Artificial Intelligence and Radiology in Indonesia

By Jum’atil Fajar*

Image: National Cancer Institute | Unsplash

Despite the common belief that Artificial Intelligence (AI) is a recent phenomenon, historical records show the use of AI in radiology started in the United States in the 1960s. In 1963, the study Computer Diagnosis of Primary Bone Tumors, reported on the progress of developing a computer program to evaluate bone cancer as shown on x-rays. To our knowledge this was the first study of its kind linking the use of AI and radiology, and setting the basis for the development of the sector.

More recently, efforts in the field have included the digitisation of 22,864 images from 1,664 radiology cases of bone tumors that were collected by Professor Henry H. Jones from Stanford Medical Center between 1955 and 2005. During the process, researchers annotated key images from 811 cases using the Annotation and Image Markup (AIM) standard. These data are now used for machine learning.

Progress in the field takes me to the question of how is AI in radiology being used in Indonesia? And how did the use of AI in radiology develop in the country? To answer these questions I conducted a series of interviews to better understand how the sector evolved over time.

The interviews revealed different approaches in public and private hospitals. In private hospitals, the practice of using AI in radiology is more widespread and supported. Public hospitals, on the other hand, are lagging behind mainly due to funding constraints and the topic not being prioritised at the moment.

Dr Pandu, a radiology specialist at Omni International hospital in Jakarta, mentioned hospital management provided X-rays equipped with AI technology. This tool has also been used for COVID-19, but has experienced some challenges along the way. 

A radiology specialist from the Doris Sylvanus Hospital, a public hospital from the Central Kalimantan province, explained that he has not been able to implement AI due to funding constraints and sub-standard facilities in the radiology room. He added that the implementation of AI in the near future is still not seen as a priority. Similar observations were conveyed by Dr Denny Muda Permana, a radiology specialist at Murjani Hospital in Sampit, Central Kalimantan. He mentioned that despite offers to implement the technology, they had been unable to do so because it is still not a priority in his department.

The issue is also one that goes beyond funding or priorities. Dr Ceva Wicaksono Pitoyo, a specialist in internal medicine and consultant in pulmonary diseases at the Cipto Mangunkusumo National General Hospital in Jakarta, argues that if AI only relies on image management, it will never be able to go beyond a radiology professor. Specialists and consultants still want to meet patients and ask questions (anamnesis) or review the patient clinical data. Dr Pitoyo also warns about only relying on AI for diagnoses. He reminds us of the important work of doctors and additional information doctors need to make a diagnosis, such as medical history, physical examinations, data on understanding pathogenesis and pathophysiology, laboratory data, and anatomical pathology data. The comments provided by Dr Pitoyo highlight the unique human skills and capacities that should always be present as part of the diagnosis process.

As stated in this study:

while the ultimate goal of machine learning algorithms and artificial intelligence may be to automatically learn from the data with limited or no human interaction, it needs to be recognized that achieving accurate results for complex image interpretation tasks such as medical images may require higher levels of cognitive processing.

In this context, the “human-in-the-loop” integration or so-called “interactive machine learning” becomes important and shows potential promise for other complex interpretation tasks in radiology. Approaches like this could help ensure we use AI in radiology to assist in the process without forgetting about the importance of the human presence.


*Jum’atil Fajar is an AI enthusiast. He holds a Masters degree in Health Sciences. He helped develop the hospital management information system. He currently manages the Hospital Accreditation Data Management Information System.

Next Event: AI Leadership Webinar Series
Friday, 30th July @ 16.00 CEST

Register here for this free event.

Register here for this free event.


Announcement: New Summer Schedule

Dear Readers,

For the months of July and August, the Leading with AI newsletter will move to a bi-weekly schedule. Our next issue will follow on Thursday (as usual) 12th August. Wishing everyone a safe and healthy summer.

Warm wishes,
The Leading with AI Team

Share Leading with AI