Archive

When in doubt: How machine learning and AI are helping diagnose cancer better

Multi-disciplinary teams, joint tumour boards, and machine learning give peace of mind for cancer patients.

Written by : TNM

We are living in the age of artificial intelligence and self-learning computers. Virtual reality has become a part of lives and instant connection to friends and strangers around the world is now possible. However, an age-old problem still haunts us. Doubt. Is this the right decision I am taking? What if things aren’t what they seem?

Doubt is a huge risk factor for patients, especially patients with severe ailments like cancer. While medical science - especially in cancer care - has advanced so much in the last decade that we are now approaching a point where one can diagnose cancer with a drop of blood and get a simple tablet to cure it, patients and their loved ones still need reassurance.

This is especially the case because till recently, many treatment options for cancer significantly affected the quality of life for the patient, and one wrong decision could be extremely dangerous.

Medicine and health care - especially doctors who approach their work with commitment to serve people and better lives - have always stressed on getting a second opinion. And that’s also good, sensible practice. Problems can be solved in many ways, and a specialist may see one course of action while another may see a different way to treat the same disease, and if the life of a patient improves with one or the other, that is the ultimate reward for medical professionals.

At Apollo Cancer Hospitals, Senior Medical Oncologist Dr. T Raja knows this only too well. He knows what cancer can do to a patient’s life and knows how to cure it. He also knows fully well that patients and their loved ones need all the reassurance they can get.

Which is why, Apollo Hospitals and Dr. T Raja have instituted special protocols - global best practices and have enlisted the help of the most cutting-edge technology to give patients a second opinion, and a much more reasoned, thought-out course of medical care.

Machine Learning and Artificial Intelligence

Machine Learning is a branch of computer science which creates a “sentient” system. That is, a computer can record, store, and read truly staggering amounts of data across many parameters, and seek out patterns. Over a period of time, the computer becomes adept at pattern-seeking it can also “predict” what the data will be given only one or two parameters.

The computer can “learn” from past information and apply that learning, to a future problem to arrive at a solution much quicker. And the more data and more patterns it sees, the better its ability to “solve”.

According to a paper in Pubmed, “Machine learning is a branch of artificial intelligence that employs a variety of statistical, probabilistic and optimization techniques that allows computers to “learn” from past examples and to detect hard-to- discern patterns from large, noisy or complex data sets. This capability is particularly well-suited to medical applications, especially those that depend on complex proteomic and genomic measurements. As a result, machine learning is frequently used in cancer diagnosis and detection. More recently machine learning has been applied to cancer prognosis and prediction.”

This makes machine learning artificial intelligence a highly useful tool for cancer prognosis.

As Dr. Raja says, “Each area of cancer is becoming so deeply researched, well researched. We know that, in the world of computing, Artificial intelligence is coming into it a big way. If we can apply that same idea for medical knowledge and present the essence of current data and ever-growing knowledge, to be relevant to the precise need of the patient, we can achieve a good deal more. That’s where machines come in.”

Follow Dr. T Raja on Twitter 

This is not entirely new. Computer-aided diagnosis and computer aided medical decision making has been used in cancer care for at least 20 years. The same pubmed paper says, “Artificial neural networks (ANNs) and decision trees (DTs) have been used in cancer detection and diagnosis for nearly 20 years.”

What does this mean for the cancer patient?

Every symptom, every manifestation of the disease, every blood test and other tests, X-Rays, MRIs, Scans, are entered into a large database. The database collects and stores, and classifies this data, across a huge number of patients, not just at Apollo Cancer hospital in Chennai, but every Apollo centre, and every other cancer hospital around the world.

Further, past research, current research on cancer, medical and scientific texts, journals, publications, websites, are all also part of this database.

This database is constantly updated, and as it records and stores, the computer also sees patterns emerging. For instance, it notes that all patients with more-than-normal estrogen levels in the blood are at higher risk of developing breast cancer. It correlates this information with other data, such as risk profiles for women in India, global disease burden for cancer, and so on, and is able to arrive at a decision.

So, the next time a patient consults Dr. T. Raja, this data is available to him, as well as to the patient. This allows both doctor and patient to arrive at a much more informed, well-reasoned, well-judged decision on what the next steps are to be.

Dr. T Raja says, “Medical field is moving. Cancer care is moving in that direction. We need artificial intelligence, already now, but very very soon in the near future, to help us decide, in the best possible way, in what we call evidence-based medicine, or evidence based clinical decision making.”

But Dr. T Raja gives a word of warning. He says aritificial intelligence and machine learning cannot be used as a diagnostic tool. It can be used as an aid for the doctors to keep up to date with the changes in the world of cancer, and to catch potential symptoms and effects of the disease, but it cannot replace a doctor’s valuable instincts and diagnosis.

“This helps in decision making! You can’t stand in front of a machine and let the machine make a diagnosis,” he says. And that is where, an Oncologists specialised knowledge, and years of experience comes into play. At Apollo Cancer Hospitals, this is combined with breakthrough cancer treatments, global best practices, and a trust - above all - that every patient deserves the best.

Multi-Disciplinary Teams & Joint Tumour Board Practice: leaving nothing to chance

One of the global best practices that Dr. Raja and other oncologists at Apollo follow, is the Joint Tumour Board and Multi-Disciplinary Team. This is another innovation that is aimed at improving patient outcomes at Apollo, and allows for a more nuanced, reasoned decision making for doctors.

An English phrase “When you have a hammer, all problems look like nails”, talks of the unique situation a specialist faces. While a specialist has arrived at his or her niche field with a lot of hard work, extensive knowledge, and wide experience, they may often only see the more difficult course of action for even a routine, simple problem.

The Multi-Disciplinary Team and Joint Tumour Board system solves that situation.

At Apollo, for a majority of cases, a team of doctors with multiple specialisations, will jointly discuss treatment outcomes and arrive at a consensus. This means that a patient can breathe easy knowing that no options are overlooked, and nothing is left to chance.

“Instead on one person working alone, you assemble a few experts on the same team, who can work together,” says Dr. Raja.

When a patient comes to Dr. Raja, relevant information, tests, and other data is reviewed by a multi-disciplinary team comprising, say a surgeon and a radiologist. This team will then discuss various treatment methods - a radiation expert may suggest a course of radiation. The team decides what is best for each given patients. 

The relative merits and demerits of each course of medicine will be discussed within the group, and a consensual decision may then be communicated to the patient accordingly.

This will mean that the patient not only gets the best medical care, but also that their quality of life is not adversely affected, and further, there could be potential savings of time, effort, and money.

“You form a team and decide what will be good for this particular patient,” says Dr. Raja. “Each patient there may be different decision that all of you come together, and come at a joint decision.”

This is known as a Multi-Disciplinary Team. Studies around the world have shown significant improvement in diagnosis for patients.

MDT approach has also resulted in sometimes changed diagnosis, but has often confirmed a medical oncologist’s first diagnosis, thereby ensuring that the patients arrive at informed, judged consent for any and all procedures.

As Dr. Raja puts it, “In Apollo, for most patients MDT takes the relevant decision.”

Dr. Raja further adds that this is almost a habit for all doctors at Apollo, and an everyday practice for almost all patients.

A further iteration of the MDT, is the Joint Tumour Board. At Apollo, all cancer care experts - oncologists, radiologists, technical experts, and other doctors, meet at least once a month to review all the cases and discuss and debate the medical care prescribed. Further, experts from across the Apollo Network in Chennai, Bangalore, Delhi, Mumbai, Kolkata, and other centres, especially oncologists, meet every month at a common centre to discuss the best medical care for their respective patients.

“A few brains thinking and deciding together, is always better than one,” says Dr. Raja.

For Dr. Raja, and other Oncologists at Apollo Cancer Hospitals, Machine Learning is a promising future where the best of computer science complements the very best of medical knowledge and experience. This, together with MDT and Joint Tumour Boards mean that Dr. Raja is able to offer his patients the best promise: accurate diagnosis, expert care, and peace of mind.

This article has been produced by TNM Marquee in association with Apollo Hospitals.

Gautam Adani met YS Jagan in 2021, promised bribe of $200 million, says SEC

Karnataka Congress wins all three bye-polls, Nikhil Kumaraswamy loses again

Palakkad bye-poll: Congress’ Rahul Mamkootathil wins with historic margin of 18,840

LDF retains Chelakkara as CPI(M)s UR Pradeep registers emphatic win

Narayana Murthy is wrong: Indians are working too long and hard already | LME EP 50