Artificial Inteligence, or AI, is no longer the mysterious technological unknown that it once was. In fact, it is now arguably woven into the fabric of our daily lives. Alexa tells us the weather and puts together playlists based on our music choices. Tesco or Ocado personalise our shopping discounts using data from previous purchasing behaviours. Apps predict menstrual cycles and Facebook even serves us ads informed by our browsing habits.
AI in Health
Through effective automation of processes and problem solving, AI is pushing the boundaries of innovation across practically every industry imaginable. And none more so than healthcare where its use is already relatively widespread. Spanning drug discovery, through automation of diagnosis to supporting patient engagement and adherence, the potential for impact in the sector is exponential.
With new proven use cases springing up on a regular basis, one area that has shown real promise is screening and diagnostics. A study published in Nature journal earlier this year highlighted the high potential for human error in the identification of breast cancer via mammograms screening. When AI was then introduced as a direct comparator, the technology demonstrated accuracy that surpassed human experts. At a time when the health system is already bursting at the seams and trained professionals are in short supply, integration of AI into the diagnostic journey would seem to make good sense.
It is also important to note that AI should not be viewed as a replacement for the work carried out by medical professionals. Instead, it should become an additional tool to enhance their capabilities. In the case of medicine, an AI diagnostic model can serve as an additional layer of support and validation for qualified doctors or nurses. This in turn can leave them to focus on other aspects of their roles that an AI machine cannot provide, like quality patient care.
Improving Outcomes for Patients
A collaboration between Moorfields Eye Hospital NHS Foundation Trust and DeepMind Health set out to find new ways to utilise the power of AI to support clinicians in their care for patients. The resulting programme is now able to recommend the correct referral decision for over 50 eye diseases with 94% accuracy, matching world-leading eye experts. Here again, AI continues to demonstrate its potential for the revolution in eye care diagnosis, enabling conditions to be spotted earlier and prioritisation of patients with the most serious conditions.
At TympaHealth, our focus is on the transformation of ear and hearing care. As with many specialisms, the journey to diagnosis for patients can be long, requiring many consultations with numerous doctors across a variety of specialities. Therefore, it’s our mission to bring ear and hearing care into the community. This includes placing better and faster access to diagnostic and treatment services. AI evidently has an important role in facilitating that.
There have been limited studies on the use of AI in the context of hearing and ear care. The use of smartphone otoscopes promise to help develop this sector further. Recent studies, including a recent collaboration from the author, use smartphone otoscopes to improve the medical learning environment. In many cases ENT experience is limited in medical school to just one week as a special study block, the advent of artificial intelligence would certainly help in recognising conditions of the ear and streamlining referrals.
Another study, published in Otology and Neurotology last month, showed that machine learning helped to predict post-operative performance of a cochlear implant, as well as identifying the influencing factors. This shows the sector is receptive to this change similar to specialties such as Dermatology and Opthalomology.
Building on these studies, at TympaHealth we have a team to help us embed the technology into our own processes and platforms effectively. In turn this helps us improve diagnostic capabilities and ultimately improve patient outcomes. Much like the Moorfields and DeepMind collaboration for eye health, we aim to use machine learning to assist us in identifying ear conditions. However, that is only the very start. We’re also exploring new ways to harness the power of AI to develop a fully integrated machine learning platform and well as using algorithms as predictors of future ear health deterioration.
A Vision For the Future
In the recent Topol Review which set out a vision for the future of digital health, Eric Topol reinforced the importance of preparing our health and care workforces for a digital future.
There is clearly a bright future for AI in healthcare, which has proven itself time and time again, especially at a time where technology and digital innovation is beginning to move front and centre in the mission to provide better and faster care. In the UK, our NHS system is already overstretched during usual times, and now with the additional pressures caused by the pandemic, it is imperative that we find new solutions that enable people to access the care they need, whilst relieving the burden on the healthcare professionals.