[IBM Japan] Started operation of AI that generates interactive disease explanations for patients in joint research for the “AI drug discovery platform business”
*IBM Japan*
Press release: August 26, 2024
**
Started operation of AI that generates interactive disease
explanations for patients in joint research for the “AI drug discovery platform business”
National Institute of Biomedical Innovation, Health and Nutrition (Location: Ibaraki City, Osaka Prefecture, Chairman: Yusuke Nakamura, hereinafter “National Institute of Biomedical Innovation, Health and Nutrition”)
”), Osaka International Cancer Center, Osaka Prefectural Hospital Organization (Location: Chuo-ku, Osaka City, President: Nariaki Matsuura, hereinafter referred to as “ Osaka International Cancer Center
Announced that the actual operation of “Interactive disease
explanation generation AI” for breast cancer patients started in August in “Patient-reducing, clinically-oriented circulation system using generation AI (AI drug discovery platform business)” I did.
Currently, drug development relies on trial-and-error drug discovery research and clinical trials, which takes an enormous amount of time and cost. In order to increase the success rate of drug discovery, it is essential to collect clinical information over time in real time and analyze large amounts of high-quality data. In order to build such a medical database, the National Institute of Biomedical Innovation, Health and Nutrition will collaborate with the Osaka International Cancer Center to back up electronic medical record data existing at medical institutions onto the cloud, structure it, and consolidate it into one. By doing so, we aim to build a foundation that will contribute to the development of medicine as real-time medical big data. In addition, this data is expected to serve as a powerful backup for disaster countermeasures, etc.
On the other hand, the system for collecting clinical information at medical institutions in this infrastructure construction work will reduce the workload at medical sites and will require informed consent so that universities and research institutes can appropriately utilize the collected data. Patient consent will be required. To this end, it is important to build a system that collects information with appropriate and optimal content, such as providing explanations to patients, obtaining consent, and collecting clinical data through interviews.
In response to these issues, IBM Japan’s role in this project is to utilize generative AI to provide explanations and obtain consent from patients, so that doctors and nurses can obtain the necessary data appropriately and without effort. “Interactive disease explanation generation AI” and “Patient explanation/consent acquisition support AI”, “Interview generation AI” that generates and analyzes web interview results entered before the hospital visit, and “Interview generation AI” that generates information between medical
professionals regarding patient treatment and care. We are developing and verifying the operation of “Nursing voice input generation AI” that automatically inputs the content of nursing conferences to be shared and automatically creates records of telephone interactions between nurses and patients.
As part of progress in this project, in August of this year, we began the actual operation of “interactive disease explanation generation AI” for breast cancer patients. The National Institute for Biomedical Innovation, Health and Nutrition, Osaka International Cancer Center, and IBM Japan will continue to work together to promote clinical research and improve work efficiency for medical professionals. Masu.
* -About the introduction of interactive breast cancer disease explanation generation AI-*
The “Interactive Breast Cancer Disease Explanation Generation AI,” which began operation in August for first-time outpatient patients in breast and endocrine surgery (breast surgery), is an interactive conversation system that combines an AI avatar and a generated AI chatbot. Patients can access a web browser using a QR code before visiting a doctor, watch a video explaining the disease at any time before the treatment, or enter their questions into a chatbot using the keyboard or voice to interact with the generated AI. Asking questions can help you better understand your disease and treatment. Interactive disease explanation generation AI system is based on IBM’s AI
Build an AI foundation with IBM watsonx, a data platform, and IBM watsonx.ai It leverages the latest large-scale language models (LLMs) supported by .
Breast cancer is the most common cancer among Japanese women*, Osaka International Cancer Center Department of Breast and Endocrine Surgery The number of breast cancer surgeries will exceed 600 in 2022. The treatment for breast cancer is complex, considering not only curability but also cosmetic properties, and choosing a treatment method that suits the patient’s lifestyle and wishes, so it takes about an hour to explain the disease and obtain consent. Ta. By introducing the “Interactive Breast Cancer Disease Explanation Generation AI,” we aim to reduce the time required for explanations and obtaining consent by 30%. Furthermore, while the number of breast cancer cases is increasing, the number of breast specialists is decreasing nationwide, and there are many areas where sufficient medical treatment is not available. By utilizing online “AI that generates interactive breast cancer disease explanations,” it is expected that breast cancer treatment will be equalized and regional disparities in medical care will be rectified.
Patients who have used the “Interactive Breast Cancer Disease Explanation Generation AI” have said, “It’s useful to be able to obtain reliable information when the internet is full of uncertain medical information,” and “I don’t understand the generated AI.” “I feel more confident when I answer, ‘I don’t know,'” and “It’s beneficial to be able to explain the disease and obtain consent while waiting, and for family members to be able to resolve their doubts together.” “I used to feel bad about asking questions to the doctor during my consultation, but by asking the AI multiple times beforehand, I was able to ease my anxiety.” Additionally, a medical professional commented, “It is beneficial for doctors to be able to understand the content of the question in advance by linking it to the inquiry number.”
* -Future work schedule by IBM Japan-*
In the future, we will use the “Interactive Breast Cancer Disease Explanation Generation AI” to analyze patient questions in detail and further improve accuracy. In addition, the “Interactive Breast Cancer Disease Explanation Generation AI” and the “Patient
Explanation/Consent Obtaining Support AI” established by the “Interactive Breast Cancer Disease Explanation Generation AI” will be used to comprehensively collect the medical information of many patients visiting the Osaka International Cancer Center. We are preparing to provide patients with explanations and consent for “prospective studies” that will analyze specific cancer types using the latest technology from the National Institute of Biomedical Innovation, Health and Nutrition. In addition, we handle the esophagus, stomach, large intestine, etc.
Gastrointestinal Internal Medicine
” plans to start operating the interactive disease explanation generation AI from January 2025.
In February 2025, we plan to deploy three generative AI systems. – “Question generation AI” that supports patient-centered
examinations by using generated AI to analyze the web interview results entered before the visit and allowing the doctor to understand the patient’s condition before the consultation.
–
“Nursing voice input generation AI” that automatically inputs the content of nursing conferences to support the creation of nursing records, and automatically creates records of telephone interactions between nurses and patients to streamline telephone recording operations.
– “Document creation/summary creation” that supports document creation by selecting items necessary for various medical documents from electronic medical record information
In order to utilize generated AI in actual clinical settings, it is necessary to correctly understand the actual situation in the medical field and provide truly useful services, while understanding the risks of using AI and creating a safer operational system. need to be built. In the future, the three parties will introduce generated AI into medical settings and aim to create a system that allows patients and medical professionals to safely use AI services that are useful.
*: Ministry of Health, Labor and Welfare “National Cancer Registry Incidence Number/Rate Report”
IBM, the IBM logo, ibm.com, and watsonx represent International Business Machines in the United States and other countries.
Corporation trademark or registered trademark. Other product and service names may be trademarks of IBM or other companies. IBM at the moment
For a list of trademarks, see http://www.ibm.com/legal/copytrade.shtml (US).