AI Cancer Care Solutions Changing Prevention and Treatment Strategies

By Laura Soto Published on Jun. 26, 2024

Cancer stands as a prominent global cause of mortality, yet, the landscape of oncology is evolving with numerous AI cancer care solutions, altering how medical experts tackle this disease. The realm of oncology, from preventative measures to treatment strategies, is progressing toward precision and individualized medical approaches. Historically, treatments were designed to cover a large group of individuals with a similar disease based on the average results from large populations. Nowadays, we understand that cancers vary between individuals, showcasing significant heterogeneity in terms of genetic expression and mutations, even within the cells of a single tumor. This variability underscores the importance of precision medicine, which considers a patient’s environment and its omic profile, concluding that, despite sharing the same disease, two patients can respond differently to the same treatment.

RELATED: Revolutionizing Healthcare with Generative AI: The Creative Code for Staying Well and Spry

The increasing quantity of data drives this more precise and potentially more successful approach. For instance, we can identify 120 different types of brain tumors, between benign and malignant. If we look only at one particular type, for example, glioblastoma multiforme, we see that only in eight genes there are 446 possible mutations. If we scale this to all kinds of tumors, the data grows exponentially, leading to a need for efficient analysis and interpretation. In addition to the genetic differences, the tumoral cells also showcase differences in phenotypes.

Moreover, novel drugs are continuously developed in the therapeutic space, accompanied by an increasing number of clinical trials. These examples showcase how oncology-related studies generate vast amounts of data, from genomic variations to clinical trial outcomes, each contributing to a complex and intricate assortment of information. This highlights the need for advanced analytical tools that can derive meaningful insights from all the information generated.

AI cancer care: Harnessing artificial intelligence to fight cancer

Due to its prevalence and complexity, cancer has become a target of innovative efforts by academia and the pharmaceutical industry. Artificial Intelligence, or AI, holds great power in this space. Our understanding of AI encompasses different terminologies, such as machine learning, a computer algorithm that enables computers to learn, and deep learning, a machine-learning method that forms networks to forecast performance. These techniques recognize patterns and connections from different models and datasets, playing a pivotal role in all the steps of the cancer journey.

1. How AI helps researchers during drug development

The fast evolution of scientific technologies has significantly impacted the oncology field. For instance, advanced multi-omics technologies like next-generation sequencing have enhanced our ability to discover predictive and prognostic factors. This evolution has enabled scientists to delve deeper into cells’ genetic, phenotypic, and behavioral characteristics, crucial for distinguishing cancerous cells from healthy tissue.

ai cancer_artificial intelligence data visualization

AI has emerged as a powerful tool supporting scientists across basic and translational research endeavors. By leveraging this technology, researchers can identify mutations, errors in gene expression, and abnormal behaviors associated with malignancy. AI also plays a pivotal role in optimizing laboratory workflows and extracting valuable insights from vast datasets, aiding in comprehending complex biological processes and predicting drug efficacy.

Within research, an emerging focus lies on drug repurposing, which involves identifying new indications for drugs already approved for certain health conditions. With the aid of algorithms for AI in oncology, the process of analyzing molecular properties of medications, scientific literature, and clinical data becomes faster and more cost-effective. This facilitates the discovery of novel diseases and conditions for which existing drugs may prove effective.

Startup highlight:

Owkin is an AI-based platform that works with deep, multimodal data collected from academic centers. They develop and design novel drug targets that can successfully work for a specific subgroup of patients. Besides the pre-clinical approach, Owkin also has an AI diagnostic module to pre-screen for biomarkers, giving clinicians the power to predict patient outcomes more accurately and treat them with targeted therapies.

2. AI cancer prevention and screening solutions — tailoring risk assessments to your lifestyle

We all possess a unique background and lifestyle. From inherited genetic variations to environmental factors, such as chemicals, alcohol, diet, or infectious agents that produce epigenetic modifications, or in other words, our behavior and environment that alter the way our genes work. Due to these differences, the risk factors associated with disease also vary from one person to another. Moreover, some types of cancer are more correlated with some risk factors than others, making it extremely difficult to track all possible prevalences at an individual level. Thanks to AI algorithms, it is possible to consider these differences, aggregating information on lifestyle and medical data to obtain personalized risk assessments and disease prevention recommendations.

Startup highlight:

Gabbi equips women with the tools for the prevention and early detection of breast cancer. With its proprietary algorithm based on claims data and user-provided information, Gabbi creates a risk assessment model that can assess the risk of developing cancer based on demographic factors like race or age, medical history such as genetic alterations, and family history (given the disease is highly heritable). With this information, women can know the prevalence of developing a disease and be aware of the necessity of periodic checks.

3. Supporting imaging and diagnosis

Time plays a crucial role in oncological diagnosis and treatment. A study published by Matthew Gitlin and colleagues in 2023 shows that US cancer patients experience an average diagnosis time of 5 months. AI offers a significant advantage in this space by reducing the time of diagnosis and allowing earlier treatment.

DOWNLOAD OUR EBOOK: Unleashing the Future: AI Innovation Reshaping Industries Today

There are several techniques for cancer detection, as Mayo Clinic summarizes; the evaluations can begin with a physical exam, and consecutively, most of them will have laboratory tests alongside, which can include blood tests or biopsies. It is important to discern healthy cells from cancerous ones for histopathological samples. Many of these tests are based on stainings and detected by antibodies, cell shape, or tissue distribution; however, it can sometimes be difficult to interpret them with the human eye. AI is a supportive tool that, when trained with images, can easily detect the different cell types, making the process less time-consuming for the experts.

Another evaluation technique is imaging; some of the most common techniques used are computerized tomography (CT) scans and magnetic resonance imaging (MRI), which are less invasive analyses used mainly for the hard accessible organs (e.g., brain or liver). Although protocols vary among countries, two clinicians usually need to agree on whether there is a disease condition based on these scans and other tests. Sometimes, they don’t come to an agreement, requiring a third clinician to determine whether there is a cancerous result or not. This not only takes time, but it is not infallible. Having AI tools as support can help accelerate the decision-making process, identifying abnormalities that can even be missed by the human eye, aiding in the correct diagnosis, and alerting on the most urgent ones.

Startup highlight:

In the past five years, AI (mostly deep learning and convolutional neural networks) has spread to automate breast cancer detection. In this space, Vara, a workflow platform for streamlining breast cancer screening through its AI power, is a great example. Combining AI into routine breast screenings reduces human subjectivity in reading mammography results. It does a pre-reading as well as a post-reading of the scans to screen cancer cases and identify potentially missed exams.

4. The role of AI in oncological treatment

Typically and whenever feasible, treatment begins with surgery and is then followed by radiotherapy and chemotherapy, with the aim of eradicating any remaining cancerous cells and preventing recurrence. In recent years, AI has been incorporated into the treatment cycle. Its assistant for surgeons, coupled with robotics, allows more precise guidance during operations.

Startup highlight:

MediView, born out of the Cleveland Clinic, is the first platform with FDA clearance for using live imaging and augmented reality for surgical guidance in soft tissue and bone structures.

Besides the traditional treatments listed above, there are emerging therapeutics such as immunotherapy, targeted therapy, and hormone therapy that are also utilized in contemporary cancer management. With the increasing number of available therapies, AI aids in selecting the most effective drug for a cancer subtype, aiming for personalized treatment, improving medical outcomes, and minimizing adverse effects.

5. Optimizing clinical trials: Getting drugs faster and better to patients

AI can serve as a potent tool in optimizing and streamlining clinical trials, enhancing efficiency and providing better patient outcomes. This technology is employed for various purposes, including the identification of patient subgroups that are most likely to respond favorably to a drug, optimizing the search for clinical trials by physicians and patients, or optimizing the study design to enhance resource utilization and communication between all stakeholders.

Startup highlight:

R.grid is a SaaS research platform that automates clinical trial administrative tasks. It offers real-time data capture and reporting, GDPR compliance and security, automated document filing, customization and branding, and anonymized data, among other functions, enabling all stakeholders involved to automate workflow.

6. Follow-up care and disease management

Pereira Cabral et al. highlighted follow-up services as one of the top fields where AI can have a huge impact in the next ten years. For the remote monitoring of patients, AI is developed for the analysis of digital biomarkers and chronic disease management (cancer is considered chronic once it is controlled with treatment, becomes stable, or reaches remission). It can aid in establishing personalized education on disease management, as well as tailored care plans, which enhance patient engagement and medication adherence through reminders for medication, follow-up, and visit planning.

The evolution of AI in the cancer journey

ai cancer-national-cancer-institute-cell

Artificial Intelligence holds great promise in transforming oncology care, particularly in predicting treatment outcomes and enhancing cancer screening and detection. With its capability for biomarker assessment and molecular characterization of cancer, the usage of this technology has vast potential to provide invaluable insights for prognosis and treatment decisions. However, substantial obstacles remain.

AI model development and training are key parts of the process, and it is essential to validate that the data used for it is curated and of quality. When data is not adequately checked, the model can be subjected to bias, potentially leading to discriminatory healthcare decisions. This has promoted a long-standing trend in testing AI algorithms to ensure the accuracy and fairness of healthcare decisions.

Integrating AI models into clinical practice is also complex. With many healthcare systems already in place, integrating AI becomes a time-consuming job. Interoperability becomes essential; seamless integration with electronic health records maximizes the applicability and scalability of AI applications.

Even when the interoperability constraint is overcome, there may still be some resistance to the adoption of AI within healthcare settings. This hesitation comes from concerns surrounding security, job replacement, or uncertainty about AI's effectiveness. It is crucial to highlight that AI serves as a support tool to streamline time-consuming tasks rather than as a replacement for the expertise and work of healthcare professionals. These professionals should be heard throughout the entire process to gather their feedback and address their concerns.

While the realization of AI’s full potential in oncology care may be years away, addressing these implementation challenges is crucial for unlocking its transformative impact on cancer treatment and patient outcomes.

AtPlug and Play’s Health vertical, we fast-track innovation in the healthcare industry.Join our platform today.