If someone asked you what illnesses or ailments scared you the most and you hope to never experience, what would you say? A frequent response is fearing the loss of something you were blessed with and don’t know how to live without. Imagine losing your eyesight or going deaf. Maybe it’s with age, but what if your body falls victim to a degenerative disease?
According to the World Health Organisation (WHO), approximately 1.3 billion people live with some form of visual impairment. 36 million of those are diagnosed with blindness. It might surprise you to know that around 80% of all vision impairments are actually considered avoidable globally, so what can we do to prevent this disability taking over?
For those of you who know your artificial intelligence, I’m sure you will have heard of Google Deepmind or more specifically, Deepmind Health. This AI technology is making increasingly positive strides in many aspects of the healthcare sector. The intention of this tech was originally to analyse images of eyes it was presented with. However, similarly to the way in which Google categorises images on a Google search, this AI unintentionally began to organise data based on whether or not the eyes showed signs of a specific visual impairment – Diabetic Retinopathy. It is a common condition that derives from damaged blood vessels in the eyes that result in a gradual loss of sight. Deepmind acquired 128,000 images of both healthy and unhealthy eyes and then assessed a further 12,000 images. It diagnosed both the disease and its severity, matching up to real-life diagnoses or sometimes even exceeding that of an expert!
Another illness that affects around 300 million on a global scale is asthma, with a further 100 million suffering from COPD (Chronic Obstructive Pulmonary Disease). COPD is a term for several progressive lung diseases that generally feature symptoms such as increased breathlessness, wheezing and chesty coughs. So where may you ask, can AI come into this? Any asthma sufferer knows that administering medicine through an inhaler is the most effective way to treat the lungs, yet around half of patients do not take the necessary daily medication. Harnessing AI with an inhaler enables the patient to be monitored in both technique for administering the drug and also ensuring they are actually taking the prescribed dose.
There is also the use of open application programming interfaces (APIs) in services such as Air. Digital therapeutics company Propeller have launched a service which takes on a preventative approach for asthma patients. Air can receive local, real-time updates of the environment around you to determine whether medication is required, or if the patient is likely to suffer on a particular day due to the atmospheric conditions.
The world of medical imaging is also seeing major advances in AI technology. Not only is it being utilised to combine CT-Scans and Computational Fluid Dynamics (tools that help pulmonologists see both the structural and functional aspects of the lungs), but it is used in regard to other illnesses as well. The Radiological Society of North America/RSNA showcases new software to be trialled and tested in imaging departments in hospitals. Samsung has just displayed the S-Detect this month at RSNA 2018, which uses ultrasound to analyse breast lesions. The software allegedly increases the accuracy of diagnosis from anywhere between 83-87% for doctors that have up to four years’ experience.
As we find ourselves living longer due to the great advances in medicine, diseases of the brain are becoming common and remain an underlying fear in the general public. Memory loss, Dementia and Alzheimer’s to name a few. One of the biggest problems with a degenerative brain is that time really counts. Without the recipient undergoing treatment in the early stages, it seems almost impossible to recover those neurons that used to fire valuable information. Not all hope is lost though, as there is promise in AI and PET scans making early predictions for Alzheimer’s disease later in life! By applying deep learning AI to find metabolic changes in the brain (signs that can be evident of Alzheimer’s), the technology could potentially predict the disease years before it strikes. The research conducted so far has trained the algorithm to measure the uptake of FDG injected into the bloodstream and then highlighted by brain cells – an indicator of metabolic activity and of course, potentially an indicator of Alzheimer’s.
The last use of AI I want to discuss is more generalised but is nonetheless important. There is an estimated shortage worldwide of over 7 million healthcare professionals, which results in overworked staff and some patients feel they are not receiving the care they truly need. No matter how big or small your suffering, whether it’s a cold or a degenerative issue, maybe we should go back to where it all begins – the doctor’s diagnosis. Doc.ai is a service that analyses data and through machine learning can provide answers to medical queries/concerns for a patient. Imagine that, having a doctor on your phone wherever you go! Their mission as an app state:
‘We are an AI company with the mission to decentralize precision medicine onto the blockchain. We believe that in the near future, human biological profile will be consumer-controlled, blockchain-based, AI-powered and omics-data-centric. It will be algorithmic rather than symptomatic, bottom up rather than top down, quantitative rather than qualitative.’
I have only covered a minuscule selection of illnesses and disease in this post, but they are perhaps some of the most concerning for the masses. Despite the negative stigma that tends to surround artificial intelligence, we can look at work fields such as healthcare and see that it is truly making a positive impact. Without AI, we would never have achieved these successes on our own, because this technology provides a unique way of thinking that humans do not naturally possess.