Medical health — new data set 23
70,000 medical journal articles, covering textual medical literature, clinical case reports, medical images with complete transfer authorization
5 million genuine structured medical examination data, gender balance (1:1 for men and women), multi-age stratification, over $1,000
The largest medical NLP training package, 50 million cases of de-sensitized electronic medical records, covering principal complaints, current medical history, past history, admissions, discharges, etc.
Covers medical textbooks, clinical guides, synopsis articles, drug manuals, etc.
5 million contour electrocardiogram images and diagnostic reports, including automatic analysis and medical examination of the diagnosis
Coverage of skin lens images, skin-dermal clinical photographs and standard pathological gold diagnostic reports
5 million chest X-ray DICOM images and structured diagnostic reports, double-waiver standard
5 million head CT sweeps + enhanced DICOM sequence images and diagnostic reports covering brain infarction, brain haemorrhage, intracircle tumor, etc.
High-quality eye colour photographs and structured diagnostic reports based on ICDR International Standards for Diabetes Retinal Disorder
Full slice digital images (20 times + scan rate) and complete pathological diagnostic reports, including immunisation results, TNM phase
5 million periscope videos + key frameshots + diagnostic reports with saliva position/size/spectrum
5 million cases of upper abdominal multiple organ cancer, covering liver/chat/insulin/spleen/adrenal/renal/diarrhea lymph nodes
5 million pathological diagnostic full-process images + text covering complete diagnostic links
Text of 5 million cases of acute disease MDT, training in rare/complex diseases, scarce data on difficult cases
1 million longitudinal tracking data for chronic diseases, 7+ consecutive hospitalizations
Qupath fine area label, HCC/ICC/MD/TLS/MVI category VI
One million liver pathologies are marked, liver cancer and angioma.
1 million cases of mammography marked, breast cancer AI screening and diagnosis
20 million full-cycle hospital text/PDF, covering patient history control/DRG/DIP
5 million + medical entities, 20 million + relationships, ICD-10/11 and SNOMED CT mapping, 7 major knowledge Library
100 million + medical NLPs covering 12 types of tasks such as entity extraction/dialogue understanding/think chain, 66 subdata Set
CT/MRI/X-ray/ ultrasound/ endoscope/pathology complete coverage, with classification/detection/division labels
3 million + structured data covering 26 priority disease queues such as diabetes/minus/tumour, 5 + long-term follow-up
Medical health — stock data sets 11
1 million + in-depth medical examination reports, 800 + structured medical fields
328,000 electronic medical records, diagnostic reports, medical advice, summary of literature NER and relationship extracts Set
52,800 DICOM images and professional medical indications
52,800 multi-series MRI DICOM images of liver cell cancer and marked by a specialist
Universal representation data for medical images covering CT, X-Ray, ultrasound, MRI 4 quartiles Set
1 million + a visual frame for surgery, 24-type anatomy structure separated from the pixel level of the surgical device
One-million-degree obstetrical clinical data sets covering all pregnancy scenarios such as prenatal emergencies, high-risk pregnancies, delivery and post-partum
1 million + voice/face images and 50,000 + professional labels
1 million +DICOM image slices covering 7 large organs, thin layer CT 512 x 512+
Pharmaceuticals NEW 6
Newly launched, covering three major pharmaceutical data assets: structured drug data, traditional Chinese medicine formulas, and medical device consumables。
5 million + Chinese drug data, 15 + core dimensions, fully covered by ATC level 5 to support rational use of drugs and drug alerts
2 million + tablets, formulae and self-agents, covering taste classification, classification of efficacy, composition of agents, land area
1 million + medical equipment registration and consumption data, covering registration number, management category, scope of application, technical parameters
Medical Insurance NEW 4
Newly launched, covering medical insurance settlement lists, DIP disease grouping, DRG grouping codes and medical insurance payment standard data。
10 million + medical insurance settlement lists and payment data, 14 + core dimensions, covering the health insurance catalogue/DIP cluster/DRG code/cost component
International language material NEW 1
Newly launched, covering 5 million + multilingual medical corpus in 10 languages, supporting medical NMT and cross-language knowledge map construction.
5 million + multilingual medical materials, 10 languages, 8 major types, ICD-11/SNOMED-CT terminologies, quality of S/A translation