Using AI to standardize cancer care globally

Print Friendly, PDF & Email

Elekta and IBM Watson Health team up to harness data to guide treatment decisions

The collective intelligence and data of decades of scientific research and clinical experience could be harnessed to guide cancer treatment as easily for a patient in Bangladesh as for a counterpart in Brooklyn or Munich. Elekta’s and IBM Watson Health’s vision of cancer care is an individual in any part of the world receiving the same care – a treatment standardized across the globe by virtue of the tremendous speed of artificial intelligence (AI).

Attendees to the 2018 HIMSS (Health Information and Management Systems Society) meeting learned about this vision from Elekta’s Andrew Wilson, Vice President, Global Marketing Software and IBM Watson Health’s Steven Schudlo, Executive Director, Strategic Alliances, North America, in a talk entitled “MOSAIQ and Watson for Oncology: Standardization of oncology care means closing the oncology treatment gap for all.”

The essence of the companies’ collaboration is the intention to apply IBM Watson Health’s sophisticated cognitive computing to Elekta’s treatment planning and automation.

“Elekta is the first radiation therapy company that will offer a tested and approved solution that combines our MOSAIQ OIS with AI, deep learning algorithms and cognitive computing,” Wilson said.

“Elekta and IBM Watson are ready to synch and scale their already existing technological assets to close the treatment gap around the world.”

A world of disparity in cancer care, driving transformation

To set the stage, Wilson and Schudlo presented a wealth of statistics and facts about the accessibility and success of cancer management in different regions of the world – preeminent of which was the World Health Organization’s (WHO) sobering forecast of new cancer cases (excluding melanoma).

“By 2030, WHO predicts 24.6 million new cancer cases, which, ironically, is based on improved diagnosis, demographics and higher life expectancy globally,” he noted. “To adequately address this future patient volume, it will be imperative to first attempt to standardize how patients are treated in various parts of the world.”

Wilson said there exist two extremes when it comes to cancer care – one in which patients demand personalized care and another in which the patient simply wants care.

“In regions such as North America and several countries in Europe, competition among cancer care providers is intense and there is a need for personalized therapy that combines multiple data points and a multidisciplinary approach – in short, a highly sophisticated treatment,” he said. “Conversely, in regions like India, Southeast Asia and sub-Saharan Africa, there is a stark lack of available services and resources that need to be filled to afford patients safe and effective treatment. The result is that two women with stage III breast cancer could have completely different treatments and eventual outcomes based on where they live.”

The impact of the cancer patient’s country of residence has historically been a rather dramatic difference in survival rates, Wilson noted, referencing data from WHO and the International Agency for Research on Cancer (IARC).

“Countries whose residents have higher incomes have significantly greater 5-year survival rate,” he said. “Residents of low-income countries also don’t receive the same public service information on certain cancers. Lack of public awareness of breast cancer, for instance, remains a critical obstacle in developing countries where it is commonly diagnosed at an advanced, less treatable stage.”

Other disparities can be seen in the box below.

Additional challenges in standardization persist, Wilson added, noting that 35% of treatment plans deviate from evidence-based guidelines and that 45% of patients do not comply with complete prescribed treatments.

Closing the gap

The presenters observed that there are encouraging signs that cancer management capacity – particularly access to radiation therapy technology and techniques – is increasing in developing countries as the funds, space and human resources are becoming available to enable the construction of new medical facilities. The remaining major challenge has been staffing cancer clinics with the highly skilled professionals required to treat using today’s rapidly changing modalities and techniques.

“Cancer care is a complex endeavor all around,” IBM Watson Health’s Steven Schudlo remarked. “And this is especially apparent in the volume of information used and transmitted for the delivery of radiotherapy to a single patient. The data comes from multiple providers over many locations and visits and entails multiple discrete episodes of care for the same diagnosis. In addition, there is an array of different touch points over the course of the patient’s therapy and plans are constantly being adapted and modified. For radiotherapy, the data needed are current dose and fraction, doses to organs apart from the target, images, and plans related to treatment delivery and a full, true record of a treatment, including MLC leaves and couch position.”

To further complicate things, the specific radiotherapy related data is subsumed under a vastly larger fund of information that could be brought to bear in a single patient’s case. The convergence of genomics, precision medicine techniques and easier access to large electronic health record (EHR) datasets are driving rapid change in the oncology arena. Unfortunately, clinicians have limited time to consume and analyze all of the data that could help inform treatment decisions for individual patients.

“Additionally, concurrent, multi-modality treatment management is common, leading to an explosion of potential therapeutic options,” Schudlo said. “As these data sets swell massively in size, today’s oncologists may face challenges in consistently following treatment recommendations that are dictated by the most current clinical practice guidelines.

“We are in the unusual situation in which Big Data is causing big problems – oncologists are drowning in a sea of data,” he continued. “This has increased the demand for enhanced decision support capabilities that can distill millions of data points into patient-specific, evidence-based cancer care.”

Powered by extensive oncology data

An innovative solution, according to Schudlo and Wilson is a synergy between Elekta’s MOSAIQ OIS and Watson for Oncology, the latter the oncology application of IBM Watson Health’s Watson computer system.

Watson was designed by IBM in 2011 as a question answering computer that could answers questions posed in natural language; its first application was to answer questions on the quiz show Jeopardy! In addition to natural language processing, Watson’s abilities included information retrieval, knowledge representation, automated reasoning and machine learning technologies.

“Watson for Oncology is powered by the broadest field of oncology data,” Schudlo explained. “It searches over 300 medical journals, over 250 textbooks and 15 million pages of text in seconds to uncover insights relevant to the patient’s unique case. The product is ranked treatment options – supported by evidence – for consideration by the treating physician.”

From the patient perspective, Wilson added, Watson for Oncology will provide access to even more data-driven care regardless of their geographic location or where they are treated.

“This may allow patients to receive optimum care close to home, limiting or eliminating the need to travel to larger, specialty cancer centers for treatment and/or follow up,: Schudlo says. “Because Watson for Oncology expands insights into cancer treatment options, clinicians using the system can advise their patients on a range of cancer treatment options ranging from conventional treatment to clinical trials to precision medicine – such as genomics.”

By leveraging the remarkable power of AI, Elekta and IBM Watson Health are hoping to harmonize the standard of cancer care globally for all patients – regardless of where they receive their care.

“The two women with advanced stage breast cancer described previously – the ones who would receive disparate treatments because their therapies were arrived at using vastly different criteria applied in an inconsistent manner – will now have something in common,” Wilson said. “With MOSAIQ and Watson for Oncology AI, they both will be able to claim: ‘I get the best care I need.’”


Source: Globocan 2012. Within 5-year of diagnosis, International Atomic Energy Agency (IAEA) Expanding global access to radiotherapy. Lancet Oncol. Vol 16, Sept 2015. DIRAC (Directory of Radiotherapy Centres) 2012/IAEA

Next Previous Home