{"url":"/dataset/mimic-iii","name":"MIMIC-III","full_name":"The Medical Information Mart for Intensive Care III","description_markdown":"The Medical Information Mart for Intensive Care III (**MIMIC-III**) dataset is a large, de-identified and publicly-available collection of medical records. Each record in the dataset includes ICD-9 codes, which identify diagnoses and procedures performed. Each code is partitioned into sub-codes, which often include specific circumstantial details. The dataset consists of 112,000 clinical reports records (average length 709.3 tokens) and 1,159 top-level ICD-9 codes. Each report is assigned to 7.6 codes, on average. Data includes vital signs, medications, laboratory measurements, observations and notes charted by care providers, fluid balance, procedure codes, diagnostic codes, imaging reports, hospital length of stay, survival data, and more. \r\n\r\nThe database supports applications including academic and industrial research, quality improvement initiatives, and higher education coursework.\r\n\r\nSource: [MIT Laboratory for Computational Biology](https://imes.mit.edu/supporting-clinical-research-with-the-mimic-iii-critical-care-database/)","description_withheld":null,"homepage":"https://mimic.physionet.org/","introduced_date":"2016-05-24","introduced_date_note":null,"introduced_by":{"paper":"/paper/mimic-iii-a-freely-accessible-critical-care","title":"MIMIC-III, a freely accessible critical care database","first_author":"Alistair E.W. Johnson","url":null},"license":{"name":"MIT","url":"https://physionet.org/content/mimiciii/view-license/1.4/"},"modalities":[{"name":"Medical","url":"/datasets/modality/medical"},{"name":"Tabular","url":"/datasets/modality/tabular"}],"tasks":[{"name":"Multivariate Time Series Forecasting","url":"/task/multivariate-time-series-forecasting","datasets_with_task":"/datasets/task/multivariate-time-series-forecasting"},{"name":"Blood pressure estimation","url":"/task/blood-pressure-estimation","datasets_with_task":"/datasets/task/blood-pressure-estimation"},{"name":"Multi-Label Text Classification","url":"/task/multi-label-text-classification","datasets_with_task":"/datasets/task/multi-label-text-classification"},{"name":"Medical Code Prediction","url":"/task/medical-code-prediction","datasets_with_task":"/datasets/task/medical-code-prediction"},{"name":"Mortality Prediction","url":"/task/mortality-prediction","datasets_with_task":"/datasets/task/mortality-prediction"},{"name":"Length-of-Stay prediction","url":"/task/length-of-stay-prediction","datasets_with_task":"/datasets/task/length-of-stay-prediction"},{"name":"Multi-Label Classification Of Biomedical Texts","url":"/task/multi-label-classification-of-biomedical","datasets_with_task":"/datasets/task/multi-label-classification-of-biomedical"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["MIMIC-III"],"data_loaders":[{"repo":"https://github.com/moustafa100/Data-Science-Advanced-Analytics","url":"https://github.com/moustafa100/Data-Science-Advanced-Analytics","frameworks":["tf"]}],"num_papers_in_archive":1041,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/medical-code-prediction-on-mimic-iii","task":"Medical Code Prediction","dataset_variant":"MIMIC-III","rows":18,"metrics":["Micro-F1","Macro-F1","Micro-AUC","Macro-AUC","Precision@5","Precision@8","Precision@15","mAP"],"first_row_in_archive_order":{"model":"GKI-ICD","paper":"/paper/a-general-knowledge-injection-framework-for-1","metrics":{"Macro-AUC":"96.2","Macro-F1":"12.3","Micro-AUC":"99.3","Micro-F1":"61.2","Precision@15":"62.4","Precision@8":"77.7","mAP":"66.1"},"code_links":[{"title":"xuzhang0112/GKI-ICD","url":"https://github.com/xuzhang0112/GKI-ICD"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/mortality-prediction-on-mimic-iii","task":"Mortality Prediction","dataset_variant":"MIMIC-III","rows":13,"metrics":["F1 score","Precision","Recall","Accuracy"],"first_row_in_archive_order":{"model":"Random Forest","paper":"/paper/early-hospital-mortality-prediction-using","metrics":{"F1 score":"0.97","Precision":"0.97","Recall":"0.97"},"code_links":[{"title":"RezaSadeghiWSU/Early-Hospital-Mortality-Prediction-using-Vital-Signals","url":"https://github.com/RezaSadeghiWSU/Early-Hospital-Mortality-Prediction-using-Vital-Signals"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/length-of-stay-prediction-on-mimic-iii","task":"Length-of-Stay prediction","dataset_variant":"MIMIC-III","rows":5,"metrics":["Accuracy (LOS>3 Days)","Accuracy (LOS>7 Days)"],"first_row_in_archive_order":{"model":"EHR-Graph Transformer (pre-trained)","paper":"/paper/unsupervised-pre-training-on-patient","metrics":{"Accuracy (LOS>3 Days)":"71.4%"},"code_links":[{"title":"chantalmp/unsupervised_pre-training_of_graph_transformers_on_patient_population_graphs","url":"https://github.com/chantalmp/unsupervised_pre-training_of_graph_transformers_on_patient_population_graphs"},{"title":"chantalmp/unsupervised-pre-training-on-patient-population-graphs-for-patient-level-predictions","url":"https://github.com/chantalmp/unsupervised-pre-training-on-patient-population-graphs-for-patient-level-predictions"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/multivariate-time-series-forecasting-on-mimic","task":"Multivariate Time Series Forecasting","dataset_variant":"MIMIC-III","rows":5,"metrics":["MSE","NegLL"],"first_row_in_archive_order":{"model":"GraFITi","paper":"/paper/forecasting-irregularly-sampled-time-series","metrics":{"MSE":"0.396 ± 0.030"},"code_links":[{"title":"yalavarthivk/GraFITi","url":"https://github.com/yalavarthivk/GraFITi"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/blood-pressure-estimation-on-mimic-iii","task":"Blood pressure estimation","dataset_variant":"MIMIC-III","rows":4,"metrics":["MAE for SBP [mmHg]","MAE for DBP [mmHg]","Mean Squared Error","MAE"],"first_row_in_archive_order":{"model":"Deep RNN","paper":"/paper/long-term-blood-pressure-prediction-with-deep","metrics":{"MAE for DBP [mmHg]":"6.7","MAE for SBP [mmHg]":"8.54"},"code_links":[{"title":"psu1/DeepRNN","url":"https://github.com/psu1/DeepRNN"},{"title":"akrlowicz/ppg-blood-pressure-estimation","url":"https://github.com/akrlowicz/ppg-blood-pressure-estimation"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/multi-label-text-classification-on-mimic-iii","task":"Multi-Label Text Classification","dataset_variant":"MIMIC-III","rows":3,"metrics":["AUC","Macro F1","Macro Precision","Macro Recall","Micro Precision","Micro Recall","Micro-F1","P@5","Precision","Recall"],"first_row_in_archive_order":{"model":"HLAN","paper":"/paper/explainable-automated-coding-of-clinical","metrics":{"AUC":"0.919","Macro F1":"57.1","Macro Precision":"65","Macro Recall":"51","Micro Precision":"72.9","Micro Recall":"57.3","P@5":"62.5"},"code_links":[{"title":"acadTags/Explainable-Automated-Medical-Coding","url":"https://github.com/acadTags/Explainable-Automated-Medical-Coding"},{"title":"dmcguire81/CS598DL4H","url":"https://github.com/dmcguire81/CS598DL4H"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/multi-label-classification-of-biomedical","task":"Multi-Label Classification Of Biomedical Texts","dataset_variant":"MIMIC-III","rows":1,"metrics":["Micro F1"],"first_row_in_archive_order":{"model":"Convolutional Neural Network with per-label Attention","paper":"/paper/predicting-multiple-icd-10-codes-from","metrics":{"Micro F1":"0.537"},"code_links":[{"title":"3778/icd-prediction-mimic","url":"https://github.com/3778/icd-prediction-mimic"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/multi-label-classification-of-biomedical-1","task":"Multi-Label Classification Of Biomedical Texts","dataset_variant":"MIMIC-III","rows":1,"metrics":["1:3 Accuracy"],"first_row_in_archive_order":{"model":"Similarity","paper":"/paper/mimic-iii-a-freely-accessible-critical-care","metrics":{"1:3 Accuracy":"95"},"code_links":[{"title":"MIT-LCP/mimic-code","url":"https://github.com/MIT-LCP/mimic-code"},{"title":"mit-lcp/mimic-iii-paper","url":"https://github.com/mit-lcp/mimic-iii-paper"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/a-general-knowledge-injection-framework-for-1","title":"A General Knowledge Injection Framework for ICD Coding","date":"2025-05-24","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/domain-knowledge-integrated-cnn-xlstm-xatt","title":"Domain Knowledge Integrated CNN-xLSTM-xAtt Network with Multi Stream Feature Fusion for Cuffless Blood Pressure Estimation from Photoplethysmography Signals","date":"2025-05-13","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/an-unsupervised-approach-to-achieve","title":"An Unsupervised Approach to Achieve Supervised-Level Explainability in Healthcare Records","date":"2024-06-13","rows_on_this_dataset":1,"code_links":2,"syntology":null},{"paper":"/paper/functional-latent-dynamics-for-irregularly","title":"Functional Latent Dynamics for Irregularly Sampled Time Series Forecasting","date":"2024-05-06","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/forecasting-irregularly-sampled-time-series","title":"Forecasting Irregularly Sampled Time Series using Graphs","date":"2023-05-22","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/knowledge-injected-prompt-based-fine-tuning","title":"Knowledge Injected Prompt Based Fine-tuning for Multi-label Few-shot ICD Coding","date":"2022-10-07","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":2,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/automatic-icd-coding-exploiting-discourse","title":"Automatic ICD Coding Exploiting Discourse Structure and Reconciled Code Embeddings","date":"2022-10-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/unsupervised-pre-training-on-patient","title":"Unsupervised Pre-Training on Patient Population Graphs for Patient-Level Predictions","date":"2022-03-23","rows_on_this_dataset":2,"code_links":2,"syntology":null},{"paper":"/paper/code-synonyms-do-matter-multiple-synonyms-1","title":"Code Synonyms Do Matter: Multiple Synonyms Matching Network for Automatic ICD Coding","date":"2022-03-03","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/effective-convolutional-attention-network-for","title":"Effective Convolutional Attention Network for Multi-label Clinical Document Classification","date":"2021-11-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/neural-flows-efficient-alternative-to-neural","title":"Neural Flows: Efficient Alternative to Neural ODEs","date":"2021-10-25","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/read-attend-and-code-pushing-the-limits-of","title":"Read, Attend, and Code: Pushing the Limits of Medical Codes Prediction from Clinical Notes by Machines","date":"2021-07-10","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/medal-medical-abbreviation-disambiguation","title":"MeDAL: Medical Abbreviation Disambiguation Dataset for Natural Language Understanding Pretraining","date":"2020-12-27","rows_on_this_dataset":5,"code_links":1,"syntology":null},{"paper":"/paper/explainable-automated-coding-of-clinical","title":"Explainable Automated Coding of Clinical Notes using Hierarchical Label-wise Attention Networks and Label Embedding Initialisation","date":"2020-10-29","rows_on_this_dataset":3,"code_links":2,"syntology":null},{"paper":"/paper/predicting-multiple-icd-10-codes-from","title":"Predicting Multiple ICD-10 Codes from Brazilian-Portuguese Clinical Notes","date":"2020-07-29","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/a-label-attention-model-for-icd-coding-from","title":"A Label Attention Model for ICD Coding from Clinical Text","date":"2020-07-13","rows_on_this_dataset":2,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":5,"samples_ran":2,"samples_unverified":3,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/icd-coding-from-clinical-text-using-multi","title":"ICD Coding from Clinical Text Using Multi-Filter Residual Convolutional Neural Network","date":"2019-11-25","rows_on_this_dataset":1,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":0,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/blood-pressure-estimation-from","title":"Blood Pressure Estimation from Photoplethysmogram Using a Spectro-Temporal Deep Neural Network","date":"2019-08-04","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/mimic-extract-a-data-extraction-preprocessing","title":"MIMIC-Extract: A Data Extraction, Preprocessing, and Representation Pipeline for MIMIC-III","date":"2019-07-19","rows_on_this_dataset":3,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/gru-ode-bayes-continuous-modeling-of","title":"GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series","date":"2019-05-29","rows_on_this_dataset":1,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":2,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/early-hospital-mortality-prediction-using","title":"Early hospital mortality prediction using vital signals","date":"2018-03-18","rows_on_this_dataset":8,"code_links":1,"syntology":null},{"paper":"/paper/explainable-prediction-of-medical-codes-from","title":"Explainable Prediction of Medical Codes from Clinical Text","date":"2018-02-15","rows_on_this_dataset":6,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/an-empirical-evaluation-of-deep-learning-for","title":"An Empirical Evaluation of Deep Learning for ICD-9 Code Assignment using MIMIC-III Clinical Notes","date":"2018-02-07","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/patient-subtyping-via-time-aware-lstm","title":"Patient Subtyping via Time-Aware LSTM Networks","date":"2017-08-13","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/long-term-blood-pressure-prediction-with-deep","title":"Long-term Blood Pressure Prediction with Deep Recurrent Neural Networks","date":"2017-05-12","rows_on_this_dataset":1,"code_links":2,"syntology":null},{"paper":"/paper/mimic-iii-a-freely-accessible-critical-care","title":"MIMIC-III, a freely accessible critical care database","date":"2016-05-24","rows_on_this_dataset":1,"code_links":2,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":6,"samples_harvested":12,"samples_ran":8,"samples_unverified":4,"pointer_only_for_licence":1,"papers_with_no_sample_that_ran":1,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}