The use of Artificial Intelligence in our Hospitals.

The use of artificial intelligence neural networks to diagnose diseases and personalise medicine will revolutionise the whole hospital system because doctors will be able to more accurately determine what disease a patient is suffering from, the prognosis of the disease (how far along their disease is) and also create a personalised treatment plan for the patient in a small amount of time. One of the best things about this is that this advice and information will come with certain degree of certainty that can be measured. This degree of certainty would however have to be higher than what the doctor can offer. We trust doctors because we believe that they have been through rigorous training in order to become a final authority on our health and wellbeing. There are stringent laws that govern doctors and how they practice in order to make sure that they take the upmost care when treating and advising patients. This training and these legislations are the reasons why we put a high level of trust in them. For AI systems to be held on the same if not a higher pedestal of trust, there needs to be public evidence of vigorous testing that reinforces their safety and accuracy.

Artificial Intelligence brings the possibility of more accurate diagnoses that are not affected by a doctor’s bias or inexperience or a patient’s inability to report their symptoms in detail. Patients will not always be able to accurately explain their symptoms and they may forget to announce a certain symptom that they deem unimportant but could have immense implications on the doctor’s diagnosis. Therefore doctors may misdiagnose a patient because they don’t have all the information available to them. The AI system can instead use information from the patient’s blood test or medical scans to draw a conclusion on what their ailment is and what treatment course is best suited. 

Blood tests results and medical scans also assist doctors but the doctor has to interpret the results according to their understanding and what they learnt/ where taught in medical school. The truth is that the medical field is always changing as scientific research is always underway and new discoveries are being made. A doctor that graduated in 2015 cannot use that same knowledge in 2022, this is why doctors are always studying and being tested. People may wonder why it’s such a hard and gruelling process to become a doctor but it is that way because doctors give advice and provide treatment for the most important part of our person. 

The question is that once we have perfected our Artificial Intelligence systems to a high degree of accuracy and we have gained the trust of the public, should we do away with training people in the medical profession on how to diagnose people and prescribe them to treatment plans. Will we still need GPs? If a patient can have a machine at home that can take a scan of their body using computer vision and a small blood sample, providing them with a diagnosis and prescription, they would rarely need to book an appointment to see a doctor at the clinic. The role of a General Practitioner would then have to evolve and change to fit these new medical technologies in order for them to remain relevant.

The involvement of doctors in the creation of these AI systems will be crucial as when AI systems are training they are known for basing their predictions on random and irrelevant factors rather than medical pathology if the data used to train them isn’t properly chosen. An example could be that an AI model is trained to detect COVID in chest radiographs and instead of it looking at the relevant pathology that a radiologist does, the system focuses on the text blocks in the radiograph and uses that to predict the presence of COVID. Therefore, we need doctors and healthcare professionals to help perfect these systems. 

These AI systems will also have the ability to find new correlations in data that scientists and doctors could investigate the validity of and use to identify new causes of diseases and possible treatments. AI will add much more to us than it takes away and the implementation of these systems shouldn’t be seen as competition for the people who currently occupy these roles but as an evolution of our healthcare systems to massively improve patient care. 

‍

Key

Personalised Medicine - the tailoring of medical treatment to the individual characteristics of each patient