MemberlistThe Library Index  FAQFAQ   SearchSearch   RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 
Artificial intelligence - is it any way intelligent at all? (NEW CONCEPTS)
Reply to topic
View previous topic :: View next topic  
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

Health care is a complex adaptive system.

Diagnosis

Clinical diagnosis is one aspect of this system and is an additional layer of sophistication, as it relies on complex interactions between clinician and patient.


Making a diagnosis is usually a ‘process’ rather than an ‘event’. It involves aspects of deductive reasoning, hypothesis testing, intuitive thought and pattern recognition; in addition to re-testing on the basis of new information provided by patient responses, physical examination, laboratory results and imaging.


The problem of undifferentiated patient populations


Pre-hospital care, primary care and care delivered in emergency departments involves the fullest possible range of clinical diagnostic acumen, as these settings provide advice and treatment on completely undifferentiated patients. That is, patients with no diagnostic pre-filter.

People who ring 999 or turn up at a GP practice or Accident and Emergency with no predesignated diagnostic label pointing their first contact care clinicians to a specific diagnosis or narrowed range of options.

Edging towards a clear diagnosis in these undifferentiated patient populations, is often therefore very challenging particularly where an ageing demographic (often with a range of co-morbidities) present to the clinician.

Where does AI fit in?


Faced with these difficulties it is unsurprising that science and the lay community look to computing advances for assistance. The dawn of precision medicine, pharmacogenomics, and especially artificial intelligence (AI) seem to offer new tools for the struggling clinician in first contact care.

Indeed, early reports seemed to suggest that we could one day dispense with the primary care clinician and instead rely on super-specialists to deliver ‘cures’ once the AI engine had made a definitive and accurate diagnosis.



AI can be defined as technologies which can perform specific tasks quicker or better than humans. A subdivision of AI is the concept of machine learning (ML). Which is the application of algorithms to recognize complex relationships or patterns from empirical data, in order to make accurate decisions.


Using ML, a computer is provided large amounts of data and a set of algorithms in order to perform a task. The data then reinforces correct answers, so that continued education of the machine occurs with no additional programming.


Current AI applications are usually to be found in defined diagnostic groups to direct therapeutic advantage, for example refining the assessment of changes in diabetic retinopathy over time or machine learning enhancing the diagnostic accuracy of potential breast carcinomas on mammography.

These are focused and specific to predefined patients with a clearer and more linear diagnostic question in place. Such as ‘has this person got a breast cancer?’ Yes or no. Has this person’s diabetic retinopathy worsened since the last image was taken? Yes or no.


These areas of medicine are clearly important, and this use of AI could have a definitive and positive impact on these patient groups. However, even within these narrow areas of clinical activity issues of rare disease prevalence creates a significant boundary for AI. Other issues of reliability, and the potential hidden nature of the decision made by the machine, so-called ‘black box decision-making’ have raised concerns more widely about the applicability of AI in medicine.


Is it all just sci-fi?

Beyond these concerns, it is also a very long way from a patient with back pain, headache or fatigue presenting to an AI machine and then being given a definitive diagnosis based on their clinical history alone.

This is categorically the case now, but what can we expect from AI in the future?


Arthur C Clarke advises that ‘when a distinguished scientist says that something is possible, he is almost certainly right. When he says that something is impossible, he is very probably wrong.’


This implies that it would be foolish to deny that developments in AI will push medicine forwards in machine learning and associated tasks However, will we ever see AI which fully replicates autonomous decision making and diagnostics in an undifferentiated patient population?

I think it unlikely.

My new thoughts

I think that instead we should look forward to an augmented AI approach, where it used as a tool by clinicians across medicine. Employed judiciously in this way AI should improve diagnosis, enhance effective therapies and ultimately improve outcomes for patients. I do not however, expect GPs, paramedics or emergency care clinicians to be replaced anytime soon, if ever.

Thoughts?
Send private message
N R Scott


In: Middlesbrough
View user's profile
Reply with quote

Is this the long way of saying will human doctors be replaced by robot doctors?

The doctor will eventually be replaced ..by the patient. Certainly if the trend in people googling their own symptoms is anything to go by.
Send private message
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

No, I really cannot conceive that doctors will be replaced with robots.

As for being replaced by patients, an interesting idea but sadly all the Boolean algorithms in the world will fail to resolve a diagnosis someone who has undifferentiated symptoms.

Especially sometimes when as patients we do not know what our symptom denotes.

For example, breathlessness can be caused by hundreds of different diseases (and other factors).

AI will become a useful tool in the armamentarium deployed by real-life doctors - my thinking is that anyone who proposes that it will be 'the answer' to healthcare diagnosis and treatment is unlikely to be correct and may have a hidden agenda in pushing certain solutions or nostrums.
Send private message
Wile E. Coyote


In: Arizona
View user's profile
Reply with quote

It seems to me we should chip everybody and then run yearly diagnostic tests.

We would catch problems early and save NHS budgets.
Send private message
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

If only life were that simple...

Screening is not easy and can result in unintended consequences.
Wilson and Junger proposed criteria which are worth considering, which are

The condition sought should be an important health problem.
There should be an accepted treatment for patients with recognised disease.
Facilities for diagnosis and treatment should be available.
There should be a recognisable latent or early symptomatic stage.
There should be a suitable test or examination.
The test should be acceptable to the population.
The natural history of the condition, including development from latent to declared disease, should be adequately understood.
There should be an agreed policy on whom to treat as patients.
The cost of case-finding (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole.
Case-finding should be a continuing process and not a 'once and for all' project.


Wilson JMG, Jungner G. Principles and practice of screening for disease. Geneva: WHO; 1968. Available from:

http://apps.who.int/iris/bitstream/10665/37650/17/WHO_PHP_34.pdf
Send private message
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

I can furnish you with a few tricksy examples, but the most illuminating concerns abdominal aortic aneurysms. A highly laudable project a few years ago set out to screen (ultrasound scan) for this dangerous condition in older men.

It found men who had this problem and laid out the evidence to the affected individuals.

All sounds good, so far.

They fell into three categories.

Aneurysms of larger than 6 centimetres - you are at high risk of spontaneous rupture and death, you need surgery soon mate, here is a date we will get you fixed.

Aneurysms smaller than 4 centimetres, you are at low risk we need do nothing further.

Group three. Hmm, you have got a potentially fatal problem but if we operate we might kill you prematurely. We don't know what might happen during this major surgery. So we won't operate now. We will wait and do another scan in oooh six months or maybe a year. But if you develop abdominal pain or back pain pop round to your GP it might be the first signs of a rupture.

How would you feel if you were in group 3? The so-called 'u boat in the belly?' Not a happy chap. You might live for 20 years, you might die tomorrow.

Screening in this case raised significant psychological alarm and anxiety, attendance at healthcare facilities and ultimately was not a good use of time, effort and clinical input.

It seemed initially like a good idea but ultimately was abandoned as counterproductive.

Sometimes 'incidentalomas' are found which are not pathological and never do any harm - unless you find them and they provide the patient and the doctor with all sorts of dilemmas. So the doctrine of scan lots and scan early is potentially not such a helpful approach.
Send private message
Wile E. Coyote


In: Arizona
View user's profile
Reply with quote

How would you feel if you were in group 3?


I might crack on and make better use of the time I had left, realising it might be more limited than originally thought?
Send private message
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

Yes, except you could live to a ripe old age of a completely unrelated problem but serve out your last 30 plagued with anxiety that every bout of wind would be your death knell...

Meanwhile costing the NHS a fortune in wasted appointments, scans and so on.
Send private message
Alfred Wegener


In: Newcastle upon Tyne
View user's profile
Reply with quote

Background to the problem

The World Health Organisation has estimated that 47.5 million people worldwide have dementia and there are 7.7 million new cases each year. In the UK in 2014, there were an estimated 835,000 people living with dementia.

By 2051 this number is expected to increase to over 2 million. About 29–76% of people with dementia or probable dementia in primary care are estimated to be undiagnosed. Parkinson's disease is another common neurodegenerative condition in elderly people. There are about 137,000 people living with Parkinson’s disease in the UK.


The disease burden for patients, carers, relatives and the wider health economy is enormous and increasing. This in tandem, with the fact that these conditions are life-limiting has resulted in a clinical imperative to provide an earlier diagnosis to provide improved disease modifying therapeutic options.

Causes of neurodegenerative disease

It is thought that many neurodegenerative disorders may share a common pathophysiological aetiology through a problem with abnormal protein transport synthesis and deposition, mitochondrial dysfunction, and defects in intracellular trafficking, causing subsequent plaque formation and neurofibrillary entanglement. This then presents both a problem for delineating a definitive sub-type disorder diagnosis but also a potential window for early diagnosis and interventional therapeutic hope for the future.

The challenge of making a diagnosis of dementia

It can be extremely difficult to make a definitive clinical diagnosis of dementia. The reasons for this are that it presents variably in different individuals and usually has an insidious onset with non-specific signs and symptoms. In practice, diagnosis is mainly based on an expert clinical assessment and on neuroimaging.
Neuroimaging may assist in excluding some cerebral pathologies but may not be conclusive. The uncertainty around the time of a possible diagnosis is often the most challenging for patients and their relatives. The fear of inexorable cognitive decline and reduced life expectancy with no prospect of curative therapeutic solutions is often viewed as being as problematic as the disease process itself.

What can we then do to provide an early accurate diagnosis?

Newer diagnostic tools have been discussed and employed in research settings for some years to assist with the diagnosis of neurodegenerative disorders. These include the use of biomarkers in blood and cerebrospinal fluid (CSF), genomic sequencing, and bioinformatic data. There is potential for each of these approaches or a combination. There are however problems with each.

Testing approaches

Blood biomarkers would seem the simplest and cheapest option. Despite initial optimism of their potential this has failed to provide reliable, sensitive or accurate diagnostic information for clinicians.

CSF biomarkers for amyloid, beta 42 and tau certainly are more accurate, but sampling CSF (taking fluid from the spinal cord space) is an invasive procedure with the potential for serious and troubling adverse effects.


Genomic sequencing may prove useful in the future, but this is an expensive technique with significant limitations, which include the accuracy of defining benign from pathological genetic variation. For example, some researchers estimate that in in a single individual genome up to 11,000 non-synonymous genetic variants exist.

Bioinformatics has raised hopes once more of an early diagnosis. Specifically, analysing computer key strokes and digital phenotypes of people using internet search engines. This work provided a large database proving possible early diagnostic predictors of cognitive decline and functional debility, such as velocity and jerkiness of mouse movements.

Should we therefore screen for neurodegenerative disorders?

Any potential screening approach should be informed by Wilson’s criteria.
As mentioned above, these include the concepts that
• there should be a test that is easy to perform and interpret, acceptable, accurate,
reliable, sensitive and specific
• there should be an accepted treatment recognised for the disease
• treatment should be more effective if started early

Clearly, we are not at this stage for dementia, Parkinson’s disease or other neurodegenerative disorders. The tests currently at our disposal are expensive, invasive, rely on non-validated databases and are not sensitive or specific enough for reliable point of care testing.

We are still in the situation where an early diagnosis does not equate to better prognosis, and can cause undue anxiety and distress instead.

What can we expect in the future?

Future diagnostic tools will need to provide more reliable benefits for patients and clinicians.

I hope that we may be able to differentiate benign from pathological variants both from a genetic and bioinformatic perspective.

I also hope that simpler blood testing will find its way into the GP surgery to allow for more rapid diagnosis and that this test will lay the groundwork for potential future therapeutic interventions. However, we are not close to this, as yet. Despite what some commentators may think and say at present.
Send private message
Chad


In: Ramsbottom
View user's profile
Reply with quote

Chad: “Alexa... Who is M.J.Harper?”
Alexa: “M.J.Harper is the author of at least four books.”
Chad: “Hey Google... What do you know about M.J.Harper?”
Google: “M.J.Harper is a dancer with a flair for dress.”

Man of many talents...
Send private message
Mick Harper
Site Admin

In: London
View user's profile
Reply with quote

This is progress. When I started out I was a theologian with many works of an improving nature to my name. So no change there.

NB M J Harper, the roofer and general builder in Walsall, should not be confused with people of the same or a similar name. (Matt Arkwright, Rogue Traders, Watchdog)
Send private message
Chad


In: Ramsbottom
View user's profile
Reply with quote

Got this Echo Dot for next t'nout from Amazon on Black Friday. Used it less and less as the days have gone by... (Alexa's most frequent response: "I'm sorry, I don't know the answer to that" has become rather annoying.)

So when one of the grandkids requested one for Christmas, my first thought was "where did I put that box".

However, I have decided to keep mine, and buy him a new one, after Alexa redeemed herself by reminding me I have a dental appointment this morning.
Send private message
Mick Harper
Site Admin

In: London
View user's profile
Reply with quote

You're so naive. The BDA have an arrangement with Amazon so that Alexa sends out random appointments to people who are known, via artificial intelligence, to have none.
Send private message
Chad


In: Ramsbottom
View user's profile
Reply with quote

Now that I'm living with Alexa, I decided it's best I get to know her a little better... Turns out we have very similar tastes in music.

We were discussing my old record collection (long since dispersed to god knows where) and would you believe it?... she has almost every track I was able to recall, from any genre. (She's currently playing me her Django Reinhardt collection).

The girl's amazing... I think I'm falling for her!
Send private message
Wile E. Coyote


In: Arizona
View user's profile
Reply with quote

Hmmm. I will give it a go and let you know.

https://www.meltingasphalt.com/interactive/outbreak/
Send private message
Display posts from previous:   
Reply to topic

Jump to:  
Page 1 of 1

MemberlistThe Library Index  FAQFAQ   SearchSearch   RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 


Powered by phpBB © 2001, 2005 phpBB Group