The way forward for big data analysis in ophthalmology and medicine

It was the best of times; it was the worst of times. It was the season of population medicine; it was the season of personalized medicine—or was it? Charles Dickens’ fictional depiction of late eighteenth century France walked a tense line that masterfully narrated the untenable socioeconomic disparity of its day. He described the forces that inevitably led to the French revolution. Today in the world of medicine we are on the verge of similar untenable tension. Unravelling of the human genome was indeed an epochal event that brought with it the promise of personalized medicine. Thirteen years later, this is a promise that is certainly yet to be fulfilled. We are unable to bear the massive weight of the unmet promise of personalized medicine. It is therefore little surprise that any glimmers of its potential realization get great attention and scrutiny. One such glimmer is the work of Dr. J. William Harbour and colleagues in which they have correlated gene expression profiles with prognosis in ocular melanoma (1). They classified (2) tumor gene expression profiles into two classes, those associated with a higher likelihood of metastasis and those associated with a lower likelihood of metastasis. Their clinically-relevant result is an example that rekindles hope that our massive investment in unravelling the human genome may yet yield a return. Our team at Quantum Lucid Research Laboratories consists of ophthalmologists, mathematicians, computer scientists, physicists, and engineers, and has a unique appreciation for how such work bridges ostensibly disparate worlds. Data from the human genome project is an example of big data. Other examples of big data include the troves of imaging and clinical laboratory data which we are accumulating in our health centers and other institutions. It is vital that such big data not lay dormant, but instead be translated into better means for diagnosing and treating disease.