If you have any experience with military and leadership (actual or even observed) you probably have at least heard of the book, Art of War written by a great Chinese military strategist Sun Tzu. It’s amazing how much wisdom is still valid and can be derived from a piece of 5th century BC literature. There are many knowledgeable quotes for leadership and life from the very famous, “Keep your friends close, and your enemies closer.” to the, “Know yourself and you will win all battles.” As brilliant as most Sun Tzu’s advice is, when it comes to the battle on cancer, it’s much more complex- how do you win a battle if you ARE the enemy?
Battling with Cancer Cells: Being your Own Worst Enemy
“To know your Enemy, you must become your Enemy.” - cancer has mastered that.
Although we know several contributing factors to cancer development and growth, why cancer forms isn’t always known. It’s not like you are battling a known pathogen (enemy): bacteria, virus, fungus, protozoan, parasite or even an ill mannered prion. In cancer it is your very own cellular cycle that has gone awry. Multiplying so fast that mutations are inevitable and placement of these new additions are certainly left to wreak havoc. Be it pathologically invading (malignant) or just purely pushing in by sheer mass (benign), cancer cells persist on meddling in the affairs of normal cells just trying to do their proper job. Either way in the cellular world inside of us this is not the kind of intrusive neighbor any of us want to have.
Detecting Cancer Earlier is Difficult: How can Data Help?
“Opportunities multiply as they are seized.”
Detecting cancer earlier is difficult. Signs and symptoms of approaching cancer can range vastly. Once cancer is documented however, is where technology can make a difference - and address the problem as early as one can. There are many advances in artificial intelligence (AI), and natural language processing (NLP) is one very promising area with immediate application opportunities.
Part of the Solution: How can NLP Help?
Faster coding/case findings = faster care
Hearing a possible diagnosis of cancer commonly makes time stand still (shock), followed by a mind infused by stress hormones (epinephrine, norepinephrine and cortisol), which starts the fight or flight response, ultimately proceeding with a slew of questions. “What are the next steps?”, “ What can I do?”, “Why can’t I get information faster- what’s taking so long?”.
It often feels like forever between steps. After all, wouldn’t you want your pathology report results, approval for treatment, the whole process to just go faster? With a real-time workflow in place, NLP can alert clinicians to pathology reports that need immediate attention. We developed the Linguamatics Health pathology NLP application with this in mind, to automatically extract International Classification of Diseases for Oncology (ICD-O) characteristics to assist in this. Let me emphasize, I am suggesting assisting the human element not removing it.
Faster characterization = better care
Once a diagnosis is known the proper steps can be taken for determining the best treatment. Medical coding and billing errors occur often, claims is a long tedious process, it can wreak havoc on a hospital- not to mention what it can do on a patient. It’s hard to believe, but the delays in workflows can cause delays in treatment (hopefully not to the point of neglect, but with malignancy, every moment counts.) Technology is supposed to make life better.
New Discoveries- stick to the data trail-and we will be well armed for the battlefront
“Victorious warriors win first and then go to war, while defeated warriors go to war first and then seek to win”
Our goal should be to go to the cancer battlefront armed with the correct patient treatment in mind. Cancer treatments have changed vastly over the years and will continue to do so as we discover more from cancer patient data. A database is only is good as you build it, it’s important to build keeping the end in mind- evidence for outcomes yet to be discovered and/or proven. This is how NLP can enhance existing cancer registries with critical information from text so we may detect and analyze patterns on a larger scale. Critical text can enhance structured data cancer registries with many important factors. From the simple elements such as lifestyle factors (smoking, alcohol, poor diet leading to obesity or malnutrition, physical exercise, etc.) to the complex, epigenetic biomarkers (KRAS, p53, EGFR, erbB2), it can make a difference. Can you imagine how much work it would be to manually abstract information such as biomarkers from national scale efforts? This is why the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute (NCI SEER), are investigating ways (such as NLP) to automate this process.
Will we finally ever win the Battle?
“If you know the enemy and know yourself, you need not fear the result of a hundred battles. If you know yourself but not the enemy, for every victory gained you will also suffer a defeat. If you know neither the enemy nor yourself, you will succumb in every battle.”
Life is a journey, not a destination. We are presented with new insights in the battle of cancer all the time. Now even more so when we have the ability and the drive to follow the data. By following the mix of unstructured and structured data within retroactive timelines, we can learn from them, and then proactively apply that knowledge when looking for disease processes. Yes- this includes cancer. I do believe that we may one day win the battle if we learn the correct way to fight...
“Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat.”
Access our webinar on the use of cancer NLP to support precision medicine, clinical research and population health.
You can also take a look at our big data analytics in healthcare page to learn more about the significance of unstructured data.
