AI Helps Fight the War Against Cancer

Despite being still fairly new on the market, AI is often seen as being something to produce eye-catching images and impressive written content, but it also has the power to help tackle some of our big social and technical issues.

Complex political, scientific and technology issues often require solutions that are simply too difficult for even the best minds to figure out, and remain unresolved. Cold fusion, global warming issues, future international migration patterns, and advanced silicon chip design are all pressing issues, but are so complex that even our most powerful computer systems are not able to resolve them. AI systems, however, offer a whole new perspective on dealing with complex data systems and big data pools, and are starting to show how we can actually start to consider solutions to the very big issues.

Of course, one of the major problems in our society is the prevalence of cancer in its many forms, and finding cures for this insidious disease is a priority for AI systems. Luckily, a team of researchers at CellChorus – a spin-out from the University of Houston – are now firmly focused on just that, and their results could become the benchmark for how cancer research is undertaken in the future.

Under project “TIMING,” (Time-lapse Imaging Microscopy in Nanowell grids), the company has created a platform for the dynamic and in-depth investigation for single cells via a series of high-quality captured videos, and lets the researchers investigate changes over time. It generates tens of thousands of individual images by observing cell interactions using a video-array-based technology. Crucially, however, analysis of these enormous video arrays calls for the use of an automated computer vision system, and that became a major issue.  

While almost any image-analysis system is capable of assessing and comparing images one at a time, the sheer scale of the project called for something that was not only much faster, but could also maintain complete accuracy for the entirety of the process. That is a very tall order, but one that AI was perceived to be capable of. 

Initial tests of the system have proven that the AI is more than capable of not only processing the mass of data but in also identifying any anomalies present in the images. This is more than an encouraging start and the team are investigating how the project can be expanded. Through a grant of $2.5 million from the National Centre for Advancing Translational Sciences of the National Institutes of Health, the startup company has been able accelerate the development of an advanced “label-free” version of this technology in collaboration with the University of Houston.

Rebecca Berdeaux, chief scientific officer, says: “by combining AI, microscale manufacturing, and advanced microscopy, the label-free TIMING platform will yield deep insight into cellular behaviours that directly impact human disease and new classes of therapeutics.”

This method of study, known as “label-free analysis,” enables researchers to see cells in their normal condition and collect vital data regarding the mobility, interactions, and changes that occur inside them. Additionally, it will make it possible for them to examine novel molecules of interest through the use of selective fluorescent staining. This is helpful in the study of illnesses such as cancer as well as the way cells respond to certain therapies.

The AI element of the process in this case is aimed at the early identification of potentially cancerous cells, but now has the power to offer much more. In addition to assisting with diagnostics, artificial intelligence helps clinicians generate predictions about a patient’s future requirements. Additionally, it is a game-changer for other procedures that are more complex. Real-time surgical intelligence may now be provided by artificial intelligence to medical professionals, enabling them to navigate past impediments and synchronise their motions with digitally presented three-dimensional surgical routes. Surgeons are also able to identify a person’s anatomy with the use of AI scanning, which enhances their knowledge of the specific placement of the anatomy within the body, considerably reducing the risk of injury to the patient. AI is becoming one of the most powerful tools in the global fight against cancer, and is frequently being used to:

• Personalise therapies. AI helps personalise cancer treatment by analysing patient and cancer cell data to create targeted therapies. This approach attacks cancer cells while protecting healthy ones, and leads to quicker recovery. 
• Reduce false positives and negatives. AI improves cancer detection by reducing false positives and negatives in results. For example, doctors often misidentify 1 in 10 mammograms as positive. Google’s AI software has cut false positives by 6% and false negatives by 9%.
• Classifying tumours without invasive procedures. AI can classify tumours without surgery. For example, doctors sometimes perform surgery only to find a tumour is harmless. AI can prevent that trauma.
• Detecting cancerous cells in early stages. Intelligent systems can detect many cancers early by spotting changes in medical images that humans might miss. Early detection can save lives. 
• Accelerating critical diagnosis. AI speeds up the diagnosis of cancer by analysing tissue in seconds, compared to days with traditional methods. This allows for targeted therapies to be applied much quicker.

The medical world is usually very wary of new technology and typically waits for it to be fully proven before making use of it, but AI has proven its worth, and is only going to become more prevalent in medicine.

 

Unity Developers are a professional development company. If you have a software project that you need help with, come and chat to us and see how we can accelerate your development.