Mangalore University: “AI combines images with the reports available to give final results. There is no doubt AI is an artist as it has the ability of increasing images without damaging the original quality of the image from existing images and detecting disease symptoms at early stages with help of raw data and it diagnoses every image uniquely”, said Dr Pardan Setia a Diagnostic Radiologist from University of Toronto during the 5 day conference held at Mangalore University on AI in Medicine proudly brought together by the Department of Research and Computer Science of MU and Global Initiative Academic (GIAN), the conference started on the 18th ending on 23rd August 2025.
Dr. Pardan Setia is a professional radiologist and researcher who is presently pursuing a Master of Health in Translational Research. He is dedicated to bridging the gap between scientific advances and practical healthcare applications. With a strong background in medicine and diagnostic radiology, including a Bachelor of Medicine and Bachelor of Surgery, a Diplomate of the National Board of Radiodiagnosis, a European Diploma in Radiology, a Diplomate of the Indian College of Radiology and Imaging, as well as extensive experience in interpreting imaging studies.
While addressing the audience, Setia emphasized the critical relevance of applying AI to Radiology, stating that health care is now a thin science, and engineers’ knowledge with algorithm machine learning can have a direct impact on how patients are diagnosed. He stated that India now has one radiographer per lakh people, and that the disease burden is very high, with a high likelihood that they missed the early indications.He explained that radiography is now dealing with excessive workloads, reporting delays, and other issues, and that with the help of AI, substantial improvements would be made to help transform. Setia identified the most effective areas where AI may be used in healthcare, including image improvement, AI-enabled decision making that aids in the interpretation of health-related concerns, and real-world existent tools and apps such as CADX (Computer Aided Diagnosis) and CADe (Computer Aided Detection). Setia went on to explain why AI is needed in radiology, including reduced workload, increasing CT and MRI volumes that are outpacing the workflow, and increasing variability and delays since AI can speed up this process.
Setia noted that the CADX system detects symptoms that people cannot identify by detecting subtle abnormalities and changes, while also being created with few errors. “Not only is it a major changer but a safe tool for radiologists,” he told me. Unlike traditional methods of using filters to smooth noise, AI has the ability to enhance images. Setia stated that they have the ability to obtain raw data to enhance the image, similar to Photoshop. The analogy has the ability to sharpen images, increasing pixels with the goal of sharpening edges, fewer artifacts, and better signals to noise reductions, which enhances deep learning. Furthermore, it improves visuals in the health field, minimizing misdiagnosis, poor outcomes, and speeding up reporting, resulting in fewer repeat cases and reduced radiation exposure for patients.
Setia emphasized that in states like India, AI algorithms are game changers, particularly in governing sectors, because they are rapid and inexpensive. He stated that the CADEX AND CADE software has the capacity to detect cancer sooner, which is critical for patient survival, such as brain tumours, lung illness, and breast cancer. The early indicators are so small that they cannot be identified. As a result, artificial intelligence aids in the early diagnosis of these nodules. Which was typically tough due to obstructions.
AI in Health Shortcomings
Although AI is revolutionizing the medical industry, it also has hurdles and drawbacks. Setia stated that false positives identify normal results in patients as anomalies, resulting in unneeded procedures. Another difficulty is that generality and bias are high because this AI tool was trained on Western populations, and using it to foreign populations such as India may result in missing illness patterns. Aside from that, AI requires massive volumes of data sets, and some countries, such as India, may have challenges due to a lack of digitalization, and AI requires a high workflow with reporting.
AI in medications, like any other application, requires permission, and with digitalization, there are significant concerns about privacy and data security, he stated. Setia went on to say that the crucial component is that medicines must grasp the need for AI, and when it comes to accountability, the discussion over who is to blame in the event of a fault eruption is still ongoing. Health professionals and data engineers may shape AI in health care through federated learning, real-time CADE image AI, and more human-friendly approaches.
Dr Pascal Tyrrel, a Data Scientist and Director of Data Science at the University of Toronto’s Department of Medical Imaging, also contributed to the conversation by sharing his perspective as a data science engineer. Tyrrel remarked that these AI-based algorithms and data are typically secured and intended for commercial usage. Some of the data provided by IA is forbidden and should not be used without authorization. He stated that accessibility should be achieved through code swapping between data engineers and health organizations that choose to share them for study or profit. He also mentioned that other standard medical technologies are expensive, and AI has the potential to change that.
Cost-Effective Of AI
Tyrrel also underlined the cost-effectiveness of AI integration in healthcare organizations. Touching issues include the waiting period during cancer screening, as well as the cost and effectiveness that comes with early disease detection. “The AI tool tool can change practices such as rapid and accurate data, reduce diagnosis delays, and many more,” Tyrrel continued, adding that it is critical to consider negotiating with the government to purchase the AI tool, as well as the impact the AI tool has on society and their income level if they can afford the services.
Complications Of AI To The Existing Health Tools
During the discussion, participants stressed how crucial Ai is in providing accurate information because it prevents false detection, which not only makes people feel awful but can also be fined by law. While some are concerned that the Indian people in rural regions cannot purchase current health instruments such as CT MRIs, and government hospitals lack technicians and CT scanners. Stating that progressively integrating AI tools in states such as India’s health sector adds complexity to the existing situation rather than improving it.
Current Status of AI in India
In the recent media report AI has already made its way into Indian medicine with recent updates of the launch of the Department of AI at KMC Manipal in healthcare. This strategic effort of KMC, Manipal is set to reshape the landscape of Indian medicine, creating a new benchmark for medical education and training healthcare professionals to face the challenges and possibilities of the twenty-first century. AI in Indian medicine is transitioning from experimentation to tangible action. From institutional research and hospital uptake to grassroots innovation, AI is reshaping healthcare. To fully harness its potential, India must invest in infrastructure, workforce development, ethical frameworks, and public trust.
