There’s a stone in your gall bladder and two in your left kidney

I was just reading an interesting article on the World Economic Forum website which speaks about seven interesting ways in which AI is being used today in healthcare.

Imaging seems to be a core application area and it’s not just about radiology.  Skin cancer diagnosis for example. That’s an imaging and recognition problem.

Eye scans is another imaging application. It appears that the two areas where AI is already doing very well is in the diagnosis of diabetic retinopathy and age-related macular degeneration (AMD).

I found the reference to brain scans and the corresponding imaging application very intriguing. The AI looks at the brain scan and is able to make a better prediction than a human doctor on the probability that a person in a coma will recover.

The radiology and cardiology applications. As also the interpretation of CT scans are more straightforward and intuitive. Here is a link to the article –

I am looking forward to big leaps in diagnosis via imaging at a microscopic level and spectrometry. How great it would be to diagnose common infections or micro nutrient deficiencies or most of the blood work that a lab does today; by simply shining a light on a drop of blood or maybe through the skin or pointing a camera into the eye. The machine interprets the image and gives you an almost instant diagnosis. Not just of an existing disease but also the likelihood of getting one?

The nature of imaging and interpretation is changing too. Soon we will not have the need to get into big machines for images from inside our bodies. Imaging could be a continuous affair using devices we can wear. This is already true for ECG. If the work of people like Mary Lou Jepsen becomes real, that’s what we will have. And surprisingly she uses sound to look into into the body. There’s an article about her work here – link –

This will mean that diagnosis is not a matter of looking at an image and comparing it with a knowledge base of images. It will be looking at a series of images. Hundreds or thousands or millions of them to identify patterns that point at something amiss. This is the amount of data that a human being will never be able to handle. An explosion in data will necessarily require machines to interpret the data. That’s the future of imaging and AI in medicine.

The coming deluge of data in the medical world

A few decades ago, a physician or a pathologist had to deal with a much smaller knowledge base than modern medicine today.

As medical knowledge grew in size and complexity, we have seen the emergence of various specialities. And then naturally, superspecialists.

In the coming decade, we shall see an explosion of data. This is just round the corner in the biotech and medical world.

It will not be possible for human specialists to deal with this deluge of data or to make sense of it beyond a point.

It will become necessary for doctors and pathologists to work with AI machines which summarise and interpret this data for human specialists.

It will not be possible for doctors to function effectively without extensive use of artificial intelligence frameworks that help the doctor to diagnose or to recommend a course of action.

Here is an article that summarises the emerging situation well from a pathology viewpoint.



Why must we be metal and silicon?

There are many ways humans are trying to extend life. One influential school of thought is the idea that we will have the ability to replace damaged organs or extend the capabilities of our organs.

The popular imagination and science fiction has dealt with this and fantasised about this for many decades now. That humans will become cyborgs.

But that imagination has almost always been about metal and silicon. Or at least the core of the cyborg was metal and silicon. Look at Arnie in The Terminator – metal covered with flesh but not the other way around.

There was no flesh cased in a protective inorganic casing. Those were usually the evil aliens and not cyborgs or augmented humans. Our cultural biases were speaking – anything biological and as intelligent as the human is an evil alien.

What if the biggest advances are going to be in regeneration of organs. Whether it is within our bodies or grown elsewhere and transplanted into our bodies? Not just regeneration of hearts, kidneys, livers, bone and joints; but the brain too.

It appears to me that in the coming couple of decades, genetic engineering may make much bigger leaps than synthetic biology.  That the future is biological advancement of the human species and not fusion with inorganic machines.

We live in interesting times.

Here is the Ted talk that set off this thinking. Meet Dr Luhan Yang and Lika the cute pig –

Healthcare – China spends 10x and the developed world 100x of India

It is a pity that India spends so little on healthcare for its citizens.

For a country that ranks a low 130 on HDI but with superpower ambitions, it is a shame that we can’t spend enough to improve the healthcare available to our people.

The central government health budget is about ₹50K crores and the states three times that. The total healthcare spend by the centre and states put together is ₹200K crores, which is about 28 billion US dollars.  This translates to 21 US dollars per capita.

Another telling statistic is that most of the healthcare expense in India is in the private sector. Government spend of $21 is only 30 percent of the spend. Both rural and urban India depend primarily on the private sector for medical care.

As a comparison, spend by government in the US on healthcare is over 100 times more, at 2500 US dollars per capita. While total healthcare spends in the US are actually higher than 8000 dollars per capita, the share of government spend is about 2500 dollars.

In the UK, while total spend on healthcare per capita is lower than the US, the government spend on NHS is even higher than the US at above 3000 dollars per capita.

The Chinese government too, spends ten times as much as India does on healthcare per capita.

I am simply unable to understand India’s poor focus on healthcare.

True, the current government has come up with an ambitious healthcare policy but the proof of the pudding is in the budgets. What is the point of announcing a policy that needs ₹800K crores of funding but only ₹200K crores has been budgeted? Where is the balance money going to come from?

Is this then a policy or just a statement of desire?

I think it is clear and everybody understands now that our real capital is our human capital. But without prioritising health and education, the human capital will not be of good quality. Then of what use will that human capital be? Surely not capable of the productivity gains we need, if we are to become a middle income country in the next 15 years.

The biggest Indian natural disaster ever

India has faced many severe and huge natural disasters in the last 5 centuries.

The largest ones according to public memory and our history texts are the famines.

For example, the Bengal famine of 1943 was responsible for about 3 million deaths.

But what most people do not know and we are not taught in school is that the single biggest disaster was the 1918 flu.

The flu killed 5 percent of India’s population within a few months. About 18 million people.

A similar pandemic today with the same mortality rate would kill 70 million people.

The flu was caused by an H1N1 flu strain. Another strain of H1N1 causes what we all know now as the Swine Flu.

There are two theories about what caused such high mortality in the 1918 flu.

One is that it caused an extreme reaction of the immune system in young adults. If this was the case, a similar pandemic could cause equivalent rates of mortality today.

The second is that most of the deaths in 1918 happened as a result of a bacterial superinfection that followed the flu. The first antibiotics were not available until the 1930s! A similar occurence today would see most of those bacterial infections treated successfully.

It is surprising that the flu was not so devastating in the South and East of the country. Was it because of the high levels of coconut oil and coconut in the food? Monolaurin (Lauric Acid) in coconut oil is known to have antiviral properties. Did this protect them? Or was it something else?




Artificial Intelligence and cancer spotting

Going by first principles, any medical diagnosis is the interpretation of data based on known correlations. What disease or condition does that data correlate to?

The data could include stated symptoms, visual/touch/sound observations, ECG readings, ultrasound imaging, MRIs, X-Rays, CAT scans, blood reports, radiology reports, biopsy observations, reactions/cultures and more.

The correlations may not be simple or direct. Often the interpretation could simply throw up the need for more data. A different test perhaps.

Quite a few of these diagnostics involve the interpretation of images. This could be an MRI report, an Ultrasound test, an X-Ray, a photograph…

There are reports that despite spending billions of dollars, IBM Watson has not lived up to its promise in cancer detection.

However, I will argue that irrespective of whether some AI experiment succeeds or fails, machines are better than human experts already in interpreting images for applications like tumor detection. The only caveat is that this statement will be true only in those cases where there is a large enough existing dataset to learn from.

I am not saying that doctors will get replaced. On the contrary, doctors will have tools with diagnostic capabilities that have not existed so far.

People will want a human doctor to exercise independent judgement and accept or reject what a machine finds. People will want a doctor to explain the diagnosis to them, to recommend a course of action and to tell them why such an action is being recommended.

But doctors must be prepared for a world that is coming very soon, where diagnostic tests and their interpretation will by machines. They must adopt and embrace these technologies sooner rather than later. These are the tools that will enable them to greatly enhance outcomes for their patients.

here is an article on a technology that detects lung cancer tumors with 95 percent accuracy, when human experts are at 65 percent-

Extinction of a Species – By Design

I am not sure if we should be troubled about the work being done to eliminate mosquito species that are vectors for diseases.

This is no longer science fiction. There are multiple initiatives which follow variants of a basic approach : “Make males whose offspring will only be males. Then release a few of these engineered males into the wild”

This approach, as some experiments have shown, have the potential to rapidly eliminate the species within a few generations in that area. This can be a very short period. Less than a couple of years.

There is no way that this can be isolated to a particular area. Any such initiative will soon spread worldwide and make the species extinct.

These “blood suckers” which cause malaria, dengue, chikungunya are not just disease carriers. They are also food for other animals. They are a part of the food chain. 

Only the females feed on blood. Both male and female mosquitoes primarily feed on nectar. The same thing that honey bees feed on. They also perform the same function that bees do – they pollinate the flowers (Some have argued that mosquitoes are not the great pollinators they are suspected to be).

We are entering an era of genetically engineered plant and animal species of all kinds. These mosquitoes are merely the beginning. Many of these species will get into the wild without any controls because the species are created by individuals or small groups of people in DIY labs.

Somehow, the thought of man made genetic solutions that eliminate or replace an entire existing species troubles me.

What about you?

Here is the mosquito article –

Criminal or Needs Help? How Do We Know?

Someone told me the other day that the reason why we send people to jail is not to punish them. That’s not the reason at all.

The reason is protection. And it works both ways. Protecting society from criminals and protecting criminals from vigilante justice. One keeps society safe and the second keeps it sane.

That brings me to the topic of mental illness. It is possible to argue in the following manner :  Almost all criminal behaviour is deviant from what we consider “normal”.  Criminal behaviour born of thought patterns that are deviant from the normal is what we may call mental illness. And so a person who has exhibited criminal behaviour is to be treated as a victim of mental illness.

In the last century, we have recognised and accepted some forms of deviant behaviour as born of mental illnesses and recognised that some others (like homosexuality) are not illnesses at all and are perfectly normal. 

But many other behaviours are yet to be recognised and formally classified as mental illness by the criminal justice system. Even perfectly normal behaviours like homosexuality continue to be criminalized in less developed legal systems and unfortunately in India as well.

Just as in other areas of life, I expect technology to play a major role in the coming two decades in this area. So far, psychiatry has not had the technology to look at things like genetic profiles or body chemistry at hormonal and other levels. The causal relationships or correlation between these and mental illnesses is neither visible nor well understood. I think that very strong relationships most likely exist and they will be found.

This opens up the possibility of cures or methods of containment. 

Maybe in future, we will be able to treat these people as victims and help them to reverse the condition or contain it. Maybe many of the people who go to jail today, don’t really belong there.

Alok Sarin makes an interesting case here on an associated topic –


Portable Medical Imaging and Communicating Thought

I was reading this extraordinary set of claims for a new technology at this link –

Coming from Mary Lou Jepsen, you have no option but to take the claims seriously.

Imagine looking into your body at resolutions that are even better than MRIs or CAT Scans. Imagine being able to do that over extended periods of time and not just looking at organs but also movements and flows! And all this as a wearable and not while lying inside a huge machine in a hospital.

The telepathy piece is something I have difficulty believing in. I think the jump from imaging, to interpretation and then communication of complex thought patterns is not feasible using any current technology.

Speech to text is complex enough. Interpretation of the nuances of language is still a work in progress.

Here we are talking of interpreting thought, or rather the visual expression of it as seen on an imaging machine. I suppose that will work  for basic commands that you would use for playing a video game for example. It may also work for communicating simple phrases or feelings or direction etc.

I can’t see how complex, structured (or monkey brain) thinking can be interpreted or communicated.

Maybe I am reading this wrong. I am assuming that a machine is interpreting and then communicating the interpreted thought patterns to humans. On the contrary, here it is humans who interpret the raw expression of thought?  And this is easy for us because it is not much different from what we “see” in our own heads?

I would so love to be a part of this team! This work seems to be really exciting.