Datasets
Models
-
gilberto-topp
-
-
5 Follower(s)
40 Following
Datasets
MRI image collection from diverse datasets showcasing anatomical structures with varied contrasts and textures, reflecting the complexity of medical imaging.
3D
Neurology
MRI
Brain
Apache License 2.0
4
Dataset of 3064 T1-weighted contrast-enhanced MRI slices from 233 patients with glioma, meningioma, and pituitary tumors, including tumor masks and manual annotations for segmentation and classification.
Diagnosis
Classification
2
A collection of de-identified T1, contrast-enhanced T1, and T2 MRI images showing various brain tumor types (e.g., astrocytoma, glioblastoma). Images are interpreted by radiologists and intended for study purposes.
2
Dataset of 100 intracranial aneurysm patients with morphological, hemodynamic, demographic, and rupture status data to assess rupture risk and aid predictive modeling.
Demographics
Mixed
2
A dataset of CT brain scans (.jpg, .dcm formats) for detecting and segmenting cancer, tumors, and aneurysms. It supports research in neurology, radiology, and oncology.
2
800 brain MRI images labeled as normal or abnormal, aggregated from multiple sources for training CNN models in brain condition classification.
2
Manual brain MR FLAIR image segmentations from 110 LGG patients in TCGA, linking tumor shapes to genomic subtypes and patient outcomes using deep learning.
Genomics
2
This de-identified tabular dataset includes demographic and clinical risk factors for predicting stroke occurrence. It supports classification tasks based on age, blood pressure, glucose, heart disease, BMI, and smoking status.
ECG
Blood Tests
2
Brain MRI dataset categorized into tumorous, non-tumorous, and unlabeled scans to support machine learning model development for brain tumor detection.
2
This project focuses on brain tumor classification (Benign, Malignant, Pituitary) using deep learning (ANN, CNN, VGG16) on 3260 T1-weighted contrast-enhanced MRI images. The VGG16 model achieved 94% validation accuracy and a 91% F1-Score.
2
MRI dataset of 3060 brain images for tumor detection and classification into tumorous and non-tumorous using CNN and Transfer Learning for automation in diagnostics.
2
Brain CT and MRI scans organized for training CycleGAN models in image-to-image translation, aiming to convert CT images into high-detail MRI equivalents.
2
MRI brain tumor dataset of 7023 images across four classes (glioma, meningioma, pituitary, no tumor) using CNN-based multi-task classification and segmentation for tumor detection, grading, and localization.
1
Multimodal dataset of high-resolution CT and MRI brain scans labeled by tumor type and location, curated for AI-driven tumor detection, classification, and segmentation research.
2
Pre-op multimodal MRI scans of GBM from TCGA with expert-corrected segmentation labels and radiomic features, ideal for quantitative studies and biomarker analysis.
2
Pre-op multimodal MRI scans of LGG from TCGA with expert-corrected segmentation labels and radiomic features, supporting quantitative analysis and biomarker studies.
2
This TCGA dataset classifies brain gliomas into Lower-Grade Gliomas and Glioblastoma using 20 molecular gene mutations and 3 clinical features, aiming to improve glioma grading accuracy and reduce molecular test costs.
2
Engineered glioblastoma brain organoids with specific mutations model tumor progression, integrating single-cell RNA-seq, spatial transcriptomics, imaging, and histopathology to study glioblastoma development and progression.
Histopathology
2
Contains 9,832 MRI brain tumor glioma images (256x256x1) for glioma grade classification and medical image quality enhancement.
2
Dataset of 7023 brain MRI scans for classifying glioma, meningioma, pituitary tumors, or healthy brains. Aims to use transfer learning and CNNs for early tumor diagnosis.
1
This dataset contains 1,334 2D ultrasound images for automatic fetal head circumference measurement, with 999 labeled training images and 335 test images, aimed at algorithm development and benchmarking in prenatal imaging.
3
Head and neck cancer MRI dataset with T2-weighted pre- and mid-radiotherapy scans and expert tumor segmentations (GTVp, GTVn) for 150 patients, supporting two segmentation tasks in MR-guided treatment.
4
42-patient multimodal CT and T1-weighted MRI dataset with expert-labeled segmentations of 30 head and neck organs-at-risk for radiotherapy planning and multimodal segmentation challenges.
3
Dataset of CT scans from 298 head-and-neck cancer patients. Includes DICOM to PNG processing (normalization, CLAHE, stretching, upscaling, blur, antialiasing). For H&N cancer research on tumor characteristics and treatment outcomes.
2
Large CT dataset of 3,346 head and neck cancer cases from UHN, with RTSTRUCT annotations for tumor and 19 organs-at-risk, enriched with clinical and demographic metadata for imaging research.
1
Multimodal dataset integrating transcriptomics, proteomics, and radiomics from 59 IDH-mutant glioma patients to enhance recurrence prediction using RNA-seq, PET, and histology-based ML analysis.
2
CT dataset of 35 subjects from MISTIE and CLEAR trials, includes anonymized non-contrast head scans with manual brain masks and metadata, supporting brain region analysis and segmentation.
2
Labelled lumbar spine T1-weighted MRI dataset for detecting spinal stenosis, includes ground truth annotations for four regions of interest, curated for training SegNet models.
2
Pretraining dataset for RSNA 2024 lumbar spine classification; includes annotated T1/T2-weighted MRIs and CT scans with vertebrae and disc coordinates for orientation tasks.
1
Lumbar spine MRI dataset with sagittal and axial T2-weighted scans, radiologist reports, and annotations for diagnosing degenerative disorders, disc herniations, and spinal stenosis.
1
MRI dataset of 48,345 axial and sagittal lumbar spine slices from 515 back pain patients, capturing the last three IVDs and vertebrae for spinal stenosis detection studies.
1
Aligned, multi-sequence 3T brain MRIs (T1, T1-IR, T2-FLAIR) from 30 subjects with annotations for 11 brain tissue classes, including lesions, used for evaluating brain segmentation algorithms.
1
Glioma brain tumor datasets use MRI structural and radionuclide functional imaging, sometimes longitudinally, to track brain changes naturally or after treatment.
1
Cleaned metadata for RSNA 2022 Cervical Spine Fracture Detection, includes vertebrae labels, segmentation info, and model-based vertebra predictions to aid in spine injury classification.
1
Large multi-center lumbar spine MRI dataset with expert segmentations of vertebrae, discs, and spinal canal; includes degenerative grading and supports the SPIDER segmentation challenge.
1
MRI spine dataset with .dcm images and expert-labeled reports, covering multiple views and conditions like osteophytes, spondyloarthrosis, and herniations for detection, segmentation, and diagnosis tasks.
1
Multi-site, multi-vendor spine MRI dataset following BIDS format, acquired using the spine-generic protocol across subjects, suitable for spinal cord research and image processing.
1
MRI dataset from a healthy 38-year-old male using the spine-generic protocol, formatted in BIDS and intended for spinal imaging research and method development.
1
Spine MRI Dataset for fracture detection & segmentation. Contains anonymized .dcm MRI scans of spines with dystrophic changes, labeled by doctors. Includes PDF medical reports.
3
250-case multi-center stroke MRI dataset with expert-annotated lesions for ISLES 2022, including FLAIR, DWI, and ADC sequences in BIDS/NIfTI format.
1
A tabular dataset for predicting stroke risk based on patient demographics and health indicators, including age, hypertension, heart disease, BMI, glucose level, and smoking status.
1
Symptom-based, ML-ready dataset for stroke risk prediction with binary and percentage outcomes, covering cardiovascular, neurological, and sleep-related features, suitable for classification and regression.
1
RNA-seq of human glioblastoma (GBM) cells reveals that TRAF3 loss, common in GBM, dysregulates lipid metabolism by affecting ECH1 ubiquitination. This protects GBM from lipid peroxidation and immune elimination.
1
Tabular dataset with six biomechanical spine features to classify patients as normal, disk hernia, or spondylolisthesis; includes binary and multiclass tasks for orthopedic diagnostics.
1
Models
LiviaNET is a 3D fully convolutional neural network for semantic image segmentation, detailed in a 2017 Neuroimage paper. It's designed for subcortical segmentation in MRI. The code supports training, testing, and using pre-trained weights.
3
2
MedGemma is a collection of Gemma 3 variants that are trained for performance on medical text and image comprehension. Developers can use MedGemma to accelerate building healthcare-based AI applications.
5