Google has been continuously investing in artificial intelligence (AI) and further expanding its horizons, the company announced a new medical AI-powered health care research product known as Med-PaLM2. The new software was first reported by the Wall Street Journal and it was announced in April 2023.
The AI tool is being tested in Mayo Clinic Research and here is what the paper states “Large language models (LLMs) have catalyzed significant progress in medical question answering; MedPaLM was the first model to exceed a “passing” score in US Medical Licensing Examination (USMLE) style questions with a score of 67.2% on the MedQA dataset.”
“However, this and other prior work suggested significant room for improvement, especially when models’ answers were compared to clinicians’ answers. Here we present Med-PaLM 2, which bridges these gaps by leveraging a combination of base LLM improvements (PaLM 2), medical domain finetuning, and prompting strategies including a novel ensemble refinement approach,” it added.
In its essence, Med-PaLM 2 is a medical Large Language Model (LLM) that has been improved by Google so that it can answer medical questions more accurately and safely than previous models.
It became the first Large Language Model (LLM) to perform at an "expert" test-taker level performance on the MedQA dataset of US Medical Licensing Examination (USMLE)-style questions, reaching 85%+ accuracy.
It was also the first AI system to reach a passing score on the MedMCQA dataset. This includes the tests of Indian AIIMS and NEET medical exam questions and the AI scored 72.3%.
Source: Google
Med-PaLM 2 is still under development, but it has the potential to revolutionise the way that doctors and other healthcare professionals learn and practise medicine.
It could be used to help doctors diagnose diseases, develop treatment plans, and educate patients. It could also be used to research new drugs and treatments.
What are the Limitations of Med-PaLM 2?
Apart from being still in development mode, here are some limitations of this medical-based AI:
- Med-PaLM 2 is prepared on massive data text and code which means it can definitely provide the information that is available in the system but may not be able to answer questions about rare medical conditions.
- Another limitation of this AI is that it is not a medical expert. This means that the language model (LLM) can only provide general medical advice or treatment recommendations. It cannot provide personalized suggestions, such as which treatment to choose based on the individual's body needs and the severity of the disease.
- Med-PaLM 2 is trained on a dataset. But, the LLM does not have the ability to update its dataset on its own. This means that the dataset can become outdated or biased over time.
The Google Research blog states “Careful consideration will need to be given to the ethical deployment of this technology including rigorous quality assessment in different clinical settings with guardrails to mitigate against risks.”
“For example, the potential harms of using a LLM for diagnosing or treating an illness are much greater than using a LLM for information about a disease or medication,” it added.
In conclusion, this new initiative by Google can revolutionise the medical industry but the platform has its limitations. It is necessary to be aware of these limitations in order to use the system safely and effectively.
READ| What is PaLM 2, Google’s new language model after Bard?
Comments
All Comments (0)
Join the conversation