AI news illustration: Google Expands Medical AI with MedGemma Imaging and MedASR Speech Tool
Google's Medical AI Revolution: MedGemma and MedASR Unveiled
Google Expands Medical AI with MedGemma Imaging and MedASR Speech Tool
Google is advancing medical AI on two fronts, and making both openly available. The company’s updated MedGemma model now processes CT scans and MRIs with sharper accuracy: a 3% gain on disease classification from CTs, a 14% leap on MRI-based tasks. It handles chest X-ray anatomy and lab-report extraction too, all in a compact 4-billion-parameter package designed for offline use.
Meanwhile, MedASR enters the scene, an open speech-to-text model purpose-built for clinical dictation. The message is clear: Google is betting that smaller, specialized, and openly accessible models can transform how medicine sees and listens.
"We are updating our open MedGemma model with improved medical imaging support," said Google in a blog post. "We also describe MedASR, our new open medical speech-to-text model." According to Google, MedGemma 1.5 improved baseline accuracy by 3% on disease classification tasks using CT scans and by 14% on MRI-based classification compared with the previous version. The company also reported gains in anatomical localisation in chest X-rays and structured data extraction from laboratory reports.
The 4B-parameter model is designed to be compute-efficient and capable of running offline, while a larger 27B-parameter version remains available for text-heavy medical applications. MedGemma models can be deployed on Google Cloud through Vertex AI, the company said. Google also announced MedASR, an open automated speech recognition model fine-tuned for medical dictation.
Google’s move is not just about incremental upgrades. It’s a deliberate push to embed AI deeper into the clinical workflow, where accuracy depends on context, not just data. MedGemma 1.5 shrinks the gap between imaging and interpretation, while MedASR turns spoken language into structured information.
Both models are open. Both are built for offline use. That’s a signal: Google wants these tools to function where connectivity fails and privacy matters most.
The 4B-parameter version keeps costs low; the 27B version keeps depth high. Together, they form a stack that scales from a rural clinic’s tablet to a hospital’s cloud. The question isn’t whether AI can read a CT scan or transcribe a consultation.
It’s whether the system can do so without breaking the flow of care. With these releases, Google is betting that open, efficient models are the path forward, not flashier benchmarks, but tools that actually get used.
Related Reading
Browse all Google news →
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
Related Topic Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI
Related Topic Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks
From Archives Deepgram Raises USD 130M, Hits USD 1.3B Valuation in Voice AI Expansion Push
From Archives Anthropic Reshuffles Leadership with Ami Vora's New Product Role
From Archives Google's Veo Turns Portraits into AI Videos with Multi-Image Generation Tool
AI Tools & Apps
From Archives Google Adds Direct Purchase Options to Gemini and Search AI Tools
AI Tools & Apps
Common Questions Answered
How does Google's MedGemma 1.5 improve medical imaging analysis?
MedGemma 1.5 has demonstrated significant improvements in medical imaging accuracy, with a 3% increase in disease classification for CT scans and a 14% improvement for MRI-based classifications. The model also shows enhanced capabilities in anatomical localization for chest X-rays and structured data extraction from medical images.
What are the key features of Google's new MedASR speech-to-text tool?
MedASR is Google's new open medical speech-to-text model designed to transform clinical workflows by converting medical speech into accurate text transcriptions. The tool aims to simplify complex medical documentation processes and improve efficiency for healthcare professionals.
What types of medical imaging does MedGemma 1.5 support?
Google's MedGemma 1.5 has expanded its imaging support to include multiple medical imaging modalities such as CT scans, MRI, and histopathology. This broader support allows the AI model to assist medical professionals in more comprehensive diagnostic processes across different types of medical imaging.
Further Reading
- Google releases medical-specific AI model 'MedGemma 1.5 4B' and 'MedASR' — Gigazine
- Continuous advancements in AI healthcare! Google releases two major open — Futunn
- Google Releases MedGemma 1.5 and MedASR Open Models for Medical Imaging and Speech Recognition — The Rift
- Google Launches MedASR, an Open Medical Speech-to-Text Model — Slator
