| Unique ID issued by UMIN | UMIN000059009 |
|---|---|
| Receipt number | R000067489 |
| Scientific Title | Development and Evaluation of an AI-Based Online Cancer Consultation System |
| Date of disclosure of the study information | 2025/09/05 |
| Last modified on | 2025/09/05 16:32:06 |
Development and Evaluation of an AI-Based Online Cancer Consultation System
AI-Based Online Cancer Consultation
Development and Evaluation of an AI-Based Online Cancer Consultation System
AI-Based Online Cancer Consultation
| Japan |
Cancer
| Medicine in general |
Malignancy
NO
This study aims to evaluate the usefulness of a hybrid online cancer consultation system by comparing responses generated through three approaches using a dedicated web-based application: (i) generative AI, (ii) healthcare professionals, and (iii) a hybrid model combining AI and healthcare professionals. Both patients and healthcare professionals will assess these responses
Others
.The hybrid approach seeks to leverage the cross-disciplinary knowledge and rapid response capability of AI, while ensuring medical, social, and ethical validity, as well as empathy and psychological support for patients, through the involvement of healthcare professionals. By integrating these complementary strengths, we aim to develop an efficient and high-quality online hybrid cancer consultation system that addresses the unmet needs of face-to-face cancer consultation support in real-world clinical practice.
Overall evaluation score by the consultee
Overall evaluation score by healthcare professionals
Consultee-reported evaluation scores: clarity of advice, appropriateness of advice, timeliness in resolving the consultation, attitude of the advisor, willingness to recommend the advisor to family/friends, willingness to consult the advisor again, and ranking of the three types of advisors
Healthcare professional-reported evaluation scores: clarity of advice, appropriateness of advice, timeliness in resolving the consultation, attitude of the advisor, and frequency of inadequate/invalid responses
Time from consultation submission to completion of advice
Proportion of consultations completed
Observational
| 18 | years-old | <= |
| Not applicable |
Male and Female
Patients diagnosed with cancer and receiving care at our institution, as well as their family members. Eligible participants include those regardless of treatment type or treatment status, including patients under follow-up without active treatment.
Patients with cancer or their family members
Individuals capable of using a personal computer or smartphone to participate in online consultations
Individuals who have provided written informed consent to participate in this study
Age >=18 years
Individuals without cancer and their family members
Individuals unable to use a personal computer or smartphone for online consultation
100
| 1st name | Hiroya |
| Middle name | |
| Last name | Takeuchi |
Hamamatsu University School of Medicine
Cance center
431-3192
Handayama 1-20-1, Hamamatsu, Shizuoka
0534352111
karayama@hama-med.ac.jp
| 1st name | Masato |
| Middle name | |
| Last name | Karayama |
Hamamatsu University School of Medicine
Cancer Center
431-3192
Handayama 1-20-1, Hamamatsu, Shizuoka
0534352111
karayama@hama-med.ac.jp
Hamamatsu University School of Medicine
Japan Society for the Promotion of Science
Other
Hamamatsu University School of Medicine
Handayama 1-20-1, Hamamatsu, Shizuoka
0534352680
rinri@hama-med.ac.jp
NO
| 2025 | Year | 09 | Month | 05 | Day |
Unpublished
Enrolling by invitation
| 2025 | Year | 03 | Month | 10 | Day |
| 2025 | Year | 07 | Month | 28 | Day |
| 2025 | Year | 07 | Month | 28 | Day |
| 2030 | Year | 03 | Month | 31 | Day |
Written informed consent will be obtained from participants.
Consultee participants (study subjects) will submit their questions in text format through a dedicated web application developed for this study.
For each question, an advisor with expertise in the relevant field will provide a response (advisor response). Subsequently, a response will be generated using ChatGPT (AI response), and the advisor will then revise the AI-generated response to produce a hybrid response. The type of advisor will remain blinded to the consultee.
The consultee will evaluate all three types of responses.
In addition, multiple healthcare professionals, independent of the advisors, will evaluate the responses provided by the advisors.
| 2025 | Year | 09 | Month | 05 | Day |
| 2025 | Year | 09 | Month | 05 | Day |
Value
https://center6.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000067489