IA is a tool that can complement Medicine (and other areas) the idea is to generate better diagnosis of illneses.

Dear,

Deep learning can help to avoid human mistakes and forgot, through medical image classification or some cases early detection of diseases or anomaly detection in measures or in some images.

A.I. can provide more time for medical doctors to dedicate to patients, it can provide them confidence in the diagnosis so that he will propose the best approach to the patient's treatment and it can even provide assessment on the emotional state of the patient thus providing indications on the best way to communicate and how to communicate with the patient for the benefit of his/her mental state. So in summary yes it can humanize medicine as long as the role of the doctor stands as the responsible, above all, for the interaction with the patient.

Yeah, till date, AI is just being implemented as tool in medicine domain for providing outperformed results of diagnostic tolerance levels than traditional methods. However, in near future, the AI tools like Reinforcement learning would pave the way for human like control in deciding the medicinal parameters for any disease.

Medicine has as many dimensions as have human beings. Does AI have any role in this clinical scenario (NEJM Catalyst)--must read item for scientists and practitioners of medicine.

Stone-Age Policies Stifle Modern Solutions

Article · April 3, 2019

Susan Dentzer

Robert J. Margolis Center for Health Policy, Duke University

America’s health care providers routinely practice Star Wars medicine, but on a Flintstones delivery platform, as a current saying goes. The U.S. health system needs to advance, at least, to The Jetsons futuristic vision as depicted in the 1960s, when one of the first U.S. telemedicine programs actually came into existence in Boston. More than 50 years later, it’s only now that telemedicine and other forms of virtual care are becoming more widespread. Although the evidence suggests that patients could benefit if it spread even faster, we still need to make light-years-long leaps over multiple barriers to get there.

To understand the possibilities, imagine all the people who might have trouble getting to a doctor’s office for care. Patients with Parkinson’s surely would be near the top of the list. They frequently have issues with gait and mobility; they may be unable to drive. Now, situate these patients in rural areas of the country, where they may need to travel long distances to the nearest medical facility. If patients have caregivers, the burden will probably fall on them. Small wonder that more than 40% of Medicare beneficiaries who would almost certainly benefit from visits to neurological specialists don’t get them...

https://catalyst.nejm.org/stone-age-policies-stifle-virtual-care/

With 60-70% of patients without higher education, the American (USA) medical system expects all patients to take important health decisions on their own -- fancifully termed as preferences -- with support from guidelines, such as initiation and frequency of cycstoscopy for aymptomatic microscopic hematuria.

Is this for real?? If so, how will AI help?

Bladder Cancer in Patients with Asymptomatic Microscopic Hematuria

Allan S. Brett, MD reviewing Gonzalez AN et al. Urology 2019 April

In a large study, no patient younger than 50 with asymptomatic microscopic hematuria had bladder cancer at cystoscopy.

When should a patient with asymptomatic microscopic hematuria be referred for cystoscopy to screen for bladder cancer? To address this question, urologists at Columbia University reviewed records of more than 2000 patients evaluated for asymptomatic microscopic hematuria (defined as ≥3 red cells per high-power field on microscopy). Patients with recent histories of visible hematuria, medical renal disease, infection, or stones were excluded.

Twenty-five patients (1.2%) had bladder cancer at cystoscopy; none were muscle-invasive. Rates were 0.8% in women and 1.9% in men. The youngest man and woman with bladder cancer were 52 and 56, respectively; none of 444 patients who were younger than 50 had bladder cancer. On multivariable analysis, the only predictors of cancer were age and smoking history. Probabilities ranged from 0.3% for never-smokers in their 50s to 12.5% for current smokers in their 80s. Number of red cells per high-power field did not predict bladder cancer.

Comment

These findings, which are similar to those of other observational studies, suggest that bladder cancer is extremely rare in patients younger than 50 with asymptomatic microscopic hematuria (and no history of visible hematuria). However, the American Urology Association (AUA) guideline on asymptomatic microhematuria recommends cystoscopy starting at age 35. In my view, a reasonable compromise (for patients younger than 50) would be to acknowledge the AUA's age cutoff, discuss the very low probability of bladder cancer in this age group, and allow informed patient preferences to guide referral decisions.

Can AI help in this situation?

A must read for all scientists, especially medical scientists...(see attached file)

There is a chance that AI will become gradually our guardians, which will be assessing our health condition permanently on the from-minute-to-minute basis.

We currently see the onset of devices capable to detect atrial fibrillation, blood sugar levels, blood pressure, etc. in real-time. The spectrum of measurable biosignals is gradually increasing.

Those devices using AI can even predict events like arrhythmias, epileptic seizures, diabetes coma, blood pressure decreases/increases, etc. My own research focusses on the prediction of heart arrhythmias.

Those AI wearable devices will definitely humanize medicine as they replace the constant supervision of the patient by a nurse/specialist. AI devices work tirelessly 24/7 without sleep and rest. :-)

Patients who will own such AI devices will be more self-confident to do activities that are not suitable or too risky for them now.

There might be devices predicting remaining reserves of a diseased organ that will inform the wearer about the necessity to take rest, replenish energy, eat, or even call an ambulance.

Patients suffering from heart diseases and brain epilepsy with a chance of losing consciousness will be much safer as AI devices can call help and ambulance. It will add a lot to patients' self-confidence and quality of life.

We are definitely working towards more humane medicine that will be heavily relying on AI, complex systems, and other latest development of mathematical research in biomedicine.