4 minute read
How artificial intelligence will help humanize healthcare
By Prof. Dr. med. Mathias Goyen
Chief Medical Officer, GE Healthcare Europe and Professor of Radiology at Hamburg University
Artificial intelligence (AI) allows us to “see more” and to examine more data faster. AI can lead to faster, more accurate diagnosis – and it has the potential to free up time for better dialog with the patient. As radiologists we should embrace AI.
There is an avalanche of healthcare data exploding. To give just two examples, this year 5,200 medical journals will publish more than 900,000 articles (1) . In one mammogram there is more information than phone numbers in New York City. Add to this that an algorithm can zero in on telltale patterns in the greyscale of the image much faster that even the most eagle-eyed radiologists, because it can see 256 shades of grey – while the human eye can only make out approximately 30 shades, and you can see how AI may become a strong ally for healthcare professional, including radiologists.
Digital is having a major impact on healthcare. It unlocks great growth potential and improvement for patient outcomes. As one of the largest generators of healthcare data through our medical devices, we’re well-placed to help shape the future of the industry using advanced analytics.
We may be able to see patterns in data to allow us to predict the course of disease, which would be a gamechanger.
AI will never fully replace the radiologist
AI can help make more information available for the radiologist and help him or her make the precise diagnosis earlier and pinpoint effective treatment. But in my opinion, AI will never fully replace the radiologist.
Radiologists are physicians specially trained in patient-diagnosis and care through medical imaging. At the same time, they are clinicians who interact at a human level with their patients. Radiologists now have the opportunity to spend more time on bringing their knowledge – augmented by AI – out of the dark rooms and transferring it to patient care through face-to-face communication.
Furthermore, the radiologist’s work involves much more than just image interpretation. Radiologists sit on tumor boards and they treat diseases - for instance in the huge and growing field of interventional radiology, guiding tiny catheters into an exact area of the patient’s body that requires treatment.
AI can help deliver better patient outcomes more efficiently
The healthcare sector in general is critically understaffed. Globally the sector today is short of about 7 million staff, and by 2030 the shortage is expected to be around 15 million (2) . A model where AI complements clinicians is a way to meet this challenge and contribute to better performance on several levels. AI and advanced analytics can be applied at three levels in the health system: individual – we insert AI into scanners; devices directly; departmental – assisting with workflows, and the third is enterprise level – hospital Command Centers.
Predictive analytics and systems engineering based on the huge amount of data collected can enable enterprises to implement more lean and flexible procedures and organizational systems and, in this way, enhance performance within e.g. patient volume, movements of patients through the hospital, patient safety and patient experience.
AI is becoming recognized by healthcare professionals
Bearing this in mind, it is perhaps not surprising that AI, from being viewed with some skepticism by healthcare professionals, now seems to be recognized by many for its positive effects. Nearly half (45%) of healthcare professionals in the U.S. and the U.K. have stated that AI has allowed them to increase time for patient consultations and surgeries, according to a survey carried out for a report published by MIT Technology Review Insights in December 2019 (3). And three-quarters of the participants in the survey reported that AI allows for better predictions in the treatment of diseases.
AI can help empower radiologists to do a better job. AI per se will not replace radiologist. But I believe that radiologists who do not leverage the power of AI may well be replaced by those who do.
AI can save lives
Scenario: It is during the night shift, and the radiologist is busy looking at the images of a trauma patient needing critical care. At the same time, a second patient’s X-ray images have arrived. AI-enhanced X-ray technology can analyze images and alert a radiologist to prioritize a patient with a critical condition such as pneumothorax (a type of collapsed lung).
Healthcare data explodes
- In 2010, it took 3,5 years for medical data to double (4).
- In 2020, this is projected to be 0.2 years – 73 days (5).
- In 2020, medical journals will publish more than 800,000 articles (6).
GE Healthcare
1) MEDLINE®: Description of the Database, 2019
2) Global Health Workforce Labor Market Projections for 2030, Human Resources for Health, 2017
3) The AI Effect, MIT Technology Review Insights and GE Healthcare, 2019
4) Challenges and Opportunities Facing Medical Education, 2011
5) Challenges and Opportunities Facing Medical Education, 2011
6) MEDLINE®: Description of the Database, 2019