All Competitors

Every biological foundation model, evaluated and ranked by the bio.rodeo team

Showing 17 of 7 filtered models

  • University of Oxford +5 othersJuly 1, 2025clinical_outcome_predictioncross_modality_generationdiagnosis+8

    A multimodal transformer foundation model for ECG, PPG, and clinical text, pretrained on heterogeneous cardiac data from ~1.7 million individuals.

    BiosignalsLanguage model
    26Openness
  • PhysioWave

    18711
    ETH Zurich +2 othersJune 12, 2025arrhythmia_detectionecgeeg+9

    A multi-scale wavelet-transformer foundation model for physiological signals (ECG, EMG, EEG), using learnable wavelet decomposition and frequency-guided masked pretraining.

    Biosignals
    80Openness
  • QoQ-Med

    5232391
    MITMay 31, 2025ecgfoundation_modelhistology+7

    Open multimodal clinical foundation model that jointly reasons over medical images, ECG time-series, and text reports, trained with domain-aware reinforcement learning.

    ImagingBiosignalsLanguage model
    74Openness
  • PULSE

    6427999
    The Ohio State University +1 otherOctober 21, 2024cardiologyecgecg_interpretation+6

    A multimodal LLM fine-tuned to interpret electrocardiogram images, trained on the >1M-sample ECGInstruct dataset and evaluated on the ECGBench benchmark.

    BiosignalsImaging
    84Openness
  • Eko HealthOctober 11, 2024disease_detectionecgfoundation_model+5

    Masked-autoencoder foundation models pretrained on millions of unlabeled digital-stethoscope PCG and ECG recordings, fine-tuned for cardiovascular disease detection.

    Biosignals
    22Openness
  • ECG-Chat

    8044
    China University of Geosciences +2 othersAugust 16, 2024cardiologycontrastive_learningdisease_classification+6

    A multimodal ECG-language model that aligns 12-lead ECG waveforms with clinical text for conversational cardiac diagnosis and automated report generation.

    BiosignalsLanguage model
    27Openness
  • Rice UniversityMay 26, 2024arrhythmia_detectioncontrastive_learningconvnext+6

    A foundation model for 12-lead ECG that learns signal representations via multimodal contrastive pretraining against LLM-generated cardiological text.

    Biosignals
    63Openness