Prof. Ghulam Asghar Channa*
Artificial intelligence (AI) has revolutionised the world. Its effects have permeated every sphere of life and it is reshaping every fabric of society. In the field of medicine, it is not a futuristic premise but is already in use for saving lives by helping doctors in early detection of disease, predicting prognosis, and anticipating disease outbreaks through surveillance. IBM’s Watson AI lab can analyse 20 million cancer research papers in less than ten minutes and can instantly advise doctors accordingly. Google health/deep mind, Microsoft cloud for health care, Amazon health lake, Path AI, and butterfly network are a few AI systems used for screening of cancer, predictive analysis, population health, examining histopathology slides, and ultrasound interpretation by portable ultrasounds in remote areas. Robots have revolutionised surgery to be conducted with precision. Health care is in transition in the world due to rising cost, fast growing technology, and dwindling resources. Providing equitable health services in low- and middle-income countries (LMICs) is a challenge.

Prof. Ghulam Asghar Channa
Advances in AI and mobile computing power has raised the hope of meeting this challenge and galvanised researchers to use machine learning, LLM (large language models), and processing visual and auditory computers signals for health interventions in diseases common in LMICs. Over the past decade, pragmatic use of artificial intelligence (AI) has begun to record improvement in health delivery in resource-constrained situations. AI tools—from image-based diagnostics, telemedicine platforms, and wearable sensors to monitor people for the development of early signs of disease—have shown improved health outcome. However, research in ethical, regulatory, and societal effects of widespread use of technology have not been adequately addressed.1
Currently more than 130 countries in the world have digital health strategies and AI partnership. Following are a few practical examples of the use of AI in LMICs.
• Nepal, adopted the IKAROS software, with Deep Neural Networks to automate chromosome analysis and to predict cancer and genetic disorders in the early stages.
• India’s Qure.ai, a health tech company, has leveraged AI powered medical imaging solutions to detect tuberculosis, pneumonia, and other lung lesions on chest X-rays (qXR) in remote clinics. Using its work flow tools (qTrack), it tracks the patient outcome and referral pathways. Similarly, in South Africa RAD -AID and AI-based radiology tools are used to detect early lung changes for diagnosis of tuberculosis, cancer, pneumonia, and other lesions.
• Rwanda’s Babyl Health programme, a 10-year public private partnership, combined AI chatbots and video-consultations to extend primary care to rural districts, enabling 2.8 million Rwandans to access health care in remote areas.2
• In Nigeria, based on voice analysis using voice sensors of smart phone, normal cry of a new-born is differentiated from the one due to hypoxia in the early stages; the change alerts workers to provide oxygen to save the infant from deleterious effects of hypoxia.
• Similarly, Indonesia, Brazil, and Philippines are using AI tools for early detection of diseases and provide consultation through chatbots to reduce work load in hospitals and that of doctors.
These projects have improved accuracy in diagnosis, have saved time, and expanded access; at the same time, they have generated local data on health to refine algorithms for developing more AI tools.
Pakistan in accordance with the World Health Assembly resolution of May 2018 is using digital technology to achieve Universal Health by 2030.
The national centre for artificial intelligence with its head quarter in National University of Science and Technology, Islamabad, has been established to develop AI systems for use in various fields including health.
Telemedicine platforms like Sehat Kahani, Marham, and Nayajee are used to connect patients and doctors to decrease waiting time for treatment and detect some diseases like retinopathy in susceptible individuals. These efforts showcase the potential of AI for wide spread use in multiple diseases, workflow management, and early detection of diseases.
Barriers in harnessing AI in Health and possible solutions
Given my experience as a surgeon who has worked in a university and district hospitals, and taking into consideration the growth of AI, and review of the current literature, I believe that the following are the possible barriers in widespread implementation of the AI in healthcare system.
• AI systems are subject to bias and likely to miss the subtle content in health which may mislead the user.
• Ethical reasoning, critical thinking, problem solving, empathy, and understanding the nuances in complex health issues are the virtues of a clinician. Targeted training of medical leaders to use AI tools as assistants rather masters will promote adaptation of AI in healthcare delivery.
• Feeding personal information raises legitimate concerns for the misuse of privacy.
• Tertiary care centres and university hospitals must be equipped with a team of AI technologists, data engineers, bioinformatics, audio and video communication engineers, ethicists, and physicians. For these new posts can be created or partnership with AI companies be explored. On site AI solution will address issues of privacy and data protection.
• The three tier healthcare system should be connected; as the district are tagged to medical colleges, primary and secondary health care units should be tagged to tertiary hospitals.
• Pakistan introduced virtual asset bill of 2025, to stay out of the grey list of FATF. The experience revealed that legislative reforms and global cooperation in AI increase the capacity to curb misuse of technology. On similar lines, virtual health regulatory bill, global cooperation, and institutional strengthening will increase trust and control the privacy of people in Digi health.
CONCLUSION
In developing countries using AI in healthcare has showcased tangible wins. Policy for adapting AI and capacity-building of staff will give momentum for achieving health-related SDGs by 2030. Transparent public private partnership, engagement of AI companies and regular audit of the process, (PPPEA) may make Pakistan champion of AI in health in LMICs. However, it is feared that if the avalanche of AI is not handled by the leaders of developing countries today, the population here will be adversely affected in the near future—not just in terms of technological lag and brain drain but also in deepening disparities in healthcare and missing out on life-saving innovations.
REFERENCES
- Nina Schwalbe, Brian Wahl, Artificial Intelligence and the future of global health, Lancet 2020; 395: 1579-86.
- Rwanda. Our experience, delivery model and what we learned so far, www.babyl.rw. (Cited July 18 2025)
*The author is a former Vice Chancellor of SMBBMU, Larkana.