Roadmap

✅ Publish anamnesis dataset

On January 28th, 2025, Recurv AI launched two significant datasets aimed at enhancing the capabilities of AI in healthcare to the public on Huggingface:

Recurv-Clinical-Dataset: This dataset comprises 12,600 data points.

Recurv-Medical-Dataset: This dataset comprises 67,300 data points.

✅ Complete medical LLM

Publish fine-tuned LLM on January 22th, 2025 based on Deepseek R1 and Llama 3 with the anamnesis dataset on Huggingface and host an interactive demo for healthcare professionals.

⌛ Anamnesis analysis module

Create analytics tools for medical institutions to extract insights from patient data while emphasizing privacy-preserving machine-learning methods.

🔜 Medical imaging AI

Develop deep learning models for radiology, focusing on tasks like tumor detection, organ segmentation, and fracture identification.

🔜 Clinical NLP tools

Deploy an AI-powered summarization tool to process unstructured data from Electronic Health Records (EHR).

🔜 Launch Recurv AI medical toolkit

Develop a lightweight application that integrates Recurv-Medical-Llama for use in clinical settings. The application should offer real-time Q&A, multi-turn dialogues, and context-aware suggestions features.

🔜 Partnerships

Collaborate with hospitals, universities, and research institutions to test and refine tools and secure pilot programs in diverse clinical environments.

🔜 Publications

Share findings on AI's impact on the anamnesis process and patient care outcomes and submit research papers to leading conferences such as NeurIPS, ACL, or MedInfo.

🔜 Improve the anamnesis dataset

Release an updated version with expanded scenarios and multi-lingual support for global applicability.

🔜 Community engagement

Host webinars and workshops to onboard users and developers and organize a healthcare AI summit to showcase advancements and foster collaboration.

🔜 AI-powered virtual assistant for doctors

Deploy a real-time AI assistant for medical practitioners, offering suggestions during anamnesis and diagnostics.

🔜 Impact assessment and feedback loops

Evaluate AI systems regarding diagnostic accuracy, time efficiency, and user satisfaction.