When saving time can be a matter of life and death, AI and machine-learning can be groundbreaking not only for care delivery but for healthcare as a whole
While the buzz around Artificial Intelligence has gained momentum, technology giants decided to utilise machine learning-based algorithms not just to assist organisations in driving growth but to also improve healthcare globally. Amidst such advancements, India ranks one among the early adaptors that has an integrated AI network across various growth pillars of its economy. This adoption of AI, categorically in healthcare, is growing while radically changing the face of delivery. The Union Government, during its interim Budget in February, stressed upon the top 10 areas for tech-enabled development with ‘Healthy India’ as one of the main agenda in its ‘Vision 2030’. Such progressive efforts reflect how the Government understands the potential of digital technologies like AI. To actualise this ambition, the Government is developing a National Artificial Intelligence Portal along with a national programme on AI; thereby creating an ecosystem for the adoption of technologies that support the continuum of care in a holistic manner.
Today, AI has opened up opportunities which were previously unavailable or unrecognised. These mechanisms are equipped to sense and comprehend data with human-like efficiency or even better, using big data and algorithms. Datasets, with help of algorithm intelligence, help in corroborating patient’s diagnosis with other similar cases on the basis of signs, symptoms and other reports. Some applications include using AI to analyse unstructured data and predictive modelling to manage patient flow and hospital capacity allocation. Although an intriguing field of study when it comes to life and death, in its early stages, AI is bridging the gap between healthcare service delivery and patient recovery. Diagnosing a patient for critical diseases, virtual nursing assistance, management of workflow and administrative tasks are some spokes of the healthcare wheel that ultimately help in seamless healthcare delivery to the patient. AI applications like Hybrid Artificial Intelligent Systems (HAIS) can address complex sets of data by using a combination of techniques like artificial neural networks, optimisation models etc. With their superior computing ability, such applications are unlocking limitations and assisting clinicians in a big way. Besides, they also allow for system adjustments for common sense, creating knowledge bank by mining raw data and adapting to changing environments.
Primarily, the AI application focusses on three medical tasks: Diagnosis, prognosis and therapy but mostly in the area of medical diagnosis. There are several techniques through which the medical diagnosis cycle can be replicated and the role of intelligent data system (the physician), input (patient data) and output (patient diagnosis) can be played simultaneously by a capable AI infrastructure. Diagnostic imaging with new methods of machine learning and deep learning is emerging as a powerful tool for translating large datasets into coherent knowledge for better care. This form of robotic learning will enable us to automate complex diagnostics and support optimal treatment. For instance, diagnostic AI applications gather and synthesise clinical data and compare information with predefined categories such as diseases to help with diagnosis and treatment. Raising analysis and interpretation of digital medical images to a whole new level, it paves way for quantitative, standardised yet personalised diagnostics while preventing errors in diagnosis.
Personalisation of care: Another field of healthcare that will be closely impacted by AI is personalisation of care. Fundamentally, it aims at establishing new relationship between people, professionals and system, thus leveraging the most out of the expertise, potential of people, families and communities. Going forward, in areas like cardiac imaging, AI-based image would be analysed along with lab results and aid in identifying high-risk patients, delivering personalised diagnostics and treatment and help prevent unnecessary treatments. Thus, involving diagnostic radiology more closely to not only being outcome-oriented but also reducing the healthcare costs.
Way forward: Moving firmly yet cautiously, AI research and development will guide us towards a healthcare system which is more trained and consequently intelligent. While leaving complicated matters and final call to human clinicians, AI systems can perform routine, less risky diagnostic and treatment processes. AI strongly promises higher automation, productivity and standardisation, along with unprecedented use of quantitative data; beyond the limits of human cognition in medical imaging. Here on, it is clear that the implementation of AI in practice will require interdisciplinary collaboration in which radiology experts have a significant role to play.
One of the primary drivers of AI solutions is data and, thus, appropriate handling, ensuring privacy and security is of prime importance. Challenges include data usage without consent, data selection bias and the resulting discrimination of AI models, risk of identification of individuals through data and asymmetry in data aggregation. In this interconnected world, AI tools can assist us with quicker services, diagnose problems and analyse data to identify trends that would predispose someone to a particular disease.
(The writer is Executive Director of a leading medical technology company)
