.avif)
Risk-Adaptive Therapy for Pediatric Oncology
AI-supported clinical decisions for safer, more personalized pediatric oncology care.
Starting this week, we will be posting weekly updates on this project and its progress. Each update will highlight what has changed in the project sections, the steps we have taken, and what comes next, so you can follow the work in real time. Today we have refreshed several parts of the project to reflect its current status. We are committed to keeping this weekly update cycle, so that our donors and partners always have a clear and confident view of how the project is moving forward.
11/14/2025. After months of volunteer-driven development, we are introducing this Risk-Adaptive Therapy project to the public. The early groundwork is complete, the clinical framework is in place, and the first modules have already been shaped. With this foundation ready, we are now opening the project for formal support and funding to move into the next phase of full-scale development.
Problem
Pediatric oncology protocols are powerful but unforgiving.
Too much therapy harms the child; too little risks relapse.
Doctors must balance dozens of variables:
• genetics of the tumor
• early treatment response
• toxicity patterns
• comorbidities
• timing of each treatment cycle
• risk of complications, including transplant-related events
Even experienced clinicians face uncertainty when deciding whether to escalate, de-escalate, delay, or modify therapy. In many hospitals, decisions rely on memory, intuition, fragmented records, or incomplete data.
The result:
avoidable toxicity for low-risk children,
insufficient therapy for high-risk patients,
and preventable complications.
Childhood cancer survival continues to improve — but toxicity remains one of the major reasons children suffer during treatment.
Solution
We are building an AI-enabled system that supports doctors in selecting the safest and most effective therapy path for each child.
The system analyzes thousands of clinical variables and continuously updates risk projections as new data becomes available. It does not replace medical judgment. It gives doctors a structured, evidence-based lens through which to see risks earlier, compare similar cases, and navigate complex treatment decisions confidently.
The system includes:
• multi-factor risk assessment models(toxicity, relapse probability, treatment delays)
• scenarios that update in real time as laboratory and clinical parameters change
• decision support for protocol timing (when to continue, delay, or adjust)
• a module for predicting high-risk complications
• explainable recommendations (the system shows which factors drive each prediction)
This project aims to bring clarity, safety, and personalization to pediatric oncology — without disrupting existing clinical workflows.
A risk-adaptive approach offers threelife-changing benefits:
1. Fewer severe toxicities
Children with favorable profiles can receive gentler therapy without compromising effectiveness.
2. Earlier detection of dangerous complications
The system highlights patterns that experienced doctors may notice too late in a crowded clinic.
3. More consistent decisions across hospitals
Junior doctors receive guidance rooted in real-world evidence, while senior clinicians gain a structured support tool.
Ultimately, it helps protect the child’s health today — and their future quality of life.
Our clinical decision system connects four elements:
1. Comprehensive Patient Data
Tumor genetics, staging, treatment history, laboratory dynamics, toxicities, comorbidities, transplant-related factors, and more.
2. Predictive Risk Models
These models estimate the probability of:
• severe toxicity
• treatment delays
• relapse risk under different intensity levels
• major complications in transplant cases
3. Real-Time Scenario Updating
Each new lab result or clinical event updates risk projections automatically.
4. Doctor-AI Interaction Layer
Doctors can:
• find similar past cases
• review predicted risk zones
• evaluate readiness for the next protocol step
• receive evidence-based explanations
• generate structured summaries for tumor boards
The system remains fully under physician control — it simply adds clarity and foresight.
The project covers several key areas inpediatric oncology:
• acute lymphoblastic leukemia (ALL)
• acute myeloid leukemia (AML)
• neuroblastoma
• medulloblastoma
• high-risk complications after stem-cell transplantation
Future expansion will include other solidtumors and rare pediatric malignancies.
The system is designed for multi-countryenvironments and can be deployed in academic centers, regional hospitals, and international collaborations.
For Patients
• fewer severe toxicities
• safer treatment decisions
• more personalized therapy
• better protection of long-term quality of life
For Clinicians
• decision support grounded in real-worldevidence
• early warnings about complications
• automated clinical summaries
• structured case comparison tools
• reduced cognitive load in high-pressure settings
For Health Systems
• more consistent treatment quality acrosscenters
• better use of resources
• transparent, data-driven decision processes
• easier participation in multicenter research
The system uses:
• anonymized clinical datasets
• multi-layer machine-learning models
• risk prediction algorithms
• explainability tools (showing which variables matter most)
• secure connections to hospital records
• continuous learning, as clinicians add outcomes over time
No personal identifiers are shared; the system follows strict privacy and safety standards.
The project is under development. We are assembling an international clinical expert board.
We are currently forming partnerships with:
• pediatric oncology centers
• transplant units
• academic researchers
• donors interested in supporting childhood cancer innovation
Hospitals and clinicians who wish to participate can contact our team for collaboration details.
