Results of Semi-Automatic Medical Document Generation Demonstration Experiment, Launch of "Discharge Summary" Semi-Automatic Generation Service (Planned), and Call for Participating Medical Institutions
PRiME-R's semi-automatic "Discharge Summary" generation function, developed using LLM, will be launched as an optional service for NTT Precision Medicine's "movacal.hospital" in Q1 2026, following successful trials.
📋 Article Processing Timeline
- 📰 Published: March 31, 2026 at 19:10
- 🔍 Collected: April 1, 2026 at 13:39 (18h 29m after Published)
- 🤖 AI Analyzed: April 17, 2026 at 01:15 (371h 35m after Collected)
Medical Corporation Homura Ouchi ni Kaerou. Hospital (Location: Itabashi-ku, Tokyo; Hospital Director: Shindai Mizuno; hereinafter "Ouchi ni Kaerou. Hospital"), NTT Precision Medicine Corporation (Headquarters: Chiyoda-ku, Tokyo; President and CEO: Koji Korekawa; hereinafter "NTT Precision Medicine"), and New Medical Real World Data Research Organization Co., Ltd. (Headquarters: Kyoto City, Kyoto Prefecture; President and CEO: Koji Korekawa; hereinafter "PRiME-R") have been conducting a demonstration experiment*1 since June 2025. Based on the results of this experiment, PRiME-R's developed "Discharge Summary" semi-automatic generation function will be launched as an optional service for "movacal.hospital"*2, a cloud-based electronic medical record system for small and medium-sized hospitals provided by NTT Precision Medicine.
*1 Regarding the start of a demonstration experiment for the "Case Report/Medical History Summary Support System" for semi-automatic generation of medical documents (June 13, 2025)
https://hospital.teamblue.jp/423/
https://www.ntt-precisionmedicine.co.jp/news/single/?id=2025061302
https://prime-r.inc/newsrelease/898/
*2 Electronic medical record system for small and medium-sized hospitals "movacal.hospital" https://movacal.net/hospital/
## 1. Results of the Demonstration Experiment
In the demonstration experiment, PRiME-R, as a research and development institution for SIP3*3, implemented a case report/medical history summary support system (hereinafter "this system") that semi-automatically generates documents from medical text using a Large Language Model (LLM) into movacal.hospital. We worked on the semi-automatic generation of discharge summaries using actual clinical information (present illness, past medical history, hospital course, etc.) at "Ouchi ni Kaerou. Hospital" and improved the accuracy of this system by incorporating evaluations and advice from doctors.
By utilizing this system, in addition to reducing the workload of healthcare professionals in document creation, it also reduced the time spent on fact-checking (record searching) when the attending physician was absent. Doctors participating in the demonstration experiment commented, "We were able to create discharge summaries in approximately one-third of the usual time"*4. Healthcare professionals at "Ouchi ni Kaerou. Hospital" continue to use this system in their actual work.
*3 PRiME-R "Regarding the adoption of the Strategic Innovation Creation Program (SIP) Phase 3" (December 27, 2023)
https://prime-r.inc/wp/newsrelease/707/
*4 This is a comment from a doctor who actually used this system. The time required for semi-automatic document generation may vary from a few seconds to several minutes depending on conditions such as the length of hospitalization.
## 2. Launch of "Discharge Summary" Semi-Automatic Generation Function Service
Based on the results of the demonstration experiment, the "Discharge Summary" semi-automatic generation function is scheduled to be launched as an optional service for movacal.hospital from the first quarter of fiscal year 2026. Details regarding applications and fees for medical institutions will be posted on the movacal website at a later date.
Medical institutions interested in this service are requested to contact us using the inquiry information.
## 3. Screen Image of "Discharge Summary" Semi-Automatic Generation
When the AI summary button placed on the movacal.hospital screen is pressed, the AI (LLM) analyzes the information in the electronic medical record and summarizes it. Healthcare professionals can complete the creation of medical documents by making minor adjustments to the summarized text, thereby significantly shortening the document creation time.
(Reference) Screen transition image*5
*5 This is a screen under development on movacal.hospital and may differ from the screen at the time of service launch.
## 4. Call for Medical Institutions to Participate in Demonstration Experiments for Semi-Automatic Generation of "Nursing Summaries" and "Referral Letters"
In the future, by realizing the semi-automatic generation of "nursing summaries" and "referral letters," we aim to create an optimal work environment for various healthcare professionals, not only doctors but also nurses and medical administrative staff.
Since the format and writing style vary among hospitals, extensive demonstrations at many medical institutions are necessary. Therefore, we are recruiting medical institutions that can participate in the demonstration experiments.
The target medical institutions for each demonstration experiment are as follows:
- [Target Medical Institutions for Nursing Summary Semi-Automatic Generation Demonstration Experiment]
- Medical institutions that have already introduced movacal.hospital
- [Target Medical Institutions for Referral Letter Semi-Automatic Generation Demonstration Experiment]
- Medical institutions that have already introduced movacal.net*6
Medical institutions interested are requested to contact us using the inquiry information.
*6 Home medical care electronic medical record "movacal.net" https://movacal.net/
## 5. Future Development
"Ouchi ni Kaerou. Hospital", NTT Precision Medicine, and PRiME-R will continue research and development while incorporating feedback and needs from clinical settings, aiming to achieve automatic generation of various medical documents. By automatically generating highly accurate medical documents, we will contribute to improving the efficiency of document creation tasks, which are a burden for many healthcare professionals, not just doctors, and promote work style reforms and further development of medical care.
*1 Regarding the start of a demonstration experiment for the "Case Report/Medical History Summary Support System" for semi-automatic generation of medical documents (June 13, 2025)
https://hospital.teamblue.jp/423/
https://www.ntt-precisionmedicine.co.jp/news/single/?id=2025061302
https://prime-r.inc/newsrelease/898/
*2 Electronic medical record system for small and medium-sized hospitals "movacal.hospital" https://movacal.net/hospital/
## 1. Results of the Demonstration Experiment
In the demonstration experiment, PRiME-R, as a research and development institution for SIP3*3, implemented a case report/medical history summary support system (hereinafter "this system") that semi-automatically generates documents from medical text using a Large Language Model (LLM) into movacal.hospital. We worked on the semi-automatic generation of discharge summaries using actual clinical information (present illness, past medical history, hospital course, etc.) at "Ouchi ni Kaerou. Hospital" and improved the accuracy of this system by incorporating evaluations and advice from doctors.
By utilizing this system, in addition to reducing the workload of healthcare professionals in document creation, it also reduced the time spent on fact-checking (record searching) when the attending physician was absent. Doctors participating in the demonstration experiment commented, "We were able to create discharge summaries in approximately one-third of the usual time"*4. Healthcare professionals at "Ouchi ni Kaerou. Hospital" continue to use this system in their actual work.
*3 PRiME-R "Regarding the adoption of the Strategic Innovation Creation Program (SIP) Phase 3" (December 27, 2023)
https://prime-r.inc/wp/newsrelease/707/
*4 This is a comment from a doctor who actually used this system. The time required for semi-automatic document generation may vary from a few seconds to several minutes depending on conditions such as the length of hospitalization.
## 2. Launch of "Discharge Summary" Semi-Automatic Generation Function Service
Based on the results of the demonstration experiment, the "Discharge Summary" semi-automatic generation function is scheduled to be launched as an optional service for movacal.hospital from the first quarter of fiscal year 2026. Details regarding applications and fees for medical institutions will be posted on the movacal website at a later date.
Medical institutions interested in this service are requested to contact us using the inquiry information.
## 3. Screen Image of "Discharge Summary" Semi-Automatic Generation
When the AI summary button placed on the movacal.hospital screen is pressed, the AI (LLM) analyzes the information in the electronic medical record and summarizes it. Healthcare professionals can complete the creation of medical documents by making minor adjustments to the summarized text, thereby significantly shortening the document creation time.
(Reference) Screen transition image*5
*5 This is a screen under development on movacal.hospital and may differ from the screen at the time of service launch.
## 4. Call for Medical Institutions to Participate in Demonstration Experiments for Semi-Automatic Generation of "Nursing Summaries" and "Referral Letters"
In the future, by realizing the semi-automatic generation of "nursing summaries" and "referral letters," we aim to create an optimal work environment for various healthcare professionals, not only doctors but also nurses and medical administrative staff.
Since the format and writing style vary among hospitals, extensive demonstrations at many medical institutions are necessary. Therefore, we are recruiting medical institutions that can participate in the demonstration experiments.
The target medical institutions for each demonstration experiment are as follows:
- [Target Medical Institutions for Nursing Summary Semi-Automatic Generation Demonstration Experiment]
- Medical institutions that have already introduced movacal.hospital
- [Target Medical Institutions for Referral Letter Semi-Automatic Generation Demonstration Experiment]
- Medical institutions that have already introduced movacal.net*6
Medical institutions interested are requested to contact us using the inquiry information.
*6 Home medical care electronic medical record "movacal.net" https://movacal.net/
## 5. Future Development
"Ouchi ni Kaerou. Hospital", NTT Precision Medicine, and PRiME-R will continue research and development while incorporating feedback and needs from clinical settings, aiming to achieve automatic generation of various medical documents. By automatically generating highly accurate medical documents, we will contribute to improving the efficiency of document creation tasks, which are a burden for many healthcare professionals, not just doctors, and promote work style reforms and further development of medical care.
FAQ
When will the "Discharge Summary" semi-automatic generation function be available?
It is scheduled to be offered as an optional service for movacal.hospital starting from the first quarter of fiscal year 2026.
How much can document creation time be reduced?
In the pilot experiment, doctors reported being able to create discharge summaries in approximately one-third of the usual time.
Are "Nursing Summaries" and "Referral Letters" also planned for semi-automatic generation?
Yes, we are currently recruiting medical institutions to participate in pilot experiments for these features, aiming for future implementation.