Artificial intelligence was at one time an abstract idea that lived in the imaginations of science fiction enthusiasts. It is, however, the nature of mankind to push boundaries and make what is conceived in imagination into a tangible reality. Although the idea of artificial intelligence can be and has been applied to a number of industries worldwide, its applications in healthcare are the most widely debated.
Artificial intelligence is based on the idea of deep learning, which uses architectures such as deep neural networks, deep belief networks and recurrent neural networks in applications like computer vision, speech recognition, natural language processing, audio recognition, social network filtering, machine translation, bioinformatics and drug design, where they have produced results comparable to and in some cases superior to human experts.
In theory AI systems need to “see” unstructured data, such as radiologic images, graphs, slide images, clinic notes, and literature texts, and then convert them to a knowledge base in a structured format. Companies worldwide are taking up the challenge of applying the principle to a complex and some would say, flawed healthcare system, in an attempt to streamline processes and establish order in a seemingly subjective, often chaotic science.
The focus of the tech industry has unsurprisingly been on clinical diagnostics and its applications in radiology and pathology. The idea is simple- to handover repetitive and time consuming tasks to a machine, so that the human experts can focus on the more challenging and complex cases.
Labor intensive and manual skill based tasks have been reduced with the advent of fully automated analyzers and machines. Digitalization of pathology, with the introduction of tele pathology and whole slide imaging led to faster diagnosis, giving rural areas access to these services and improving turnaround time. The introduction of e-health records has greatly helped with providing healthcare providers with patients comprehensive history and progression of treatment, and lets fewer patients fall through the cracks. It has also led to the reduction of patient identification errors.
Radiologists and pathologists diagnose by training the eye to recognize patterns and correlating them with clinical symptoms to come to a diagnosis . With the evolution of technology however, the amount of data that is generated, either from automated analyzers in pathology or the numerous images in a Pan Scan , is so voluminous, that diagnosticians, once maestros with a chest radiograph or a peripheral smear are now often visually fatigued . Deep learning , the core of AI , in the area of diagnostics is an autodidact- like an outstanding pathology or radiology resident, the more images or slides it analyzes, the better it gets.
The next step naturally progressed to the analytical aspects of diagnostics – how do we introduce technology in this sector to make the analysis and interpretation of clinical and diagnostic data simpler , faster and more effective ? Tech companies worldwide actively took up the challenge.
The most noticeable of the players on the scene is IBM’s Watson, an AI entity that has been making waves with the potential that it shows , especially in diagnostics. Patrick McNeillie, MD, clinical lead and senior architect of IBM’s Watson Health project, discussed how advanced clinical diagnostics algorithms can allow computers to relate diverse clinical literature with patient data to generate an “outcome report.” The software must also associate this with other similar “knowledge banks.” For example, cortisol can be either a hormone or drug, and can be differentiated by looking at context clues. So if cortisol is associated with “secretion” it indicates a physiological process, ruling out its use in this example as a drug. This contextual information-based approach helps the system be intelligent. The idea is to harness the speed of modern computers and programme them with cognitive algorithms and sophisticated decision trees to help solve complex medical problems.
To help address the growing number of samples from cancer, as well as other diseases, Fimmic, a Finnish software company, has developed Aiforia™ Cloud. It combines deep learning artificial intelligence (AI) image analysis cloud solution and automatised pathology image analysis capabilities. With this Fimmic aims to improve the speed and accuracy of tissue-based diagnostics and thus enable efficient, precise and more personalized care. Formerly known as WebMicroscope, Aiforia Cloud is already benefitted more than 6,000 pathologists, researchers and pharmaceutical R&D teams to manage and share digital slide collections. With the new Aiforia platform, CNN (convolutional neural network) deep learning algorithms can be easily developed in just days or less, compared to the months needed for conventional machine vision For the first time, the Aiforia deep learning algorithm can mimic the human observer in understanding the context in tissue and it can automatically and accurately perform laborious image analysis tasks in a fraction of the time. This gives users significant time back in their day for high-value work, such as analyzing more complex or rare samples.
“Digitizing the field of pathology has been an important step forward for healthcare, but we’re once again poised for real disruption in the space,” said Kaisa Helminen, Fimmic CEO. “An ageing population and rise in cancer prevalence, coupled with the radical innovation of AI technology, have created both a significant need and opportunity for digital pathology to evolve. The benefits of the system, like fully automated analysis from the whole tumor area; analysis from the area of interest, no double-staining needed for epithelium/stroma segmentation; fast, accurate and reproducible results and without the need for local hardware or software , are making it market competitive.
Path AI, world’s leading provider of AI-powered technology for the pathology laboratory, headed and co founded by Andrew H. Beck, MD PhD and Aditya Khosla PhD, is developing technology that assists pathologists in making rapid and accurate diagnosis for every patient, every time and is also building solutions to help identify patients that benefit from novel therapies, to make scalable personalized medicine a reality.
It is clear that artificial intelligence is geared to take the healthcare industry by storm. However, there are some who remain cautious about this disruption. One cannot help thinking – if the machine can do all the mental heavy lifting, what are the doctors going to do? Healthcare has always been an industry based on human contact and interaction. Human biology is a variable and inexact science and it requires physicians to collaborate and put together clinical observations and diagnostic data together to arrive at a conclusion. Clinical observations are honed by experience and are based on the physicians interaction with the patients , as well as visual and auditory cues that are picked up , sometimes subliminally. Can an algorithm really replace this?
Then there is the topic of employment. The healthcare sector is responsible for the employment of physicians, physician assistants, medical transcriptionists and technicians. Even though it leads to lower manpower costs for the industry and is an efficient business practice, it will lead to the future generations rethinking of career and education options and will change the job market as we know it. It also raises the question of accountability. At what point can we say: “It wasn’t me! It was the machine!”?
In India, even in Metropolitan cities, the idea AI seems farfetched and eons away . However, there are industry magnates who have already jumped onto the tech bandwagon. The Apollo group has been a strong advocate for the use of technology in healthcare and remains at the forefront of integration.
Most recently, Google caused waves when they announced their foray into the healthcare sector with AI, during their keynote address “Last year at Google I/O we announced Google AI, a collection of our teams and efforts to bring the benefits of AI to everyone,” Google CEO Sundar Pichai said. Healthcare is one of the most important fields AI is going to transform.”Pichai made his case by recapping the company’s work on interpreting retina images to detect diabetic retinopathy. Along with flagging relevant observations sometimes unnoticed by human reviewers, he said that the company found they could use these same eye scans to predict patients’ five-year risk of adverse cardiovascular events.“Last year, we announced our work on diabetic retinopathy, a leading cause of blindness, and we used deep learning to help doctors diagnose it earlier,” he said. “We’ve been running field trials since then at hospitals in India, and the field trials are going really well. We are bringing expert diagnosis to places where trained doctors are scarce.”
It is an exciting, historically disruptive time in healthcare. The idea that it is “ physician versus artificial intelligence” however , is counter productive. The best scenario is a synergistic relationship between the two, which showcases the best of both applied sciences, in order to do what the healthcare industry is called upon to do – improve the lives and the health of its patients.
(Disclaimer: Dr Lakshmi Vaswani, Assistant Pathologist, Bhatia Hospital. Views expressed are a personal opinion.)