Prostate cancer detection and diagnosis: harnessing image data to do it right

Prostate cancer (PCa) is a major health issue, with approximately 1,3 million new cases diagnosed worldwide every year, being the second most frequent malignancy in men.

PCa remains a critical health challenge worldwide, with physicians and researchers striving to overcome obstacles in early detection, accurate diagnosis, and treatment of advanced cases. Prostate cancer images play a pivotal role in this effort, as advanced imaging technologies allow clinicians to visualize and understand the disease more effectively.

Despite advances in recent years, physicians acknowledge that several challenges still need to be overcome to improve outcomes and maximize quality of life for patients, including early accurate diagnosis, the need for innovation in prostate cancer images, appropriate risk stratification and treating advanced disease.

Early prostate cancer detection 

As occurs with other types of cancer, early detection remains crucial in PCa, as it may dramatically increase life expectancy. According to data, the 5-year survival rate drops from nearly 100% for patients diagnosed with early-stage PCa, to 30% for those whose cancer has spread to other organs. On the other hand, it is well-known that current standard diagnostic tools are not exempted from certain limitations, and may lead to overdiagnosis, misclassification and overtreatment. Advanced imaging techniques, including multiparametric MRI (mpMRI) images, are key to improving early detection and reducing diagnostic uncertainties.

In particular, the prostate biopsy, often required to complement prostate-specific antigen (PSA) and/or digital rectal examination results and to confirm PCa diagnosis, is also a painful and stressful procedure for patients. Once PCa is diagnosed, appropriate risk stratification tools are required, that allow to discriminate aggressive versus non-aggressive cancer, with the subsequent impact on prognosis and treatment. Finally, developing new therapeutic options, especially for advanced disease, while working out how best to use existing treatments for optimal effect is a key area to focus efforts.

Quibim’s commitment to prostate cancer patients 

In Quibim, we deeply care about patients with PCa, and as such, we are fully committed to addressing all these challenges, contributing to the advancement of scientific knowledge and providing solutions through the application of cutting-edge Artificial Intelligence (AI) technologies to medical imaging. By leveraging cutting-edge Artificial Intelligence (AI) technologies and analyzing prostate cancer real images, we aim to transform health outcomes.

Our ambition is to transform health by placing non-invasive virtual biopsies as a standard of care, by developing imaging biomarkers for early detection of aggressive cancer and ultimately, by designing imaging-based digital biomarkers and companion diagnostics. Our current efforts are focused on approaching PCa across the entire spectrum of research, from discovery to translation, to improve patient’s health by generating real-world evidence (RWE) and by ensuring that this evidence is used in practice. Prostate cancer images, analyzed through our proprietary tools, provide essential insights for designing companion diagnostics and improving patient management.

Research on prostate cancer detection

Thus, within the discovery area, Quibim is actively involved in different national and international academic and pharmaceutical industry-sponsored projects that include more than 27,000 PCa cases (see table 1) and are pioneering efforts in creating large repositories of prostate cancer images. These projects serve two main objectives:

 1. The creation of big European repositories of health imaging data: In Quibim we are aware that open access to appropriate data for training, testing and evaluation of AI-based models is a key limitation to the field. To overcome this problem, Chaimeleon and ProCancer-I7, two EU-funded multidisciplinary and collaborative projects, are devoted to set-up and populate a PCa MRI repository (Chaimeleon also includes three other types of cancer and different cancer imaging tests other than MRI) facilitating access to large, high-quality sets of anonymized prostate cancer images. Additionally, Chaimeleon aims to develop and test different imaging data harmonization protocols with the ambition of setting a standard while providing resources for future AI experimentation for cancer management.

2. The development of advanced, trustworthy, AI models to address unmet clinical needs: AI models in PCa may help to increase performance in discriminating indolent from aggressive disease, early predicting recurrence and detecting metastases or predicting the effectiveness of therapies. Thus, using data from the repository, ProCAncer-I will develop robust AI models for addressing different PCa clinical scenarios based on novel ensemble learning methodologies. ProCAncer-I’s ultimate mission is to deliver a PCa AI platform where a unique collection of PCa mpMRI images and AI algorithms co-exist to support diagnosis, patient management and treatment.

Quibim’s contribution to the field underscores the importance of accessible and reliable prostate cancer images in advancing diagnosis and treatment strategies.

Prostate cancer detection solutions

In Quibim we work hard to be at the forefront of science by applying innovative and revolutionary AI tools to prostate cancer images. This the nature of ProCanAid, an EU-funded project that was conceived to develop a computational tool to create a 4D digital twin (a virtual model that reflects a physical entity) of the entire prostate of a patient, by applying novel AI-based MRI segmentation algorithms, advance multiscale Finite Elements (FE)-simulations and in silico tissue/cell models for reconstructing prostate anatomy and for detecting PCa.

ProCanAid will be the first AI-based integrative tool (clinical data – imaging biomarkers – advanced multiscale FE simulations) capable of performing accurate, automatic, medically certified and cost-effective quantitative analysis for PCa based on mpMRI, helping to improve diagnosis, prognosis and treatment response. Finally, our commitment to PCa and drug discovery is also reflected in our successful collaborations with the pharma industry, positioning us as reliable AI partners. As a result, Quibim has joined forces with Janssen in a multi-year partnership to develop an innovative non-invasive approach for prediction of biochemical relapse from baseline imaging exams in PCa patients, by combining RWE, AI and novel imaging biomarkers.

QP-Prostate structured report view

With our holistic approach, and as global leaders in whole-body medical imagining analysis, Quibim sets out to transform PCa landscape.

TABLE 1. Quibim national and international academic and pharmaceutical industry-sponsored projects
CT, computed tomography; mpMRI, multiparametric MRI; MRI, magnetic resonance imaging

Translating science into real-life solutions

Translating science into real-life solutions is in Quibim’s DNA. Thus, our medical imaging software, QP-Prostate®, constitutes an FDA 510(k)-cleared solution for prostate MRI analysis based on AI. By utilizing PCa images from MRI scans, QP-Prostate® supports  radiologists, urologists and oncologists at each step of the workflow, from visualization to quantification, with the aim of increasing diagnostic accuracy and, potentially, early PCa detection. At present. QP-Prostate® is the only tool on the market providing automated regional organ segmentation, a process that may reduce interpretation time and helps define diagnosis per region.

The QP-Prostate® solution also facilitates the early detection of aggressive forms of the disease by incorporating advanced imaging biomarkers. These features ensure that clinicians have access to state-of-the-art tools to interpret and act on images of prostate cancer, improving patient outcomes and reducing healthcare costs.

With our holistic approach, and as global leaders in whole-body medical imagining analysis, Quibim sets out to transform PCa landscape.

References

1. Sandhu, S., et al., Prostate cancer. Lancet, 2021. 398(10305): p. 1075-1090.
2. Global Cancer Observatory. Cancer Fact Sheets; prostate cancer. 2020  [cited 2022 March]; Available from: https://gco.iarc.fr/today/data/factsheets/cancers/27-Prostate-fact-sheet.pdf.
3. Cancer.NET. Prostate cancer: statistics. 2021  [cited 2022 March]; Available from: https://www.cancer.net/cancer-types/prostate-cancer/statistics.
4. Loeb, S., et al., Overdiagnosis and overtreatment of prostate cancer. Eur Urol, 2014. 65(6): p. 1046-55.Prior, F., et al., Open access image repositories: high-quality data to enable machine learning research. Clin Radiol, 2020. 75(1): p. 7-12.
5. Bonmatí, L.M., et al., CHAIMELEON Project: Creation of a Pan-European Repository of Health Imaging Data for the Development of AI
6. Powered Cancer Management Tools. Front Oncol, 2022. 12: p. 742701.
7. ProCancer-I. An AI Platform integrating imaging data and models, supporting precision care through prostate cancer’s continuum.  [cited 2022 March]; Available from: https://www.procancer-i.eu/.
8. Cancer.org. Key Statistics for Prostate Cancer. [cited 2022 May]; Available from https://www.cancer.org/cancer/prostate-cancer/about/key-statistics.html#:~:text=About%20268%2C490%20new%20cases%20of,34%2C500%20deaths%20from%20prostate%20cancer
9. Hopkinsmedicine.org. Prostate Cancer Prognosis. [cited 2022 May]; Available from https://www.hopkinsmedicine.org/health/conditions-and-diseases/prostate-cancer/prostate-cancer-prognosis#:~:text=Even%20if%20your%20cancer%20was,mark%2C%20following%20the%20initial%20therapy