Artificial intelligence at the service of radiology
Reading time: 2 minutes
Artificial intelligence, a tool in radiology
Many tools make up these modules. Here is a non-exhaustive list of the parameters that can be integrated:
- Annotation assistance (classification, segmentation detection)
- Data processing and structuring
- Automatic generation of medical report
- Detection of the most common anomalies (hernia, vertebral compression, spinal stenosis, spondylolithesis, etc.)
- Automatic measurement of parameters
Thanks to image synthesis, a global point of view can be obtained. Indeed, segmentation allows the output to be visually controlled. In addition, thanks to reconstruction and noise reduction, the patient has a shorter exposure time to the waves.
Investing upstream in future technologies
France decided to position itself very early on innovation in artificial intelligence in radiology. To do this, it was decided that it would transmit radiological “BigData” to industries. These, by developing competent artificial intelligences, introduce them into hospitals that have provided the necessary data. Today, tools are being put in place to assess the skills of radiologists and artificial intelligences. The observation is very clear, artificial intelligence is a very powerful tool, and their collaboration is very conclusive. Artificial intelligence is a tool to support medical decision-making, not diagnosis.
The relationship between programmer and radiologist
Collaborating with radiologists means enabling artificial intelligence to become a more reliable tool. Indeed, in order to guarantee an ideal medical imaging analysis tool, dialogue between data scientists and health specialists is necessary. Thus, each of the two professions will allow the other to identify different but equally important angles. Above all, it is necessary to work as best as possible to provide the best patient care. Radiologists are demanding transparency regarding medical artificial intelligence. Indeed, it is necessary for radiologists to understand how the AI they use works. The concept of a black box is unthinkable in the medical field.
Radiologists, volunteers for the idea of new technologies.
When artificial intelligence was mentioned in radiology, people panicked. Indeed, some radiologists thought that their role would be annihilated. Of course, this was a misperception. Indeed, and this is particularly true in France, radiologists are very efficient. Radiologists produce 120 million procedures per year, they obtain impressive results despite sometimes insufficient resources. Today, this idea has changed. Artificial intelligence, by appearing in institutions, allows time to be freed up and a place to be restored for relationships with patients.
Improving relationships between patients and radiologists: What is evident with the appearance of artificial intelligence in radiology is the evolution of the relationship with the patient. Indeed, while frustration could arise regarding the lack of contact with the specialist, the acceleration of processes changes this.
Now, the healthcare professional has the opportunity to devote more time to the patient and develop the social bond between them. Who says results given by a machine, says importance of an exchange with a person able to confirm and explain. Indeed, the machine makes it possible to promote the relationship, by reducing the time of proposing a diagnosis.
A revaluation of the personal aspect.
Moreover, what is most problematic for the patient is the lack of contact, communication, and feelings. Indeed, what a patient is looking for is support, transparency, and trust. Medical imaging analysis tools are only just beginning in radiology. However, artificial intelligence already brings major practical, physical, and emotional benefits.
