Cancer diagnosis has evolved as a frontline area for showcasing AI capabilities in healthcare.
Artificial Intelligence is the buzzword which we have been hearing a lot lately. It has almost encompassed in every sphere of our life knowingly or unknowingly. Companies, Researchers and Investors always remain on the lookout for high impact interventions that bring useful transformation in society. Furthermore, the application of Artificial Intelligence is making a great impact on cancer diagnosis. It also makes a lot of business sense also as a huge amount of money is invested in cancer-related research, medicine and applications. And, it makes a lot of difference to a cancer patient, if the cancer is diagnosed at an early stage.
Using Artificial Intelligence as a Diagnostic Tool
The diagnostic skills of highly sophisticated software have been tested and compared with the traditional diagnostic tools and experts, and found very helpful in diagnosis as well as prognosis of the disease. According to Beth Israel Deaconess Medical Center of Harvard Medical School led by Dr. Andrew Beck, it was discovered that the analysis of data through deep-learning had decreased the error rate in breast cancer diagnosis by 85%.
Similarly, Google is also working on developing an "Augmented Reality Microscope" which can enhance the functioning of existing light microscope "using low-cost, readily-available components, and without the need for whole slide digital versions of the tissue being analyzed. In addition, the researchers also compared these results with certified dermatologists. Surprisingly, they found that artificial intelligence can classify skin cancer as accurately as dermatologists.
Solid Tumor Solutions is SOPHIA GENETICS's (a Swiss company) approved cancer test kit that analyzes DNA samples of patients using an AI platform. This test can accurately detect mutations/ alterations in 42 genes that are linked with solid cancers. This cancer diagnostic kit is assisting doctors in more than 920 hospitals all around the world.
These neural networks are the most basic form of artificial intelligence. Machine learning is the branch of AI that is focused on teaching machines to be better at tasks iteratively. By developing algorithms that can help systems determine where they were right and where they were wrong automatically, the system could theoretically learn generations worth of data in a short space of time. Despite the theoretical soundness of the technique and the use of complex algorithms that can recognize behaviours and patterns, AI technology has only recently been able to offer the human-like insight and determinations required for it to excel in the medical field.
AI and elastography
AI has the complete potential of detecting the signs of breast cancer in females at early stages. The Ultrasound elastography is a relatively new diagnostic technique that tests the stiffness of breast tissue. It achieves this by vibrating the tissue, which creates a wave. This wave causes distortion in the ultrasound scan, highlighting areas of the breast where properties differ from the surrounding tissue.
From this information, it is possible for a doctor to determine whether a lesion is cancerous or benign.
The researchers wanted to see whether they could train an algorithm to differentiate between malignant and benign lesions in breast scans. Interestingly, they attempted to achieve this by training the algorithm using synthetic data rather than genuine scans.
Learn like a human
While it has the ability to understand the meaning of language and can develop on its own via machine learning, Watson still has a way to go before it can be introduced into the real world as an effective assistant. But even today, AI has shown in potential in some specialized medical tasks, with human help. According to a recent Northwestern University study, AI can outperform radiologists at cancer screening, especially in patients with lung cancer. The results show that using AI cut false positives by 11%. The medical field might not be so far away from having its own well-trained AI delivering proper diagnoses. It all depends on how fast AI technology advances and how quickly it can learn to diagnose like a human physician.