ONCONOURISH
Smarter Nutrition for Better Outcomes
Overview
The challenge
Patients with cancer do not have access to nutritional treatment.
Users/Testers
30
The solution
ONCONOURISH uses the IBIONS360 app as a key support tool to enhance nutritional adherence in oncology patients. The solution combines features designed to guide and empower users throughout their treatment, facilitating continuous monitoring and the development of long-term healthy eating habits.
This solution integrates technology, education, and personalisation to effectively support cancer patients in improving their well-being through nutrition.
The app’s main features include:
- Personalised notifications that remind and motivate patients to stay on track with their nutritional plan.
- Automated food logging, which differentiates between meals, cooking methods, and the presence or absence of key nutrients, providing a detailed overview of the patient’s diet and allowing immediate feedback on their food choices to improve adherence.
- Barcode scanner, which makes it easy to identify ingredients and nutritional values of commercial products.
- Educational module, offering videos and instructional content to help patients gain a solid understanding of nutrition tailored to their condition, fostering meaningful and sustained learning over time.
- Product recommendation section, based on supermarket items we have studied and selected. These recommendations are based on a detailed analysis of their nutritional composition, ensuring alignment with the patient’s clinical needs and dietary preferences.
- Recipe inspiration and adaptive recommendation system, including a curated collection of recipes that we have personally developed and tested. These recipes are designed to be practical, nutritious, and enjoyable. The system features an automated recipe recommendation engine that dynamically adapts to the patient’s nutritional requirements, dietary restrictions, and taste preferences, offering personalised suggestions to support variety and adherence.
- Integrated chatbot, which uses the patient’s clinical information and personal preferences to answer questions outside of clinical appointments. The chatbot is synchronised with the recipe recommendation system to provide personalised meal suggestions and nutritional advice in real time.
Users/Testers
30
Members
Spain
Spain
Spain
Citizen Engagement
Citizen participation has been a key component in the development and evaluation of ONCONOURISH. Two distinct pilot studies are being carried out, one of which is targeted at women, with an approach tailored to the characteristics and needs of each group.
In the first pilot, focused on middle-aged women with breast cancer, approximately 14 participants took part over a two-month period. This pilot centered on group sessions using a focus group approach, where participants received group consultations led by our nutritionist and shared experiences, needs, and perceptions. This process enabled the co-creation of relevant features and content for the application, as well as the validation of the functionalities we had envisioned.
The second pilot, currently underway, involves 10 men with prostate cancer. This pilot follows a more personalized approach, with individualized monitoring through the use of the application — more closely resembling the real patient flow we have planned. This methodology allows for a direct evaluation of the effectiveness of the proposed digital tools and their impact on the patient’s nutritional adherence.
DataU and FOODITY components
ONCONOURISH collects relevant information about patients’ eating habits, such as the types of foods consumed, cooking methods, and meal frequency. This data enables tailored nutritional monitoring and personalised recommendations without compromising user privacy.
During the pilots, no clinical or sensitive information — such as medical test results — has been collected through the app, since this type of data is typically managed directly by hospitals. The solution is designed to respect personal data sovereignty, ensuring that users remain in control of the information they choose to share.
In addition, ONCONOURISH integrates DataU, a tool from the FOODITY ecosystem that facilitates ethical and transparent data management. This integration makes it easy to apply data sovereignty principles, ensuring that decisions regarding the use of information are always in the hands of the user. A particularly interesting aspect is the user’s increased awareness and understanding of their data when we talk about DataU and why we use it.
Results and achievements
Tangible improvements
Currently, the tangible results of ONCONOURISH are based on the first pilot study conducted with women with breast cancer, as the second trial — involving men with prostate cancer — is still ongoing.
Despite the short duration of the pilot (2 months), the data obtained through scientifically validated questionnaires show a clear positive impact: 80% of participants demonstrated improved adherence to the proposed nutritional changes. This suggests increased awareness of the importance of nutrition during cancer treatment and a genuine willingness to change.
Impact Indicators
Regarding the general Key Performance Indicators (KPIs), we have successfully reached the 10% increase we had set as a goal for KPI 1: Dietary adherence (IBIONS’ method questionnaire). The same applies to KPI 2: Health outcomes (QLQ-C30), which has also reached the targeted 10% for this stage of the project.
In terms of business KPIs, the results have exceeded our expectations. For KPI 3 — user adoption rate — we reached 70% at Stage B, which is higher than initially projected. Similarly, KPI 4 — customer satisfaction — has reached 80%, also surpassing our original expectations.
As for the Technical KPIs, KPI 6: Platform uptime has performed as expected, reaching 99% availability with no major issues reported. On the other hand, KPI 5: Accuracy of nutritional analysis has not been tested extensively yet, so we are not in a position to evaluate its reliability. We are currently in the process of testing and validating its accuracy.
Outputs produced
We have developed the following features:
- Automated food intake logging module, capable of identifying meal types, cooking methods, and nutritional content.
- Barcode scanner, allowing users to access the nutritional information of commercial products.
- Personalised notification system, designed to reinforce adherence and individual monitoring.
- Educational video module, focused on delivering practical and tailored knowledge about oncology nutrition.
- Personalised and adaptive nutritional guide with a chatbot interface, enabling patients to make inquiries tailored to their individual profile.
- Intelligent nutritional information system for ingredients and recipes, designed for future integration with advanced biomarkers.
Feedback from users
“I would like to thank you for your help. I really liked your services – they are very useful, everything is easy to understand, and you make it much easier to put things into practice. Also, thank you for allowing me to have this experience through the app.
I hope everything goes great for you, and that you continue to grow and help many more people!”
– Patient testimonial
Social/environmental/economic impact
ONCONOURISH contributes directly to the goals of Food 2030 across several key areas:
Nutrition for sustainable and healthy diets: We improve nutritional adherence in cancer patients, supporting healthy ageing through personalized plans and evidence-based nutritional education.
Circularity and resource efficiency: Through our Green at Heart initiative, we promote the reduction of food waste by recommending seasonal, local, and sustainable products — encouraging choices that benefit both personal health and the environment.
Innovation and empowering communities: Our approach combines technology, education, and citizen engagement. Beyond the app, we conduct educational sessions in hospitals, patient associations, and public events, helping people gain knowledge and autonomy in nutrition and health.
Within the 11 pathways of Food2030, our solution is especially aligned with:
- 7. Nutrition and sustainable healthy diets, by delivering tools tailored to real clinical needs.
- 10. Data and digital transformation, through our integration with DataU and use of adaptive technologies that prioritize data sovereignty and ethical information management.
Materials and links
Alignment with FOODITY
- Data you control: We ensure all user data remains secure and private, stored exclusively in trusted cloud providers like Azure and AWS. IBIONS retains full ownership and control of the data at all times, with strict adherence to GDPR regulations. Your data is never used for training AI models or shared with third parties — your privacy is our priority.
- Patients at the helm: Our tools empower cancer patients to take control of their nutrition with the support of an AI-driven food analyzer, personalised meal plans, and a real-time chatbot for guidance. But our mission goes beyond providing tools — we believe education is the most powerful form of empowerment. We’re not here to keep you subscribed for life; we’re here to help you learn, grow, and become confident enough to manage your health independently.
- Ethical food choices: Our system promotes informed, healthy decisions rooted in the latest scientific evidence, offering educational guidance that encourages sustainable and health-conscious eating. We emphasize the value of local (kilometer zero), organic, and seasonal foods — not just for better nutrition, but also to support the environment and local communities. It’s about learning to eat well in a way that’s good for both you and the planet.
- Made for you: We’ve specifically tailored our tools for cancer patients, considering their unique medical needs, personal preferences, and cultural dietary habits, with a special emphasis on Mediterranean diets — the most scientifically validated dietary pattern for long-term health.
Lessons and recommendations
Lessons learned
- Adaptability is non-negotiable: We quickly learned that standard AI models needed significant adjustments, especially for tasks like food recognition, pushing us towards more versatile tools like multimodal LLMs.
- Good data is everything: Building personalised nutrition plans hinged on first meticulously cleaning and structuring diverse data sources – a foundation more crucial than we anticipated.
- Patients beed flexibility: Rigid systems fell short with the complex and changing needs of cancer patients; adaptable solutions consistently performed better and felt more meaningful.
- Experts are your best friends: Continuous feedback from the IBIONS medical and nutritional teams was vital for fine-tuning our tools, like the RAG chatbot, ensuring they aligned with real-world clinical practice.
- Ethics from day one: User trust depends on more than just robust tech; it requires clarity, transparency, and a supportive experience. We integrated ethical considerations into every design decision, from feedback framing to interface.
- AI is all about iteration: No component worked perfectly initially. Every solution underwent multiple refinements, underscoring the need for flexible workflows and constant testing.
Recommendations
- Put data sovereignty Ffrst: Always ensure users fully own and control their data from the outset. This means selecting infrastructure like GDPR-compliant cloud providers (e.g., Azure) and making privacy and transparency fundamental to your design.
- Build with adaptable Components: Design solutions using interchangeable, modular parts. This allows for easy adjustments to suit different cultural contexts, dietary needs, or user groups.
- Collaborate closely with experts: Work directly with medical professionals, nutritionists, and users. Their feedback is essential for building tools that meet real-world needs, avoid harmful assumptions, and foster trust.
- Design for everyone, personally: Avoid generic solutions. Instead, focus on personalisation and inclusivity, tailoring recommendations to individual health, cultural preferences, and even local food availability for greater relevance and effectiveness.
- Use AI responsibly: While powerful, LLMs must be handled carefully. Deploy them in secure environments with safeguards, such as private instances and curated knowledge bases, to ensure safe and ethical responses.
- Start simple, grow smart: Begin with a basic, functional prototype that solves a core problem. Then, iterate based on user feedback and gradually expand features once a solid foundation is established.
- Document everything and measure impact: Maintain clear documentation of your architecture and decisions for transparency and replicability. Crucially, track key metrics like user satisfaction and adherence to evaluate your solution's impact and guide its future development.
Main benefit of participating in the FOODITY Programme
The main benefit of participating in the FOODITY Programme has been the opportunity to co-develop an impactful, data-driven solution within a European ecosystem that prioritises ethical innovation, data sovereignty, and citizen empowerment.
Since our beginnings at IBIONS, our focus has always been on putting the patient first, with strong support from the professionals involved in the Foodity program — one of the most valuable aspects of the project. We’ve also learned that responsible data management must be another key pillar, making this experience especially valuable.
Additional comments
We would like to highlight the unique value of cross-sector collaboration fostered by the FOODITY Programme. The partnership between a data science company (Datipic), a clinical nutrition institute (IBIONS), and a cancer awareness organization (TheMoveMen) has proven extremely powerful. It ensured that our technological innovations were not only robust and state-of-the-art, but also deeply aligned with real patient needs and grounded in lived experience.
Finally, we are proud to contribute to the broader vision of ethical, European AI in health and food systems, and are enthusiastic about continuing to scale and improve our solution beyond the project timeline.