– MR imaging biomarkers profiles in patients with prostate cancer treated with androgen deprivation therapy
– CT-based radiomics models to predict progression in locally advanced head and neck cancer treated with definitive chemoradiation
– A clinico-imaging predictive artificial intelligence model of relapse in colon cancer using baseline CT scans
19 October 2023 – Quibim, a company specialising in the design of predictive panels used to enable healthcare providers to improve outcomes for cancer patients, is this week presenting three abstracts at the European Society for Medical Oncology (ESMO) Congress in Madrid. The research covers prostate, head and neck and colon cancers and has been done in collaboration with Consorcio Hospitalario Provincial de Castellon, INCLIVA Biomedical Research Institute and Hospital Clínic de Valencia.
Taking an innovative approach to develop artificial intelligence (AI) models and applying them to medical imaging, Quibim designs tools used by oncologists to improve outcomes for patients with a wide range of cancers. All research presented at ESMO 2023 has been conducted using the company’s five-step process to developing its AI models which involves data curation, feature selection, and model training followed by data validation using unseen internal and, ideally, external datasets. Finally, Quibim works with its partners to understand and explain how the model reaches its decisions and improve its reliability.
Enhanced risk assessment of prostate cancer tumours
The first of Quibim’s abstracts is a publication on MRI biomarker profiles in patients with prostate cancer treated with androgen deprivation therapy. This collaborative research highlights the critical role of accurate risk assessment and biochemical recurrence (BCR) prediction in prostate cancer management. Through radiomic features and imaging biomarkers extracted from the MRIs for the whole prostate and its different subregions, their research showcased enhanced predictive accuracy over a 10-year span. This milestone reflects the company’s commitment to advancing oncological medical imaging through a more personalised approach to managing prostate cancer.
Using radiomics to better predict head and neck cancer progression
Another abstract presented at the conference relates to the use of radiomic models based on baseline computed tomography (CT) scans to predict progression in locally advanced head and neck cancer. Quibim’s research illustrates their innovative use of CT-based radiomic models combined with clinical data as a powerful tool to provide valuable insights into the prediction of 5-years progression-free survival, showing its commitment to transforming the realm of head and neck cancer patient management through a personalized approach, ensuring better outcomes.
Predictive AI used to spot colon cancer relapse
The last of Quibim’s three abstracts is a cutting-edge, clinical-imaging predictive AI model designed to forecast relapse in individuals diagnosed with localized colon cancer. Using baseline CT scans and real-world clinical data, Quibim’s AI model demonstrated improved predictive accuracy for relapse compared to models based solely on clinical variables. By leveraging this predictive tool, clinicians can combine imaging biomarkers with clinicopathological factors to potentially identify high-risk individuals early on, enabling proactive and tailored intervention strategies. Here, Quibim showcases the potential for AI-powered tools to revolutionise how we approach and manage cancer, ultimately leading to enhanced quality of care.
Angel Alberich-Bayarri, Co-Founder and CEO of Quibim, comments on the importance of these findings and their implications for Quibim going forward: “As a medtech company with the ambition of developing imaging biomarker panels for cancer risk assessment and prediction, it is always important for us to work in symbiosis with the oncologists to ensure the technologies we develop are tailored to their needs. This is why ESMO is always so exciting for us, and this year we’re especially proud of being able to announce research which will contribute to a better stratification and treatment of a wide range of cancers, including prostate, head and neck, and colon. We are now determined to incorporate the technologies we’re presenting at the conference into our products and ensuring they can be used by oncologists internationally as soon as is possible.”
Forging strategic partnerships
During the past year, Quibim has also entered several partnerships with life sciences companies. This includes a joint project with Microsoft to build a repository of more than 10 million anonymised medical images for biobanks worldwide through Microsoft Azure cloud services; a partnership with global biotechnology company ImaginAb Inc to revolutionise immune status diagnosis and immunotherapy prediction; a collaboration with the distinguished Mass General Brigham (MGB) healthcare system (United States) for a comprehensive, multi-case, multi-reader validation study on the QP-Prostate Quibim product and a long-term strategic framework agreement with Novartis to create new AI-powered diagnostic imaging tools.
