoxemia after extubation in intensive care units

    Xia M, Jin C, Cao S, Pei B, Wang J, Xu T, and Jiang H. Development and validation of a machine-learning model for prediction of hypoxemia after extubation in intensive care units. Annals of Translational Medicine 2022; 10. Available from:https://atm.amegroups.org/ article/view/95636

  149. [159]

    Encounter for palliative care

    NHS Health Research Authority. WAVE study - Weaning Algorithms for mechanical VEn- tilation. Unpublished work. 2018. Available from:https://www.hra.nhs.uk/planning- and-improving-research/application-summaries/research-summaries/wave-study- weaning-algorithms-for-mechanical-ve...

  150. [302]

    Available from:https://pubmed.ncbi.nlm

    doi: 10.1164/RCCM.201208-1523CI . Available from:https://pubmed.ncbi.nlm. nih.gov/23641924/

  151. [315]

    1007 / S10462 - 022 - 10304 - 3 / TABLES / 11

    doi: 10 . 1007 / S10462 - 022 - 10304 - 3 / TABLES / 11. Available from: https : / / link . springer.com/article/10.1007/s10462-022-10304-3

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.