Medical Imaging, Machine Learning to Align in 10 Key Areas

12Feb - by aiuniverse - 2 - In Artificial Intelligence


Medical imaging and machine learning are on a collision course that promises significant advancements in diagnostics and precision medicine, according to a new report from Frost & Sullivan.

Advances in artificial intelligence and powerful computing technologies to support highly detailed imaging studies are driving opportunities for vendors and providers to capture a segment of a quickly growing market.

The firm predicts that the precision medical imaging market, worth $120 million in 2017, will explode into an $8 billion opportunity by 2027.

“Precision medical imaging has tremendous potential to improve all aspects of the care continuum, thus supporting emerging care approaches that are more targeted, predictive, translational, personalized and effective,” said Siddharth Saha, Vice President of Research, Transformational Health.

“AI-enriched imaging equipment will help adapt and personalize the imaging protocols and procedures while precise radio mic and phenomic datasets from the given clinical context will enable deep learning, thereby reinforcing medical imaging’s contribution to precision medicine. There are several firms in the ecosystem making very valuable contributions to the care pathways and this pool is set to exponentially grow in the short term.”

The report identifies ten areas in which medical imaging and machine learning will combine to bring greater efficiencies, more precise diagnoses, and innovative treatment options for patients.


Artificial intelligence can analyze massive volumes of clinical and imaging data to uncover patterns in the interplay between testing and long-term outcomes.  By generating evidence for when testing is needed and when it can be avoided, AI tools can not only create guidelines for ordering but can also reduce the billions spent each year on wasteful, low-value testing.


Capturing clear and complete images of physical structures can be challenging for certain populations, including children, the obese, and individuals with physical impairments as well as those with anxiety, dementia, or claustrophobia.

Advanced imaging techniques and personalized protocols for imaging acquisition, supported by machine learning, can ensure that providers can reduce patient stress while still capturing the necessary data for diagnostics and care.


Embedding intelligence into the imaging scanner itself can help to tailor imaging studies to the needs of the individual patient.  A collaboration between UC San Francisco and GE Healthcare, for example, is working to embed an AI-driven “library” of abnormal scans directly in its imaging machines to identify pneumothorax in trauma patients as speedily as possible.


Correlating imaging studies with the latest research and studies can improve the accuracy of interpretation and connect patients with optimal treatment paths.

Using AI to comb through millions of pages of academic literature to present decision support for providers can ensure informed decision-making.


Radiogenomics is a relatively new field that aims to correlate imaging studies with gene expression, particularly for cancer patients. By combining imaging studies with genomic data, providers may be able to identify cancers with much greater accuracy and offer personalized treatment options to patients based on their genetic and clinical data.


3D printing can now produce highly tailored implants for patients as well as reproduce complex structures in the body so surgeons can examine them in detail or rehearse a delicate surgery.  The more sophisticated the imaging, the more accurate and detailed the 3D reproduction can be.


Interventional oncology, external beam radiotherapy, and focused ultrasound are non-invasive techniques that require accurate imaging tools to provide guidance to clinicians.  AI can support real-time imaging that enables the delivery of these therapies while protecting surrounding healthy structures from exposure.


Correctly calculating the optimal dose of radiation therapy for cancer patients can be challenging for providers.  Machine learning tools can offer clinical decision support that accurately calculates dosages and plans therapies to ensure patients are receiving the right amount of treatment for their needs.


Theranostics is the combination of diagnostic and therapeutic tools into a single agent, while radiotracers are chemicals with radioactive components that can be watched for the rate of decay to monitor chemical reactions.  Radiotracers can be used to observe the metabolism of substances or the behavior of biological processes, giving researchers insight into the behavior of a drug, for example.

Imaging tools that can monitor theranostic radiotracers on a molecular level require AI to analyze the huge volumes of resulting data.  This area of exploration is likely to see significant growth in the next few years.


Precision therapies are intended to produce better outcomes at lower costs, but many organizations are currently struggling to generate business intelligence that would allow them to understand how their actions are affecting their bottom lines.  Artificial intelligence will play a key role in moving basic operational and financial analytics into a new realm of comprehensive insights.

As a result, organizations will be able to invest wisely in new products from vendors seeking to capitalize on the imaging analytics space.

But the ability to successfully combine machine learning with advanced imaging techniques is currently uneven across the market, Saha observed.

“While most major imaging companies are keen to make the most of the opportunities in precision imaging, they are at various levels of adoption. For instance, Siemens Healthineers has fully embraced the precision trend since it offers multi-pronged value through its solutions portfolio,” he said.

“At Philips Healthcare, a few precision hot spots have been forming, notably in image-guided therapies and oncology informatics. GE Healthcare, on the other hand, is looking to combine the precision paradigm with applied intelligence.”

Organizations looking to invest in AI-driven imaging technologies should carefully assess both their internal needs and the proffered capabilities of new products to ensure that they are partnering with a vendor that best suits their goals.

Facebook Comments

Comments are closed.