Google Working on AI Technology for Ultrasound and Cancer Therapy
Google announced partnerships with hospitals at The Check UP health event to develop AI technology for ultrasound, cancer therapy, and medical language models.
AI technology is now not only used to write papers or financial reports, but can save someone’s life in the medical field, with new technology being worked on by Google.
Based on reports from EngadgetGoogle considers AI to be critical in reading ultrasound devices in areas where there are not enough trained specialists.
Although, sensors will be more accessible than before, they will require experts to test and interpret the images.
READ ALSO:
Google’s AI models can help simplify the process by identifying data, such as detecting breast cancer and gestational age.
To make this happen, Google partnered with Kenyan not-for-profit Jacaranda Health to research AI-based ultrasound treatments for government hospitals.
In addition, the company is also working with Chang Gung Memorial Hospital in Taiwan to research how AI might work to detect breast cancer via ultrasound as an alternative to mammograms, which are limited in availability in low-resource areas. However, this alternative is less effective with high breast density.
On the other hand, Google says its medical-focused big language model (LLM) has significantly improved. Recently Med-PaLM 2, Google’s next-generation healthcare LLM managed to score 85% on physician-level medical exam questions, an increase of 18% over the previous version’s score.
“This model not only accurately answers multiple-choice and open-ended questions, but also provides reasons and evaluates its own responses,” say Corrado and Matias.
However, Google warns that this technology is still not ready to be implemented in the real world. Because, evaluation of criteria such as scientific factuality, precision, media consensus, reasoning, bias, and disadvantages is still lacking when answering medical questions.
“We look forward to working with researchers and the global medical community to fill this gap and understand how this technology can help improve healthcare,” explained Corrado and Matias.
In addition, Google is also partnering with the Mayo Clinic to explore the role of AI technology for health in planning radiotherapy for cancer treatment.
This research focuses on reducing the time-consuming steps of the radiotherapy process, especially in the contouring process.
The contouring process is a technique that requires the doctor to draw lines on a CT scan to separate the cancerous area from the healthy area that is closest to what might be damaged by radiation. This process takes up to 7 hours for one patient.
The company said it will soon publish research from the 3-year study while formalizing an agreement with the Mayo Clinic to explore more research-based radiotherapy, AI models, and commercial uses.
READ ALSO:
Google’s recently announced AI health technology is here to help X-ray screening on the chest for tuberculosis.
For this technology the company is working with an AI-based organization to make this technology widely available in the Sub-Shara African region.
These Google partners are committed to donating 100,000 free tests to help detect tuberculosis early and provide early treatment to reduce its spread. [FY/HBS]
