1. [行业](/industries/overview) 2. 医疗健康和生命科学 # 面向医疗与生命科学的 开发者资源 一套面向医疗与生命科学领域的计算平台、软件与服务组合，为从影像与仪器到基因组学与药物发现在内的各类 AI 解决方案提供支持。 细分领域： - 数字生物学 - 数字健康 - 数字设备 ## NVIDIA AI 基础架构 以下各行是该组件的原始结果记录——每条记录一行。列名为源索引字段；完整数据也可通过随后的链接下载。 | title | ai_infra | featured | use_case | x_formats | document_url | technologies | document_date | short_summary | document_title | learning_level | x_content_types | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | FQ2BAM_Meth (BWA-Meth) | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_fq2bam_meth.html | Parabricks | 2026-03-10T04:00:00.000Z | Accurate alignment of bisulfite-treated DNA sequencing reads (BS-Seq) often used in methylation. | FQ2BAM_Meth (BWA-Meth) | Developer | Library & tool, Documentation | | HaplotypeCaller | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_haplotypecaller.html | Parabricks | 2026-03-10T04:00:00.000Z | Tool from Genome Analysis Toolkit (GATK) to detect germline/genetic variants. | HaplotypeCaller | Developer | Library & tool, Documentation | | Minimap2 | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_minimap2.html | Parabricks | 2026-03-10T04:00:00.000Z | Sequence alignment tool to align DNA or RNA sequencing reads to a reference genome. Often used for long-read technology from ONT or PacBio. | Minimap2 | Developer | Library & tool, Documentation | | MSA Search NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Documentation | https://build.nvidia.com/colabfold/msa-search | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for multiple sequence alignment generation. | MSA Search NIM | Developer | Deployment, Nim, Hosted nim/api | | OpenFold3 NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Documentation | https://build.nvidia.com/openfold/openfold3 | BioNeMo | 2026-01-20T00:00:00.000Z | Hosted build.nvidia NIM/API for protein/DNA/RNA/ligand structure prediction. | OpenFold3 NIM | Developer | Deployment, Nim, Hosted nim/api | | Giraffe | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_giraffe.html | Parabricks | 2026-03-10T04:00:00.000Z | Built by UCSC, a short-read mapping tool for pangenome graph alignment. | Giraffe | Developer | Library & tool, Documentation | | CodonFM | AI Models | true | Digital Biology | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/models/nv_codonfm_encodon | BioNeMo | 2026-02-20T00:00:00.000Z | Model weights/code for codon-level RNA modeling; not hosted NIM inference. | CodonFM | Developer | Model, Model weights | | DiffDock NIM | Deployment | false | Virtual Screening | Hosted NIM/API, Documentation | https://build.nvidia.com/mit/diffdock | BioNeMo | 2025-11-15T00:00:00.000Z | Hosted build.nvidia NIM/API for protein-ligand blind docking; use NIM docs. | DiffDock NIM | Developer | Deployment, Nim, Hosted nim/api | | Boltz-2 NIM | Deployment | false | Virtual Screening | Hosted NIM/API, Documentation | https://build.nvidia.com/mit/boltz2 | BioNeMo | 2026-02-01T00:00:00.000Z | Hosted build.nvidia NIM/API for biomolecular structure and affinity prediction. | Boltz-2 NIM | Developer | Deployment, Nim, Hosted nim/api | | maxToki | AI Models | false | Single-Cell Analysis | Model Weights, Code | https://huggingface.co/theodoris-lab/MaxToki | BioNeMo | 2026-06-05T00:00:00.000Z | Model weights/code for single-cell language modeling; not hosted NIM inference. | maxToki | Developer | Model, Understand, Model weights | | warpKAN | Libraries | false | Digital Biology | Code | https://github.com/NVIDIA-Digital-Bio/warpKAN | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for GPU-accelerated Kolmogorov-Arnold Networks with NVIDIA Warp. | warpKAN | Developer | Library & tool, Code | | AlphaFold2-Multimer NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Documentation | https://build.nvidia.com/deepmind/alphafold2-multimer | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for protein complex structure; container separate. | AlphaFold2-Multimer NIM | Developer | Deployment, Nim, Hosted nim/api | | La-Proteina | AI Models | true | Generative Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/collections/laproteina_weights_data/artifacts | BioNeMo | 2026-01-28T00:00:00.000Z | Model weights/code for all-atom protein generation; not hosted NIM inference. | La-Proteina | Developer | Model, Model weights | | MolMIM NIM | Deployment | false | Molecular Design | Hosted NIM/API, Documentation | https://build.nvidia.com/nvidia/molmim-generate | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for molecule generation; docs define schema. | MolMIM NIM | Developer | Deployment, Nim, Hosted nim/api | | Parabricks | Libraries | false | Genomic Analysis | Documentation, Code | https://docs.nvidia.com/clara/parabricks/latest/ | BioNeMo | 2025-08-01T00:00:00.000Z | GPU genomics analysis docs; adjacent to BioNeMo, not a BioNeMo NIM. | Parabricks | Developer | Library & tool, Documentation | | cuik-molmaker | Libraries | false | Molecular Design | Code | https://github.com/NVIDIA-Digital-Bio/cuik-molmaker | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for molecular featurization for Chemprop-style GNN workflows. | cuik-molmaker | Developer | Library & tool, Code | | cuEquivariance | Libraries | false | AI Model Training | Code | https://github.com/NVIDIA/cuEquivariance | BioNeMo | 2025-11-10T00:00:00.000Z | Code route for optimized equivariant kernels in structure and molecular models. | cuEquivariance | Developer | Library & tool, Code | | nvMolKit | Libraries | false | Molecular Design | Code | https://github.com/NVIDIA-Digital-Bio/nvMolKit | BioNeMo | 2025-12-15T00:00:00.000Z | Code route for GPU cheminformatics: fingerprints, clustering, conformers, search. | nvMolKit | Developer | Library & tool, Code | | JEPA-DNA | AI Models | false | Genomic Analysis | Code | https://github.com/NVIDIA-Digital-Bio/JEPA-DNA | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for genomic foundation modeling with joint-embedding predictive objectives. | JEPA-DNA | Developer | Model, Understand, Code | | f-RAG | AI Models | false | Molecular Design | Code | https://github.com/NVlabs/f-RAG | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for fragment retrieval-augmented molecule generation and optimization. | f-RAG | Developer | Model, Design, Code | | BioNeMo Recipes Documentation | Training Tools | true | AI Model Training | | https://docs.nvidia.com/bionemo-recipes | BioNeMo | 2025-09-30T00:00:00.000Z | Model implementations, training recipes, and inference workflows for biological foundation models. | BioNeMo Recipes Documentation | | | | Generative Protein Binder Design Pipeline | Deployment | true | Protein Design | Blueprint, Code | https://build.nvidia.com/nvidia/protein-binder-design-for-drug-discovery | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow for de novo protein binder design; not a standalone NIM/API. | Generative Protein Binder Design Pipeline | Developer | Deployment, Blueprint | | Evo2 Protein Design Blueprint | Deployment | false | Protein Design | Blueprint | https://build.nvidia.com/nvidia/evo2-protein-design | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow around Evo2; use Evo2-40B NIM for hosted inference. | Evo2 Protein Design Blueprint | Developer | Deployment, Blueprint | | OpenFold2 NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Documentation | https://build.nvidia.com/openfold/openfold2 | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for protein structure from sequence, MSA, template. | OpenFold2 NIM | Developer | Deployment, Nim, Hosted nim/api | | Digital Biology Examples | Training Tools | false | AI Model Training | Code | https://github.com/NVIDIA/digital-biology-examples | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for Digital Biology examples that compose BioNeMo workflows. | Digital Biology Examples | Developer | Training, Code | | ProteinMPNN NIM | Deployment | false | Protein Design | Hosted NIM/API, Documentation | https://build.nvidia.com/ipd/proteinmpnn | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for protein sequence design; docs define schema. | ProteinMPNN NIM | Developer | Deployment, Nim, Hosted nim/api | | RNAPro | AI Models | false | Digital Biology | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/resources/rnapro | BioNeMo | 2025-12-01T00:00:00.000Z | Model weights/code for RNA 3D structure prediction with RNA foundation models. | RNAPro | Developer | Model, Understand, Model weights | | Evo2-40B NIM | Deployment | false | Digital Biology | Hosted NIM/API, Documentation | https://build.nvidia.com/arc/evo2-40b | BioNeMo | 2025-12-05T00:00:00.000Z | Hosted build.nvidia Evo2-40B NIM/API; use docs. Framework/Blueprint separate. | Evo2-40B NIM | Developer | Deployment, Nim, Hosted nim/api | | GenMol | Deployment | false | Molecular Design | Hosted NIM/API, Code | https://build.nvidia.com/nvidia/genmol-generate | BioNeMo | 2025-11-20T00:00:00.000Z | Molecule generation model with hosted NIM/API plus code and weights. | GenMol | Developer | Model, Design, Hosted nim/api | | GFMBench-API | Libraries | false | Genomic Analysis | Code | https://github.com/NVIDIA/GFMBench-api | BioNeMo | 2026-04-28T00:00:00.000Z | Code route for benchmarking genomic foundation models with standard tasks and metrics. | GFMBench-API | Developer | Library & tool, Code | | BioNeMo Recipes | Training Tools | false | AI Model Training | | https://github.com/NVIDIA-BioNeMo/bionemo-recipes/tree/main | BioNeMo | 2026-03-09T00:00:00.000Z | BioNeMo Recipes is a collection of reference implementations and step-by-step guides that demonstrate how to efficiently scale and optimize the training of biological foundation models using NVIDIA Transformer Engine and popular open-source frameworks. | BioNeMo Recipes | | | | Single Cell Analysis Blueprint | Deployment | false | Single-Cell Analysis | Blueprint, Code | https://build.nvidia.com/nvidia/single-cell-analysis | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow for GPU single-cell analysis; not a hosted NIM/API. | Single Cell Analysis Blueprint | Developer | Deployment, Blueprint, Code | | AlphaFold Database of Protein Complexes | Datasets | false | Protein Structure Prediction | Dataset | https://alphafold.ebi.ac.uk/ | BioNeMo | 2026-03-01T00:00:00.000Z | Dataset of predicted protein complexes for interaction biology and drug discovery. | AlphaFold Database of Protein Complexes | Developer | Data, Dataset | | Proteina-Complexa | AI Models | true | Protein Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/models/proteina_complexa | BioNeMo | 2026-03-15T00:00:00.000Z | Model weights/code route for protein binder design against protein or small-molecule targets. | Proteina-Complexa | Developer | Model, Model weights | | Genomics Analysis Blueprint | Deployment | false | Genomic Analysis | Blueprint, Code | https://build.nvidia.com/nvidia/genomics-analysis | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow for germline genomics analysis with Parabricks. | Genomics Analysis Blueprint | Developer | Deployment, Blueprint, Code | | Proteina | AI Models | false | Generative Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/collections/proteina_weights_data | BioNeMo | 2025-12-10T00:00:00.000Z | Model weights/code for flow-based protein backbone generation with fold-class conditioning. | Proteina | Developer | Model, Design, Model weights | | ProtComposer | AI Models | false | Protein Structure Prediction | Model Weights, Code | https://github.com/NVlabs/protcomposer/tree/main/model_weights | BioNeMo | 2026-02-05T00:00:00.000Z | Model weights/code route for spatially conditioned protein structure generation. | ProtComposer | Developer | Model, Design, Model weights | | Megalodon | AI Models | false | Generative Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/collections/megalodon | BioNeMo | 2025-10-15T00:00:00.000Z | Model weights/code route for 3D molecule generation with equivariant graph transformers. | Megalodon | Developer | Model, Design, Model weights | | AlphaFold3 NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Model card | https://build.nvidia.com/deepmind/alphafold3 | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for proteins, nucleic acids, ligands, and ions. | AlphaFold3 NIM | Developer | Deployment, Nim, Hosted nim/api | | VariantWorks | Libraries | false | Genomic Analysis | Code | https://github.com/NVIDIA/VariantWorks | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for deep-learning-based genomic variant calling. | VariantWorks | Developer | Library & tool, Code | | nvQSP | Libraries | false | Molecular Design | Code | https://github.com/NVIDIA-Digital-Bio/nvQSP | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for GPU QSP/PBPK ODE solvers for virtual patient simulations. | nvQSP | Developer | Library & tool, Code | | Consistency Distilled Synthetic Protein Database | Datasets | false | AI Model Training | Dataset | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/resources/proteina-atomistica_data?version=release | BioNeMo | 2025-12-20T00:00:00.000Z | Dataset of curated protein sequence-structure pairs for training and benchmarks. | Consistency Distilled Synthetic Protein Database | Developer | Data, Dataset | | ReaSyn | AI Models | false | Generative Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/resources/nv-reasyn-ar-166m-v1.ckpt?version=1.0 | BioNeMo | 2026-01-15T00:00:00.000Z | Model weights/code route for synthesis pathway prediction from building blocks to products. | ReaSyn | Developer | Model, Optimize, Model weights | | Context Parallelism (boltz-cp) | Training Tools | false | AI Model Training | Code | https://github.com/NVIDIA-Digital-Bio/boltz-cp | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for long-sequence parallelism in protein structure prediction models. | Context Parallelism (boltz-cp) | Developer | Training, Code | | DualBind | AI Models | false | Virtual Screening | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/resources/dualbind_v1_toxbench | BioNeMo | 2025-10-20T00:00:00.000Z | Model weights/code for protein-ligand affinity; scoped to HER-alpha. | DualBind | Developer | Model, Optimize, Model weights | | AlphaFold2 NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Documentation | https://build.nvidia.com/deepmind/alphafold2 | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for monomer protein structure; container separate. | AlphaFold2 NIM | Developer | Deployment, Nim, Hosted nim/api | | RFDiffusion NIM | Deployment | false | Protein Design | Hosted NIM/API, Documentation | https://build.nvidia.com/ipd/rfdiffusion | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for protein backbone/binder design; container separate. | RFDiffusion NIM | Developer | Deployment, Nim, Hosted nim/api | | AvgFlow | AI Models | false | Molecular Design | Model Weights, Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/collections/avgflow | BioNeMo | 2026-01-10T00:00:00.000Z | Model weights/code for molecular 3D conformers; not hosted NIM inference. | AvgFlow | Developer | Model, Design, Model weights | | ESMFold NIM | Deployment | false | Protein Structure Prediction | Hosted NIM/API, Model card | https://build.nvidia.com/meta/esmfold | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for fast single-sequence protein structure. | ESMFold NIM | Developer | Deployment, Nim, Hosted nim/api | | ESM2-650M NIM | Deployment | false | Digital Biology | Hosted NIM/API, Model card | https://build.nvidia.com/meta/esm2-650m | BioNeMo | 2026-06-05T00:00:00.000Z | Hosted build.nvidia NIM/API for protein language-model embeddings. | ESM2-650M NIM | Developer | Deployment, Nim, Hosted nim/api | | KERMT | AI Models | false | Binding Energy | Code | https://github.com/NVIDIA-Digital-Bio/KERMT | BioNeMo | 2025-11-01T00:00:00.000Z | Code route for ADMET property prediction; fine-tune before inference. | KERMT | Developer | Model, Optimize, Code | | Pangenome Aware DeepVariant | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_pangenome_aware_deepvariant.html | Parabricks | 2026-03-10T04:00:00.000Z | Pangenome-aware DeepVariant is an enhanced version of Google's DeepVariant that leverages pangenome reference graphs (GBZ files) to improve variant calling accuracy, particularly in complex and highly variable genomic regions. | Pangenome Aware DeepVariant | Developer | Library & tool, Documentation | | Generative Virtual Screening Pipeline | Deployment | true | Virtual Screening | Blueprint, Code | https://build.nvidia.com/nvidia/generative-virtual-screening-for-drug-discovery | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow for molecule generation, docking, and optimization. | Generative Virtual Screening Pipeline | Developer | Deployment, Blueprint | | nvFLARE | Libraries | false | AI Model Training | Code, Documentation | https://github.com/NVIDIA/NVFlare | FLARE | 2026-03-09T04:00:00.000Z | NVFlare (NVIDIA Federated Learning Application Runtime Environment) is an open-source, domain-agnostic Python SDK that provides a secure, extensible framework for collaborative computing, enabling researchers to adapt existing machine learning and data science workflows to a federated paradigm without centralizing sensitive data. | nvFLARE | Developer | Library & tool, Documentation, Code | | Biomedical AI-Q Research Agent Blueprint | Deployment | false | Reasoning | Blueprint, Code | https://build.nvidia.com/nvidia/biomedical-aiq-research-agent | Blueprint | 2026-06-05T00:00:00.000Z | Blueprint workflow for biomedical research agents; not a BioNeMo NIM/API. | Biomedical AI-Q Research Agent Blueprint | Developer | Deployment, Blueprint, Code | | FQ2BAM (BWA-MEM) | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_fq2bam.html | Parabricks | 2026-03-10T04:00:00.000Z | Part of Burrows-Wheeler Aligner (BWA) bioinformatics software package to provide fast, accurate algorithm for mapping DNA sequence reads to a reference genome. | FQ2BAM (BWA-MEM) | Developer | Library & tool, Documentation | | STAR (rna_fq2bam) | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_rna_fq2bam.html | Parabricks | 2026-03-10T04:00:00.000Z | Spliced Transcripts Alignment to a Reference (STAR). Software tool for aligning RNA sequencing (RNA-seq) reads to a reference genome. | STAR (rna_fq2bam) | Developer | Library & tool, Documentation | | DeepSAP | Deployment | false | Genomic Analysis | Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/containers/clara-parabricks-deepsap?version=v0.0.3 | Parabricks | 2026-03-10T04:00:00.000Z | Enhance RNA-seq alignment by combining it with Transformer based Splice Junction Scoring (TSJS). TSJS alone improves accuracy when applied with additional tools including STAR. | DeepSAP | Developer | Library & tool, Code | | fgbio | Deployment | false | Genomic Analysis | Code | https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/containers/clara-parabricks-umi-fgbio?version=1.4.0-1 | Parabricks | 2026-03-10T04:00:00.000Z | Developed by Fulcrum Genomics, utilize molecular tagging (UMIs) to distinguish actual biological variants from machine sequencing errors, ensuring high-fidelity genomic data analysis. | fgbio | Developer | Library & tool, Code | | Segment-CT | AI Models | false | Imaging Segmentation | Model Weights, Code | https://huggingface.co/nvidia/NV-Segment-CT | MONAI | 2026-03-10T04:00:00.000Z | This 3D foundation model provides out-of-the-box automatic and interactive point-click segmentation for 132 anatomical classes in computed tomography (CT) scans. | Segment-CT | Developer | Model, Understand, Model weights | | RetMol | AI Models | false | Molecular Design | Code | https://github.com/NVlabs/RetMol | BioNeMo | 2026-06-05T00:00:00.000Z | Code route for retrieval-guided molecule generation and property-conditioned design. | RetMol | Developer | Model, Design, Code | | parakeet-1.1b-rnnt-multilingual-asr | AI Models | false | Speech | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/parakeet-1_1b-rnnt-multilingual-asr | Riva | 2026-03-10T04:00:00.000Z | The model transcribes 25 languages in upper case and lower case alphabets along with punctuations, spaces, and apostrophes. | parakeet-1.1b-rnnt-multilingual-asr | Developer | Deployment, Nim, Hosted nim/api | | llama-3.1-nemotron-ultra-253b | AI Models | false | Reasoning | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/llama-3_1-nemotron-ultra-253b-v1 | Nemotron | 2026-03-10T04:00:00.000Z | Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) which is a derivative of Meta Llama-3.1-405B-Instruct | llama-3.1-nemotron-ultra-253b | Developer | Deployment, Nim, Hosted nim/api | | DeepVariant | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_deepvariant.html | Parabricks | 2026-03-10T04:00:00.000Z | Google's deep-learning based variant caller for germline/genetic variants (i.e., inherited diseases like sickle cell disease). | DeepVariant | Developer | Library & tool, Documentation | | Mutectcaller (Mutect2) | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_mutectcaller.html | Parabricks | 2026-03-10T04:00:00.000Z | Tool from Genome Analysis Toolkit (GATK) for somatic variants. | Mutectcaller (Mutect2) | Developer | Library & tool, Documentation | | DeepSomatic | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_deepsomatic.html | Parabricks | 2026-03-10T04:00:00.000Z | Google's deep-learning based variant caller for somatic variants (i.e., non-inherited diseases like cancer). | DeepSomatic | Developer | Library & tool, Documentation | | NVIDIA NeMo | Libraries | false | Language With NVIDIA NeMo | Documentation, Code | https://www.nvidia.com/en-us/ai-data-science/products/nemo/ | NeMo | 2026-03-10T04:00:00.000Z | NeMo is a modular, interoperable, enterprise-ready software suite for managing the AI agent lifecycle, and can be used to develop digital health solutions. | NVIDIA NeMo | Developer | Library & tool, Documentation, Code | | STAR-Fusion | Deployment | false | Genomic Analysis | Documentation | https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/man_starfusion.html | Parabricks | 2026-03-10T04:00:00.000Z | Tool designed to detect gene fusion transcripts using RNA-Seq data-particularly from Illumina sequencing platforms. | STAR-Fusion | Developer | Library & tool, Documentation | | LLAMA3-VILA-M3-13B | AI Models | false | Reasoning | Model Weights, Code | https://huggingface.co/MONAI/Llama3-VILA-M3-13B | MONAI | 2026-03-10T04:00:00.000Z | An advanced multimodal model that analyzes MRI scans to provide structured clinical reasoning and detailed explanations for complex neurological or musculoskeletal findings. | LLAMA3-VILA-M3-13B | Developer | Model, Understand, Model weights | | Segment-CTMR | AI Models | false | Imaging Segmentation | Model Weights, Code | https://huggingface.co/nvidia/NV-Segment-CTMR | MONAI | 2026-03-10T04:00:00.000Z | Expanding on its predecessor, this model automates the detailed segmentation of over 345 anatomical and pathological classes across both CT and magnetic resonance imaging (MRI) modalities. | Segment-CTMR | Developer | Model, Understand, Model weights | | Generate-CT | AI Models | false | Synthetic Data Generation | Model Weights, Code | https://huggingface.co/nvidia/NV-Generate-CT | MONAI | 2026-03-10T04:00:00.000Z | This 3D latent diffusion model synthesizes high-resolution, realistic 3D CT volumes alongside paired segmentation masks to help augment limited medical training datasets. | Generate-CT | Developer | Model, Design, Model weights | | llama-3.3-nemotron-super-49b | AI Models | false | Reasoning | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/llama-3_3-nemotron-super-49b-v1_5 | Nemotron | 2026-03-10T04:00:00.000Z | Llama-3.3-Nemotron-Super-49B-v1.5 is a reasoning model that is post trained for reasoning, human chat preferences, and agentic tasks, such as RAG and tool calling. | llama-3.3-nemotron-super-49b | Developer | Deployment, Nim, Hosted nim/api | | Reason-CXR | AI Models | false | Reasoning | Model Weights, Code | https://huggingface.co/nvidia/NV-Reason-CXR-3B | MONAI | 2026-03-10T04:00:00.000Z | a specialized vision-language model designed for medical reasoning and interpretation of chest X-ray images, with detailed explanations. The model combines visual understanding with medical reasoning capabilities, enabling healthcare professionals to access comprehensive analyses and engage in follow-up discussions about radiological findings | Reason-CXR | Developer | Model, Understand, Model weights | | Generate-MR-Brain | AI Models | false | Synthetic Data Generation | Model Weights, Code | https://huggingface.co/nvidia/NV-Generate-MR-Brain | MONAI | 2026-03-12T04:00:00.000Z | This generative model synthesizes high-fidelity, 3D T1-weighted brain MRI volumes from noise, providing a powerful tool for creating diverse synthetic datasets to train other medical imaging algorithms. | Generate-MR-Brain | Developer | Model, Design, Model weights | | nemotron-3-nano-30b-a3b | AI Models | false | Reasoning | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/nemotron-3-nano-30b-a3b | Nemotron | 2026-03-10T04:00:00.000Z | Nemotron-3-Nano-30B-A3B is a large language model (LLM) trained from scratch and designed as a unified model for both reasoning and non-reasoning tasks | nemotron-3-nano-30b-a3b | Developer | Deployment, Nim, Hosted nim/api | | Generate-MR | AI Models | false | Synthetic Data Generation | Model Weights, Code | https://huggingface.co/nvidia/NV-Generate-MR | MONAI | 2026-03-10T04:00:00.000Z | Similar to the CT variant, this generative AI model creates diverse, high-quality synthetic 3D MRI scans across various contrast modalities to supplement research data. | Generate-MR | Developer | Model, Design, Model weights | | NVIDIA Riva | Libraries | false | Speech | Documentation | https://www.nvidia.com/en-us/ai-data-science/products/riva/ | Riva | 2026-03-10T04:00:00.000Z | NVIDIA Riva is a collection of GPU-accelerated microservices for building real-time, customizable speech AI functionality into your digital health applications. | NVIDIA Riva | Developer | Library & tool, Documentation | | parakeet-ctc-1.1b-asr | AI Models | false | Speech | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/parakeet-ctc-1_1b-asr | Riva | 2026-03-10T04:00:00.000Z | Parakeet transcribes audio into text, using spaces and apostrophes where needed | parakeet-ctc-1.1b-asr | Developer | Deployment, Nim, Hosted nim/api | | magpie-tts-multilingual | AI Models | false | Speech | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/magpie-tts-multilingual | Riva | 2026-03-10T04:00:00.000Z | The Magpie TTS Multilingual model converts text into audio (speech). | magpie-tts-multilingual | Developer | Deployment, Nim, Hosted nim/api | | parakeet-ctc-0.6b-asr | AI Models | false | Speech | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/parakeet-ctc-0_6b-asr | Riva | 2026-03-10T04:00:00.000Z | Parakeet transcribes audio into text, using spaces and apostrophes where needed | parakeet-ctc-0.6b-asr | Developer | Deployment, Nim, Hosted nim/api | | magpie-tts-zeroshot | AI Models | false | Speech | Hosted NIM/API, Model card | https://build.nvidia.com/nvidia/magpie-tts-zeroshot | Riva | 2026-03-10T04:00:00.000Z | The Magpie TTS Zeroshot model converts text into audio (speech). | magpie-tts-zeroshot | Developer | Deployment, Nim, Hosted nim/api | | Raw2Insights-US | AI Models | false | Image Reconstruction | Model Weights | https://huggingface.co/nvidia/NV-Raw2Insights-US | MONAI | 2026-03-12T04:00:00.000Z | Physics-informed AI model for real-time ultrasound autofocusing using raw channel data. | Raw2Insights-US | Developer | Model, Understand, Model weights | | Raw2Insights-MRI | AI Models | false | Image Reconstruction | Model Weights, Code | https://huggingface.co/nvidia/NV-Raw2insights-MRI | MONAI | 2026-03-11T04:00:00.000Z | Accelerated MR reconstruction model supporting a wide range of MRI protocols. | Raw2Insights-MRI | Developer | Model, Understand, Model weights | | GR00T-H | AI Models | false | AI Model Training | Code, Model Weights | https://github.com/NVIDIA-Medtech/GR00T-H | Isaac for Healthcare | 2026-06-01T05:28:10.000Z | GR00T-H is a vision-language-action foundation model with specialized tooling to unify and train across diverse surgical robotics datasets. | GR00T-H | Developer | Model, Robotics, Code | | Cosmos-H-Surgical | AI Models | false | Synthetic Data Generation | Code, Model Weights | https://github.com/NVIDIA-Medtech/Cosmos-H-Surgical | Isaac for Healthcare | 2026-06-10T18:20:08.000Z | Cosmos-H-Surgical delivers high-quality video prediction and transfer for surgical scenes, including future-state simulation and control-conditioned generation across modalities. | Cosmos-H-Surgical | Developer | Model, Design, Code | | Open-H | Datasets | false | AI Model Training | Code, Dataset | https://github.com/open-h/open-h-embodiment | Isaac for Healthcare | 2026-06-04T17:44:57.000Z | Open-H-Embodiment is a community-driven dataset initiative building the open, shared foundation needed to train and evaluate a generalist Vision-Language-Action (VLA) model for healthcare robotics | Open-H | Developer | Data, Dataset | | Rheo | Deployment | false | Synthetic Data Generation | Code, Blueprint | https://github.com/isaac-for-healthcare/i4h-workflows/tree/main/workflows/rheo | Isaac for Healthcare | 2026-06-05T10:11:24.000Z | Rheo is a blueprint for smart hospital automation, providing a simulation-first pipeline for healthcare robotics under NVIDIA Isaac for Healthcare. | Rheo | Developer | Deployment, Blueprint, Code | | Holoscan SDK | Deployment | false | Synthetic Data Generation | Documentation | https://nvidia-holoscan.github.io/ | NVIDIA Holoscan | 2026-05-31T10:45:55.000Z | Holoscan is an open-source library for building multi-modal, multi-rate processing pipelines with accelerated I/O and TensorRT-powered inference in Python or C++ | Holoscan SDK | Developer | Library & tool, Documentation | [下载原始结果数据 (JSON)](https://developer.nvidia.cn/search-data/healthcare.json) ## 开始使用医疗与生命科学用例 数字生物学 数字健康 数字设备 ### 生成式蛋白质结合剂设计 智能地对蛋白质结合剂进行预优化和筛选，从而显著加速和改进药物研发过程。 - [阅读技术博客](https://developer.nvidia.cn/blog/accelerate-protein-engineering-with-the-nvidia-bionemo-blueprint-for-generative-protein-binder-design/) - [探索 NVIDIA AI Blueprint](https://build.nvidia.com/nvidia/protein-binder-design-for-drug-discovery?nvid=nv-int-txtad-232263) - [参加自主培训课程](https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-HX-04+V1) ### 基因组分析 轻松运行基本的基因组学工作流程，利用 NVIDIA® Parabricks® 节省时间。 - [探索 NVIDIA AI Blueprint](https://build.nvidia.com/nvidia/genomics-analysis) - [试用可启动组件](https://brev.nvidia.com/launchable/deploy/now?launchableID=env-2xn8JayrpXlBwGXkKASSomq6gXi) - [NVIDIA Parabricks 入门文档](https://docs.nvidia.com/clara/parabricks/latest/index.html) ### 单细胞分析 借助 NVIDIA RAPIDS 单细胞加速数据处理、聚类、维度、归约和回归。 - [阅读技术博客](https://developer.nvidia.cn/blog/driving-toward-billion-cell-analysis-and-biological-breakthroughs-with-rapids-singlecell/) - [探索 NVIDIA AI Blueprint](https://build.nvidia.com/nvidia/single-cell-analysis) - [试用可启动组件](https://brev.nvidia.com/launchable/deploy?launchableID=env-2xn8JayrpXlBwGXkKASSomq6gXi) - [参加自定进度课程](https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-HX-05+V1) ### BioMedical AI-Q 研究智能体 将真实数据连接到推理模型，实现安全、隐私保护的 AI 工作流，通过提高研发、试验和监管流程中的研究效率、见解生成和决策制定，加速临床开发。 - [阅读技术博客](https://developer.nvidia.cn/blog/advancing-literature-review-amp-target-discovery-with-nvidia-biomedical-ai-q-research-agent-blueprint/) - [探索 NVIDIA AI Blueprint](https://build.nvidia.com/nvidia/biomedical-aiq-research-agent) - [试用可启动组件](https://brev.nvidia.com/launchable/deploy/now?launchableID=env-2yHyPE6LQR0EJUgffHFsBRK1KD5) - [参加自定进度课程](https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-HX-03+V1) ### 环境医疗健康智能体 生成式 AI 推理模型使提供商和患者语音智能体能够提供情境感知型响应、自动生成临床文档、个性化患者互动，并转变工作流程以实现可扩展的高效护理。 - [探索开发者示例](https://build.nvidia.com/wandb/traceability-for-agentic-ai) - [试用可启动组件](https://brev.nvidia.com/launchable/deploy?launchableID=env-34f0WpRwoDMj851ced7kGEoovxH) ### 训练医疗健康机器人 对高级机器人系统进行原型设计和测试，使用真实数据和合成数据训练和评估 AI 模型，为手术 Dexterity 开发机器人策略，并将仿真与现实世界的部署无缝衔接。 - [阅读技术博客](https://developer.nvidia.cn/blog/introducing-nvidia-isaac-for-healthcare-an-ai-powered-medical-robotics-development-platform/) - [在 GitHub 上开始使用](https://github.com/isaac-for-healthcare) - [参加自定进度课程](https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-OV-52+V1) ### AI 辅助注释 加速并改进医学影像的标记流程 - [教程点播](https://www.nvidia.cn/on-demand/session/gtc24-dlit61433/) - [开始使用 MONAI 标签文档](https://docs.monai.io/projects/label/en/latest/) - [参加自定进度课程](https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-HX-02+V1) ## 探索医疗健康和生命科学平台 ### BioNeMo ### 适用于生命科学研究的 AI 开发平台 NVIDIA BioNeMo™ 是一个开放式开发者平台，旨在加速 AI 驱动的研发。 [开始使用 BioNeMo](https://docs.nvidia.com/bionemo-framework/latest/) ### Holoscan ### AI 传感器处理平台 NVIDIA Holoscan 是一个多功能边缘 AI 计算平台，用于实时处理流式传输的医疗设备数据。 [Holoscan 入门](https://docs.nvidia.com/holoscan/index.html) ### 面向医疗健康行业的 Isaac ### AI 机器人的仿真、训练和部署 适用于医疗健康领域的 NVIDIA Isaac™ 是一个专为医疗健康领域打造的 AI 机器人开发平台。 [开始使用 Isaac 打造医疗保健行业](https://isaac-for-healthcare.github.io/i4h-docs/workflows/) ### MONAI ### 开源 AI 框架 MONAI 是一个特定领域的开源框架，用于开发、训练和部署医学影像深度学习模型。 [开始使用 MONAI](https://monai.io/) ### NVIDIA Parabricks ### 基因组学软件套件 NVIDIA Parabricks 是一款用于二级分析的可扩展基因组学软件套件，可提供可信开源工具的 GPU 加速版本。 [开始使用 NVIDIA Parabricks](https://docs.nvidia.com/clara/parabricks/latest/index.html) ## 更多资源 ![代表开发者社区的装饰性图像](https://developer.download.nvidia.com/icons/m48-developer-1.svg) ### 加入 NVIDIA 开发者计划 ![表示培训和认证的装饰性图像](https://developer.download.nvidia.com/icons/m48-certification-ribbon-2.svg) ### 获取培训和认证 ![代表开发者时事通讯的装饰性图像](https://developer.download.nvidia.com/images/isaac/m48-ai-startup-256px-blk.png) ### 加速您的初创公司 ## 联系我们 ![](https://developer.download.nvidia.com/icons/m48-email-settings-ffffff.svg) 注册以接收医疗健康新闻 – 从激动人心的产品发布到重要的新闻更新。 [随时了解最新情况](https://www.nvidia.cn/industries/healthcare-life-sciences/healthcare-news-sign-up/) ![](https://developer.download.nvidia.com/icons/m48-question-support-ffffff.svg) 有疑问？我们的医疗健康专家随时为您提供帮助！ [联系我们](https://www.nvidia.cn/industries/healthcare/contact-sales/)