Artificial Intelligence and Technology In Medicine

Definitions

Artificial Intelligence: The development of computer systems to perform tasks normally requiring human intelligence such as visual perception, identification, translation and decision-making.

Current Uses

Diagnostics:

  • Used in ophthalmology to interpret eye scans to diagnose some conditions.

  • Used to classify types of cancers from photos in dermatology and CT scans.

  • GIDEON is a software that helps identify rare communicable diseases from data of where the patient has travelled from and their symptoms.

  • AI is used in IVF screening to determine the embryo with greatest chance of implantation.

  • Supermarket club cards are now being used to suggest risk factors of certain diseases (e.g the frequency and amount of painkillers you buy can indicate a risk factor for cancer before you consider going to a doctor).

  • Motion capture cameras (the same they use in CGI film making) can be used to diagnose motor diseases well before a doctor can. Traditionally, a doctor has to notice deterioration in movement to diagnose the disease which can take a year - this technology can do it in months.

Virtual Nursing:

  • “BioButton” is a small device that sticks to the chest just below the neck and can track heart rate, respiratory rate, skin temperature and a range of other biometrics. This can monitor patients overnight, removing the need for nurses to check the patient every few hours. There is suggestions that wearables like this can be used to monitor patients from the comfort of their own home.

  • Improvement in sleep technology to advise people about their rhythms and type of sleep.

  • CarePredict has created a wrist wearable (similar to a FitBit) for elderly people to monitor when they eat, go to the toilet, fall etc. which caregivers can access. It can also calculate risk factors of certain diseases from this data.

  • Suggestions to use robots to operate NHS 111. However, current AI technologies are not advanced enough and there is a negative public opinion of automated call services.

AI Assistants:

  • Assist doctors to search hospital guidelines and write prescriptions.

  • Instantly search for available physicians and schedule appointments.

  • AI is being trialed in America to transcribe patient consultations - recording only the important information.

Surgery:

  • The da Vinci Surgical System can be used for prostatectomies (removal of the prostate), cardiac valve repair, renal and gynaecological procedures.

  • Robotic surgeries have a 5 fold reduction in complications.

  • Augmented reality can also be used in some surgeries as well as navigation and tissue feature tracking.

Education:

  • Amazon Alexa has partnered with the NHS so that any medical information will be read from the NHS website. This helps the elderly and blind people access a trustworthy source of medical advice.

  • Ideas to develop a translation system in order to communicate with patients who don’t speak English (currently translators are used which are limited in supply and hospital dependent).

  • There has been a move towards sharing patient notes and scans with patients. AI can aid with the scanning and sending of this information to patients.

Will Robots Replace Doctors?

With artificial intelligence being integrated in our day to day society (Amazon Alexa, Chat GTP etc.), there are concerns the doctors job will be rendered obsolete, however:

  • Currently patient notes are kept on a range of interfaces: paper, different software and apps. These will need to be standardised to enable the transferal of information.

  • Concerns about data privacy and collection; encryption needs to be improved and mandated.

  • Doctors provide an empathetic and trusting environment, one that an AI would lack (e.g humans aren’t ready to receive cancer a diagnosis from a robot).

  • There are many ways to describe the same symptoms. Furthermore, many symptoms are made on observations alone, (how they walk, their facial expressions, the shallowness of their breathing, examinations etc.). As of current AI is not advanced enough to make these complex diagnoses.

  • It is very expensive to develop these systems and there is currently not enough public interest for them to be profitable. Therefore, there are no big development projects.

This makes AI replacing the job of doctors very unlikely (at least on the near future). However, it is probable that AI will continuously become more integrated in the healthcare and care of patients.