HomeArtificial Intelligence and Precision Medicine Help the Fight Against Cancer

Artificial Intelligence and Precision Medicine Help the Fight Against Cancer

~~~Learn how AI can accelerate precision medicine to unlock the inner workings

Once likened to the Apollo space program that brought humanity to the moon, the completion of the Human Genome Project (HGP) in April of 2003 is considered to be a major milestone of medicine. The knowledge gained from subsequent research has shaped genomics, and is now making its way into clinical medicine.

By helping to map the entire human genome, researchers can now begin to understand the genetic basis for health and the pathology of disease. Instead of taking a one-size-fits-most approach to developing treatments for diseased populations, medicine has taken a turn towards prevention, personalization, and precision. A shift that can be put into everyday practice with the use of artificial intelligence (AI).

While traditional medical practice focuses on finding clinical solutions and treatment options based on the needs of the average person showing a generalized set of symptoms, precision medicine, driven by genome sequencing and data analysis, allows for personalized diagnosis and treatment according to specific genetic, environmental and lifestyle factors.

In this article, we’ll take a look at how AI and precision medicine can work together in the fight against cancer.   


How AI and precision medicine can fight cancer

Personalized treatment based on individual genetics and lifestyle factors can improve patient outcomes.

Artificial Intelligence and Precision Medicine Help the Fight Against Cancer

The National Institute of Health (NIH) defines precision medicine as an emerging approach to disease treatment and prevention that considers variability in patients’ individual lifestyles, environments and genetic makeup. Precision medicine enables researchers to more accurately predict which course of treatment and prevention strategies will be most effective for certain populations.   

Because of the complex analysis required for vast amounts of patient data, precision medicine will come to rely on AI-powered supercomputers and self-learning algorithms to help treat and prevent disease.

For example, cancer patients are typically prescribed a treatment regimen based on the type of cancer and its stage. A typical course of cancer treatment consists of a combination of surgery, chemotherapy, and radiation, and each treatment comes with its own set of side effects and challenges for the patient.

As you may have already guessed, patients’ responses to the same course of treatment and disease may look vastly different across two different bodies. The promise of precision medicine is to one day be able to tailor treatment based on a patient’s individual genetic response and spare them from treatments that aren’t effective.

Takeaway: Precision medicine will enable the accurate diagnoses and selection of the safest and most effective treatment for patients based on their specific genetic makeup.

AI in cancer research and diagnosis

Cancer is a genetic disease that is caused when abnormal cells grow and divide uncontrollably. Currently, pathologists diagnose cancer by examining the shape, number, mass, and appearance of cells for indications of whether or not the tissues are benign or malignant.

The process of finding the right treatment can take extended amounts of time. In some cases, treatment can change the biology of the tumor, rendering the initially prescribed drugs ineffective.

Recently, however, scientists at Weill Cornell Medicine and New York-Presbyterian developed an AI-powered computer program that examined more than 13,000 pathology images of various cancers and distinguished each type of cancer with 100% accuracy. In addition, it was able to distinguish cancer subtypes and biomarkers over 90% of the time.

Another powerful example of AI’s potential for diagnosing cancer lies in managing the lung cancer epidemic. Lung cancer is the leading cause of death in China, claiming over 600,000 lives each year due to high levels of air pollution.

Faced with the challenge of rural patients not seeking treatment until it was too late, along with a lack of radiologists, Chinese startup Infervision created a deep learning and image recognition algorithm that can be integrated with hospital systems to spot early signs of lung cancer in patients. This innovation has extended the reach of radiologists by providing prompt and accurate diagnoses.

Over time, advances in human genome sequencing technologies have made sequencing drastically more accessible. Whereas it used to cost $1B for complete genome sequencing, the price has fallen significantly over the last 15 years to only $1,000.

The combination of AI and human genome sequencing can further accelerate precision medicine.

According to a study done by IBM, New York Genome Center (NYGC) and The Rockefeller University, researchers were able to use IBM’s cognitive computing skills to examine genome frequency data and variants in only 10 minutes, compared to 160 hours of manual organization and human analysis.

AI and machine learning can automate the process of extracting pathological features from patients’ tissue samples objectively and consistently. Simply put, artificial intelligence can provide fast and efficient analyses of vast amounts of data that will advance the development of precision medicine.

AI and cancer treatment

Variations in genetics, responses to disease and treatment should mean personalization. There are now more than 200 different types of cancers that have been identified; worse yet, those cancers are constantly evolving. Advances in precision medicine and disruptive technologies will slowly put an end to the one size fits all approach to treatment.

Drawing personalized information and understanding genomic information plays a key role in cancer treatment. Treatment methods may be more effective for some patients than others due to variations in genetics and how a specific patient responds. The use of machine learning and predictive analytics can help doctors provide the best treatment recommendations.   

Machine learning has great potential to advance cancer treatment. According to Thomas Brown, MD, Executive Director of the Swedish Cancer Institute, “Few oncologists today have the complete training or time necessary to decipher complex results of a tumor’s biologic fingerprint.” In a recent collaboration between the Swedish Cancer Institute and precision medicine company GNS Healthcare, causal machine learning is being used to help support the best course of treatment.

Artificial intelligence can also help identify effective and personalized cancer-drug combinations.

There are countless numbers of cancer drugs available, but finding ways to combine them is challenging. For example, 100 drugs can result in a somewhat manageable number of 2-drug combinations, but when researchers want to test three or four-drug combinations, that number grows exponentially. Weill Cornell Medicine researchers have leveraged AI to predict the best possible anti-cancer drug combinations.

AI and cancer prevention

Bioinformatics is the science of taking sophisticated computing methods to large sets of biological data such as genetic sequences, cell populations, or protein samples to predict and discover new biology. In recent years, bioinformatics has proven significant in the advancement of cancer research by advancing the field of human gene mapping.   

The wealth of data from cancer genome studies can be used in conjunction with patients’ medical histories and clinical data to improve methods of predicting cancer risk, prognosis and response to treatment. Powerful AI algorithms will propel this field forward by helping researchers store, sort, analyze and make predictions based on data.

Takeaway: Cancer research, diagnosis, treatment, and prevention will be powered by artificial intelligence.


Illinois State University Awarded Over $660,000 Grant to Build a Cutting-Edge Microscopy Lab

Opportunities and challenges lie ahead

Unlocking the inner workings of health and disease is exciting, but consider the caveats.

Precision medicine powered by artificial intelligence will disrupt the way scientists research, diagnose, treat and prevent cancer. Despite AI’s potential to revolutionize the field, challenges do remain.

For example, the vast amounts of patient health data being generated are outpacing methods to analyze it all. Artificial intelligence, machine learning and cognitive computing will be essential to uncovering insights within all this data. Simply put, many treatments will depend on AI and data.

Integrating cancer genome data with electronic health records is crucial to enable better clinician decision making and treatment outcomes. Before any analysis can be made, researchers say there is a need for better tools to help them manipulate, study and analyze data, calling for the collaboration between researchers and informatics experts.     

The need for diversity in data presents another challenge of precision medicine and AI. While the goal of increasing our understanding of health and disease is exciting, it is imperative to be aware of biases that could result in gathering and analyzing health data. Recruiting diverse participant pools is critical in large-scale precision medicine in order to ensure fair representation in medical research.  

Takeaway: The advancement of precision medicine for the good of all patients will rely on improved data integration across sources, collaboration between researchers and informatics support, and recruiting diverse participant pools.

The future of AI and precision medicine

The completion of the Human Genome Project 15 years ago transformed the world of medicine, as the human genome paved the way for a more personalized approach to treatment. Now, the objective, efficient and automated data analysis powered by AI will change the field of precision medicine.

While the potential of unlocking the inner workings of cancer through precision medicine and AI innovation is revolutionary, several caveats still must be considered. Seamlessly integrating patient data from multiple sources, collaboration between researchers and informatics support, and recruiting from a diverse participant pool can help further advance precision medicine towards a cure in this lifetime.

We can’t wait to see what the future holds.