You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Educational symptom self-check on a medical knowledge graph: pick a body part and your symptoms, and it ranks candidate diseases by how much of your input they explain. Not a diagnosis. 医疗知识图谱驱动的症状自查:选身体部位、勾选症状,按「能解释你多少症状」给候选疾病排序,并筛掉不可能的科室;覆盖 7733 种疾病、4697 个症状。仅供科普参考,不能替代诊断。
The Labinator Medical Symptom Checker helps you understand the possible causes of symptoms in children and adults. It uses the USWDS UI components and programmed in JavaScript.
A Health Care website, providing information on several diseases and medicines. It contains a simple symptom checker and a general quiz section. It uses HTML, CSS, JavaScript, PHP, MySQL and AJAX.
A Vietnamese dataset of over 12 thousands questions about common disease symptoms. Perfect for researchers and developers building Vietnamese healthcare chatbots or disease prediction models.
A One Stop Medical Assistant. Features include: Diary, Symptom Checker, Allergy Finder, Hospital/Pharmacy Finder, and Medication Tracker. Implemented with the use of Firebase and Google ML, and Clarifai.
SymptoCare is an #AI-powered health assistant that provides #symptom-based #risk predictions and mental wellness support using GPT/Gemini. Built with #React, #Flask, s#cikit-learn, and #Firebase, it offers smart insights and doctor booking features.
Deep learning–powered platform for skin disease image classification and symptom-based disease prediction using a conversational LLM chatbot. Combines computer vision (MobileNetV2, EfficientNet) with natural language processing for interactive, user-friendly medical guidance.
AI-powered healthcare consultation SaaS that enables users to receive medical guidance, analyze symptoms, and interact with an intelligent health assistant through a modern web platform.