The term AI was coined way back in 1957. But it’s only in the last decade that we have seen an explosion of data, and data is the key fuel for AI and ML algorithms. As patient data and data collected through research is digitised, these algorithms can use it to detect patterns, and then assist health workers with early detection of warning signs as well as clinical decision-making.
Issues with public health programme
- Public health programmes are complex and dependent on committed human resources, who are in short supply and fairly difficult to keep motivated.
- These constraints limit the impact of large-scale health programmes, often leaving out families that need these.
- The progress made in the field of artificial intelligence (AI) and machine learning (ML) in the last decade can bridge this gap.
Usage of AI
From precision medicine, medical record storage and retrieval, medical report diagnosis, and robotics in clinical settings, to virtual consultations and personal fitness trackers that can be used at home, AI is making its presence felt:
Diagnostics and screening: Identifying or predicting diseases based on symptoms;
Health worker performance: Tracking the data captured by health workers, and using it to direct their efforts where they are most needed;
Improving client adherence: Identifying gaps in people’s health-seeking behaviour and suggesting who might drop out of a health programme or course of treatment.
The Astana Declaration on Primary Health Care identified technology as a key driver to improve accessibility, affordability and transparency towards achieving #HealthForAll.
Benefits of AI
- With the kinds of applications outlined above, AI and ML can be an excellent tool for the health workforce, making their lives easier and their work effective—when a few conditions are met.
- It can automate repetitive tasks, figure out patterns in huge datasets, and aid clinical decision-making in specific areas, particularly radiology and pathology. What conditions health professionals using AI/ML should ensure?
1. Get the right data: AI and ML algorithms are smart, but only as smart as the data that feeds them. The principle of GIGO (garbage in, garbage out) is applicable here, too. Any bias in the data—method of collection, populations and contexts covered, human error—will make the algorithm biased.
2. Be ethical: New developments like the EU’s General Data Protection Regulation are forcing investments in data security and privacy, but as public health professionals it’s important to think about ownership, access and use of people’s health data, before collecting it.
3. Get everyone on board: Getting non-IT people to accept the outputs of AI and ML can be an issue. If algorithms and processes are complicated (they often are), try and demystify AI and ML for teams that work on the ground.
4. Be clear about your objective: It’s important to not fall in the trap of setting huge objectives (like finding cure for cancer), but aim for low-hanging fruits and start with something well-defined and achievable.
AI and ML can seem daunting to those who don’t dabble in technology, so organisations should get some tech experts on board. They can help define achievable outcomes, design usable systems, and navigate the complex maze of resources available to turn those ideas into reality. What health professionals bring to the table is their understanding of the needs and context, their on-ground networks that enable co-creation, and their experiential insight into how these technologies will affect the lives of communities and health workers. Through such powerful partnerships, we can harness AI to power the movement towards Health for All.