Radiation Planning Assistant (RPA)

by UT MD Anderson Cancer Center  · Based in United States → — Cancer Treatment & Cancer Research Hospital
Oncology
Website
Claim this tool
Is this the maker? Claim the company →

Overview

The Radiation Planning Assistant (RPA) is a web-based artificial intelligence (AI)-assisted tool developed by Laurence Court, Ph.D., and his team at the University of Texas MD Anderson Cancer Center. It offers automated contouring and planning solutions for various radiotherapy indications, primarily designed to improve access to high-quality radiotherapy in low- and middle-income countries (LMICs) where there is a shortage of skilled specialists.

The RPA streamlines the multi-step process of radiation planning, which traditionally involves clinicians analyzing medical images to design personalized treatment plans, including targeting radiation, avoiding specific areas, and determining dosage. The platform allows clinicians to upload a CT scan and complete a service request form. The de-identified data is then processed on RPA servers.

For complex cases, the system first performs automatic contouring, which is then reviewed and finalized by the clinician before the system delivers the final treatment plan. For simpler cases, the plan is automatically created without this intermediate review step. The RPA is powered by AI tools for preparation and contouring tasks, with plan optimization and dose calculation automatically performed in conjunction with the Eclipse Treatment Planning System (Varian Medical Systems). It also incorporates quality assurance checks to flag suspicious contours or plans for user investigation.

Once the plan is generated, it is sent back to the clinician for recalculation of dosages and any necessary edits before clinical utilization. The RPA is intended to improve workflows and support clinical teams, not to be used as a primary treatment planning system, with final evaluation and approval remaining with the user institution.

Reviewed by Pouyan Golshani, MD — Interventional Radiologist

Key Features

  • Automated contouring of normal tissues and clinical target volumes
  • Automated radiotherapy treatment planning (simple and advanced, e.g., 4-field box, VMAT)
  • AI-powered dose optimization
  • Integrated quality assurance checks to flag suspicious plans
  • Web-based platform for accessibility
  • De-identification of patient data upon upload
  • Supports both one-step (fully automated) and two-step (user-reviewed contouring) planning workflows
  • Integration with commercial treatment planning systems (e.g., Varian Eclipse) via API
  • Designed for use in resource-limited settings

Use Cases

  • Streamlining radiation therapy planning workflows
  • Increasing access to high-quality radiotherapy in low- and middle-income countries (LMICs)
  • Automated planning for head and neck cancer (VMAT)
  • Automated planning for cervical cancer (4-field box, VMAT)
  • Automated planning for breast cancer (chest wall, tangents)
  • Automated planning for brain metastases (whole brain)

What Physicians Need to Know

Key Capabilities
The Radiation Planning Assistant (RPA) offers a suite of fully automated contouring and radiotherapy planning tools. It can generate complete plans for simpler treatments, such as 4-field box plans for cervical cancer, with minimal user interaction. For complex treatments like VMAT planning for head and neck cancers, the system performs automated contouring that is reviewed and edited by the user before final plan generation. The RPA is powered by advanced artificial intelligence to automate data preparation, contouring, and key planning steps, and performs multiple quality assurance checks, automatically flagging potentially suspicious contours or plans for user review.
Clinical Utility
The RPA aims to deliver consistent, high-quality treatment planning and enhance clinical workflows, particularly in low- and middle-income countries (LMICs) where there is a shortage of skilled radiotherapy specialists. It helps reduce planning time, standardize workflows, and supports clinical teams, allowing them to focus on refining plans and dedicating more time to complex cases, teaching, and research. The system is designed to expand access to safe and effective radiotherapy worldwide.
Integration Options
The RPA is a web-based, AI-enabled platform. It integrates with commercial radiotherapy treatment planning systems, specifically performing plan optimization and dose calculation within the Eclipse Treatment Planning System (Varian Medical Systems). It also interfaces with commercial secondary dose verification software. Final contours and treatment plans are returned to the user's own treatment planning system for review, editing, and dose recalculation prior to clinical use.
Compliance Status
The RPA received FDA 510(k) clearance in 2023 and launched clinically in 2024.
Pricing Model
MD Anderson intends to offer the service for free to clinics in low- and middle-income countries (LMICs) that would otherwise not be able to afford commercial solutions. They are investigating how to make the RPA Program financially self-supporting without making it unaffordable to clinics in LMICs, with the goal of making it available at minimal or zero cost to these clinics.
User Experience
The workflow begins with the user uploading a CT scan and completing a Service Request form. Patient data are securely transferred, de-identified, and processed within the RPA system. For complex cases, users review and edit automatically generated contours within their local treatment planning system before the RPA generates the final plan. For simpler cases, the system can generate a complete plan with minimal user interaction. The final plan is downloaded from the RPA website and imported into the user's local treatment planning system for dose recalculation, review, and any necessary modification.
Support Quality
The RPA is intended to enhance clinical workflows while maintaining full clinician oversight; final plan evaluation and quality assurance remain the responsibility of the treating institution. Dedicated, purpose-built algorithms perform multiple quality assurance checks, automatically flagging potentially suspicious contours or plans for user review. The registration process for collaborators includes independent audits of radiation outputs and the local treatment planning system.
Implementation Complexity
The RPA is a web-based tool, designed to be easy to use and deploy across diverse clinical environments. The registration process is currently open to select collaborators and includes independent audits to ensure clinics can use the RPA in their workflows. Training is provided, including videos on generating and reviewing treatment plans and detecting errors.
Evidence Base
The RPA is powered by advanced artificial intelligence and its algorithms have been developed and evaluated through various publications. Studies have been conducted to develop and validate checklists for use with automatically generated radiotherapy plans to improve error detection during plan review. Research indicates high rates of clinical acceptability for auto-contours and treatment plans for certain cancer types, such as prostate and cervical cancer, with many meeting per-protocol compliance criteria.
Physician Tip

Physicians should remember that while the RPA automates many steps, it is an *assistant* tool. Final plan evaluation, quality assurance, and any necessary modifications remain the clinician's responsibility. Always review the automatically generated contours and plans thoroughly within your local treatment planning system, especially for complex cases, and perform dose recalculations before clinical use. The system is particularly beneficial for streamlining routine cases and can free up valuable time for more complex patient care, teaching, and research.

The RPA is designed to integrate with existing radiotherapy treatment planning systems, specifically Varian Eclipse, for plan optimization and dose calculation. It also works with commercial secondary dose verification software. The web-based nature of the tool facilitates access, but clinics will need to ensure their local systems are compatible for seamless data transfer, review, and final plan implementation.

Details

Category Oncology AI, Radiology & Imaging AI
Pricing Unknown
Free Trial No
Mobile AppYes
Compliance
BAA Available Yes AI-estimated
HIPAA Compliant Yes AI-estimated
FDA Status Unknown
Integrations
EHR Not specified
Specialties Oncology

What the Web Says

The Radiation Planning Assistant (RPA) is a web-based, AI-powered tool developed by MD Anderson Cancer Center to automate radiotherapy contouring and treatment planning, particularly for clinics with limited resources. Physicians generally find the RPA-generated contours and plans largely acceptable, either as-is or with minor edits, indicating its clinical relevance and safety. The system aims to improve global access to high-quality cancer care by streamlining the labor-intensive planning process.

Overall: Positive

Strengths

  • Automates labor-intensive contouring and treatment planning, saving time and resources.
  • Improves access to high-quality radiotherapy, especially in low- and middle-income countries.
  • Produces plans that are largely acceptable to physicians, either as-is or with minor edits.
  • Enhances consistency and quality of cancer treatment.
  • Can be integrated with existing treatment planning systems.
  • Reduces the demand on highly specialized clinical staff.

Limitations

  • Some plans may require minor edits or adjustments by physicians, particularly for complex cases like head and neck cancers.
  • One reviewer scored a lower percentage of RPA-generated whole-brain plans as acceptable compared to others, indicating some variability.
  • Final review and approval by qualified clinical staff (radiation oncologists, physicists) remains essential for safety and effectiveness.
  • The system currently has specific treatment site and modality limitations (e.g., head/neck VMAT, cervical cancer 4-field box).
  • The role of Radiology Assistants (RAs) or Radiology Practitioner Assistants (RPAs) in general radiology, as discussed on Reddit, faces challenges with credentialing, state laws, and insurance reimbursement, which might indirectly reflect on the broader adoption of 'assistant' roles in radiology, though the RPA is a software tool.
  • The system is still being developed and optimized, particularly for specific users, patient populations, and clinical practices in LMICs.

Based on reviews from: MD Anderson Cancer Center, Cancerworld, PubMed, PMC, Engineering News, UCL (University College London), Imaging Technology News, Reddit, Journal of Visualized Experiments, Digital Commons@TMC, Oncology News Central, YouTube, TrustRadius

Last updated: 2026-10-26

Ratings & Reviews

No reviews yet. Be the first to review this tool!

Rate Radiation Planning Assistant (RPA)

Clinical Value
Ease of Use
Integration
Support & Docs
Value for Money

Press & Coverage

MD Anderson Cancer Center
Radiation Planning Assistant: Delivering on the promise of AI
MD Anderson's web-based AI tool, the Radiation Planning Assistant (RPA), launched clinically in 2024 after FDA clearance in 2023, aiming to streamline radiation planning and improve access to treatment globally, particularly in low- and middle-income countries.
2024-09
Radiating Hope
Radiating Hope Teams Up with MD Anderson's RPA Team to Improve Global Cancer Care
Radiating Hope announced a collaboration with MD Anderson's Radiation Planning Assistant (RPA) team to bring AI-powered radiotherapy to low- and middle-income countries, offering free access to the web-based tool for streamlined planning.
2024-09
Engineering News
AI-powered Radiation Planning Assistant system being used at Universitas Academic Hospital
The Radiation Planning Assistant (RPA) from MD Anderson has been clinically integrated at Universitas Academic Hospital in South Africa, marking the first site outside the US to use this AI-powered platform for cancer treatment planning.
2025-08
University of the Free State News Archive
UFS medical physics and oncology teams at the forefront of AI-powered cancer treatment planning
The University of the Free State, in collaboration with MD Anderson Cancer Center, has integrated the Radiation Planning Assistant (RPA) into clinical practice, becoming the first worldwide to use this AI platform for radiotherapy treatment planning.
2025-07
UT System
The Radiation Planning Assistant: AI to improve access to high quality radiotherapy
A presentation from May 2025 highlights the Radiation Planning Assistant (RPA) as an AI tool to improve access to high-quality radiotherapy, particularly in low- and middle-income countries, with its 510(k) clearance in May 2023 and first patient treated in South Africa in May 2024.
2025-05
medRxiv
AI-Powered Radiotherapy for Resource-Limited Settings: Advancing Cervical and Prostate Cancer Treatment Planning with the Radiation Planning Assistant (RPA)
This study details the expansion and clinical validation of the Radiation Planning Assistant (RPA) to include end-to-end, AI-driven workflows for prostate and cervical cancers, aiming to improve efficiency and accessibility in low- and middle-income countries.
2025-10
FDA
Radiation Planning Assistant (RPA) (K222728)
The FDA 510(k) premarket notification for the Radiation Planning Assistant (RPA) from MD Anderson Cancer Center outlines its use for planning radiotherapy treatments for various cancers using CT images, emphasizing its role as an assistant tool for clinicians.
2023-05
ASCO Publications
Addressing the Global Expertise Gap in Radiation Oncology: The Radiation Planning Assistant
This article discusses the Radiation Planning Assistant (RPA) project, conceived in 2015 and funded in 2016, to address the global expertise gap in radiation oncology by using automated contouring and treatment planning algorithms to support oncologists in low- and middle-income countries.
2023-07

Videos

Product demos, reviews, and walkthroughs for Radiation Planning Assistant (RPA).

No videos found. Search YouTube directly

View all on YouTube

Frequently Asked Questions

RPA is typically designed to integrate as a plug-in or standalone module that interfaces with standard DICOM-RT data from your existing treatment planning system. It automates specific steps, such as contouring or dose prediction, providing a proposed plan that can then be reviewed and refined by the physician.
RPA holds relevant regulatory clearances, such as FDA 510(k) or CE Mark, depending on the region, affirming its safety and efficacy as a medical device. It adheres to strict data privacy protocols, often processing anonymized or de-identified data and ensuring secure data transfer in compliance with regulations like HIPAA and GDPR.
RPA's limitations can include potential biases if trained on unrepresentative datasets, difficulty with highly complex or unusual anatomies, or a lack of adaptability to novel treatment techniques. Physicians should always review AI-generated plans critically, especially for cases outside the system's validated scope or those requiring highly individualized clinical judgment.
RPA often differentiates itself through its specific AI algorithms, the breadth of planning tasks it automates, or its integration capabilities with particular vendor ecosystems. While traditional planning relies entirely on manual contouring and iterative dose calculation, RPA aims to significantly reduce planning time and improve consistency by leveraging machine learning for these tasks.
The cost structure for RPA typically involves an annual subscription fee, which may vary based on the number of users, treatment sites, or modules purchased. There might also be initial integration costs and ongoing support fees, with some vendors offering tiered pricing based on feature sets or usage volume.
RPA's accuracy and safety are supported by extensive internal validation studies and often by peer-reviewed publications demonstrating its performance against manual planning or other benchmarks. These studies typically evaluate metrics like contouring accuracy, dose distribution quality, and planning efficiency across diverse patient cohorts.

Related Tools

SOPHiA DDM
SOPHiA GENETICS
Oncology AI
SOPHiA DDM™ is an AI-driven and cloud-powered platform that helps clinicians generate real-time, actionable insights for patients with cancer and rare disorders by analyzing complex genomic, radiomic, and multimodal data.
Paid
CerviCARE AI
NTL Healthcare
OB/GYN AI
CerviCARE AI is an AI-powered cervical cancer screening system that analyzes smartphone images of the cervix to predict findings, aiming for early detection.
Lumonus AI
Lumonus
Clinical Decision Support & Reference
Lumonus AI is an AI-powered software that assists healthcare providers with prescriptions, dosing, and treatment for cancer patients, with a module (Sigma) that suggests treatment protocols based on prior plans for similar patients.
DeepBio DeepDx® Prostate
DeepBio
Oncology AI
DeepDx® Prostate leverages deep learning to standardize prostate cancer diagnostics by detecting abnormalities in digital slides with high precision, providing automated Gleason grading, tumor quantification, and tumor length measurements.
AIRA Matrix AI Solutions
AIRA Matrix
Lab & Diagnostics
AIRA Matrix offers AI solutions to improve efficiency and diagnostic accuracy in pathology lab workflows, primarily for prostate and breast cancer, with services like DPaaS and predictive toxicology.
Surgical Reality Viewer
Surgical Reality
Oncology AI
Surgical Reality offers CE- and FDA-certified AI-powered 3D medical imaging and surgical planning software that transforms CT scans into patient-specific 3D models for thoracic surgery, enhancing precision in complex lung procedures.

See all Oncology AI tools →

Suggest an Edit → | Report an issue | Last Verified: 2026-04-19 | First Added: 2026-04-19
AI Tool Finder
AI-powered search. Results may not be comprehensive.