Technology
Development of an AI Image Processing System for Radiation Therapy
Developing AI technology that processes medical images used in radiation therapy to make them easier for physicians to interpret, as well as automating complex medical judgments and treatment planning based on these images
Strength
①
A highly specialized AI system for the field of radiation therapy using unique information standards
Medical images captured through MRI or CR in radiation therapy utilize a highly specialized information standard known as DICOM-RT. Due to its complexity, handling these images was limited to skilled physicians, and the development of systems by AI researchers in this field had been virtually impossible
②
High predictive accuracy AI
By utilizing clinical data from Tohoku University Hospital, renowned for providing top-tier, high-precision radiation therapy, the AI achieves exceptional predictive accuracy
③
Comprehensive one-stop support for diverse workflows in radiation therapy
By deeply understanding the specialized and multifaceted workflows of radiation therapy, an AI support system has been developed to assist with each process
Use Case
Developing a one-stop support system tailored to the workflows of radiation therapy
1.
Automated formulated radiation therapy plans
• AI chatbot for supporting treatment plan decisions
• Prediction of treatment efficacy and side effects
2.
Automatic contour extraction (Auto-Contouring)
• AI-based image diagnostics automatically identify the location and contours of tumors and organs from CT images through image processing
3.
Automated creation of radiation therapy plans
• Reduces the time required to develop an optimal plan from an average of 6 hours manually to just 20 minutes using an AI system
• Contributes to ensuring consistent medical quality in a process that traditionally relies on expertise and is prone to variations in quality
4.
Automated verification and adjustment of discrepancies between radiation therapy plans and actual patient data
• AI automatically verifies whether the predicted values from the treatment plan align sufficiently with the actual measurements taken from the patient’s body and adjusts the treatment plan as needed
Business Model Examples
This section shows examples of companies using this technology for its businesses. The sections is intened for illustrating how the technology is used in the real world as food for thoughs for the applicants. Applicants doesn't necessarily follow the exact business model when developing their own business ideas.
Airato Inc.
Business Contents
Selling an AI support system to hospitals performing cancer radiation therapy, aiming to reduce the workload associated with radiation therapy and improve treatment outcomes
Social Issues
While Intensity-Modulated Radiation Therapy (IMRT), which provides treatment outcomes comparable to surgery with less burden on patients, has become increasingly widespread, Japan has faced challenges in delivering this treatment to many patients due to:
①
Variations in treatment quality caused by operator experience levels,
②
Overwork among medical staff
Radiation therapy for cancer and other tumors requires an extensive amount of time to
create treatment plans, taking up to six hours per session
Solution
By utilizing AI radiation therapy planning support software, automation of treatment plan creation has been successfully achieved, reducing work time by 94%
Tasks that previously required six hours can now be completed in approximately 20 minutes
Value
Contributing to improved accuracy and safety in cancer treatment while alleviating overwork amongst healthcare professionals
Message from the Lab
What I Want to Know
Department of Radiation Oncology, Tohoku University Hospital,
Assistant Professor
Mr. Noriyuki Kadoya
• Challenges in radiation therapy practices across countries
• Overview of hospitals providing radiation therapy across countries
– Number of hospitals offering radiation therapy
– Decision-making process for introducing medical equipment in hospitals
Desired Candidate Profile
Students interested in Radiation Therapy × AI Image Processing
• Our laboratory offers an environment where you can learn both fields simultaneously.
• We welcome individuals who are eager to study both areas or those with knowledge and skills in either radiation therapy or AI image processing who wish to further specialize
scroll