Big data has been a game-changer in business, research, technological innovation, and even policymaking. Big data analytics has transformed all industries in the last few years. All the major organizations are implementing a Big Data strategy. However, there is still a large part that has no clue about Big Data. In this article, we will look at the applications and impact of big data analytics in healthcare and medicine. Additionally, we will also look at the list of top universities offering Masters in Health Data Analytics.
Big Data Analytics in Healthcare and Medicine
Currently, the healthcare system is facing several challenges. There is an increasing gap between healthcare costs and the treatment outcomes. The pharmaceutical sector is still fighting against a limited understanding of the biology of diseases. According to the Nature Journal, the top 10 highest grossing drugs prescribed in the US help only a small percentage (< 25%) of the patients. For cholesterol drugs, the success rate is even lower (only 2% of patients). So the probability of success is very lower compared to the expenditure made on research, approval, and marketing activities.
Big data, advanced & predictive analytics (big data analytics), and other new-age technologies (machine learning, artificial intelligence, and cloud technology) do hold the key to solve the current problems. If you are not too familiar with these terms, feel free to read the beginners guide on big data, machine learning, deep learning, artificial intelligence and other new-age technologies.
From its role in new wearable technology to epidemics and research, big data analytics is having a tremendous impact on the medical field. Big data can help in aggregating information around multiple scales for what constitutes a disease—from the DNA, proteins, and metabolites to cells, tissues, organs, organisms, and ecosystems. Additionally, big data can be applied to early diagnosis, predictive modeling clinical decision support, disease or safety surveillance, drug discovery, public health, and biomedical research. In this post, we will look at the major seven applications of big data analytics in healthcare and medicine.
Big Data Analytics in Diagnosis (Predictive Analytics)
At present, we go for check-ups only when there are serious symptoms. Even physicians or diagnostic labs can detect some abnormality only when lipid levels or glucose are way off the charts or something extreme.
Hospitals, clinics, and other healthcare establishments collect patient data such as patient data as name, age, disease description, diabetic profile, medical reports and family history of illnesses etc. Such data do not necessarily provide a true picture of the patient’s medical condition.
For instance, in case of a patient diagnosed with cardiovascular disorders, the traditional healthcare system would collect family history, diet, symptoms, age and other existing diseases. While such information provides a detailed view of the disease, the data is unable to provide other perspectives into the problem.
If a patient has suffered a stroke, the hospital could collect additional data points like the time of stroke, gap between strokes in case of multiple strokes in the past, influencing events preceding the stroke such as a psychologically stressful event or heavy physical activities.
Aggregation of all such data points would help the physicians and healthcare personnel to take definite steps to prevent strokes in the near future
Big Data Analytics in Personalized Medicine
The objective of big data analytics in medicine & healthcare is to understand as much about a patient as possible and as early in their life as possible. Picking up warning signs of serious illness at an early enough stage would make the treatment far simpler (and less expensive) than if it had not been diagnosed until later.
Personalized medicine involves tailoring medicines to a person’s unique genetic makeup – and is developed by integrating a person’s genetic blueprint and data on their lifestyle and environment, then comparing it alongside thousands of others to predict illness and determine the best treatment.
In this particular scenario, wearable devices and various mobile apps are the future. Wearable devices and engagement through mobile apps can provide a way to longitudinally monitor the actual state with respect to many different dimensions of your health. Big data has got the capability to provide a much better and much more accurate profile. Any slight deviations from the baseline could predict a disease state or slide into a disease state.
When the biomedical devices (or apps) can measure how far you walk in a day, track calorie intake, and monitor blood glucose levels on a regular basis.
In the future, this data points can be shared with the doctor, who will use it as part of his or her diagnostic toolbox when an individual visit them with a sickness or some serious symptoms.
Even if there’s nothing wrong with the individual, access to huge, ever-growing databases of information about the state of the health of the general public will allow problems to be spotted before they occur, and remedies – either medicinal or educational – to be prepared in advance.
For example, if patient’s blood pressure increases alarmingly, the system will send an alert in real time to the doctor who will then take action to reach the patient and administer measures to lower the pressure. This part also comes under telemedicine (described later).
This kind of modeling would not be possible unless the individuals get phenotyped on a longitudinal and long-term basis. To leverage the big data analytics, data of an individual must be taken from various sources such as medical and insurance records, wearable sensors, genetic data and even social media use. Then the data can be used to draw a comprehensive picture of the patient as an individual, in order to offer a tailored healthcare package.
This individual’s data won’t be analyzed in isolation. It will be compared and analyzed alongside thousands of others, highlighting specific threats and issues through patterns that emerge during the comparison.
This will enable the doctors and data scientists to practice predictive analytics. In simple terms, a doctor will be able to assess the disease progression and come up with the best treatment regimen on the basis of data from other patients with the same condition, genetic factors, and lifestyle.
Big Data Analytics in Telemedicine
I hope you have heard about the Internet of Things (IoT). The commercially available IoT devices deal with smart home items (e.g. smart TV, smart thermostat, smart refrigerator, AC etc.). The Internet of Medical Things aka IoMT (Medical version of IoT) deal with remote patient monitoring, which is one of the most critical components of telemedicine.
A major goal of telemedicine is to eliminate unnecessary traveling of patients. Telemedicine relies on data acquisition, data storage, data display and processing, and data transfer. The communication between the doctor and a patient (or healthy individual) happens remotely via transfer of text, reports, voice, app notifications, images and video. Read more about big data analytics and telemedicine.
Big Data Analytics in Bioinformatics
The new generation of sequencing technologies enables the processing of billions of DNA sequence daily. The increasing application of electronic health records (EHRs) is documenting large amounts of patient data as well. The storing, managing, and analyzing a massive amount of datasets is a big challenge for the biomedical scientists.
In such cases, artificial intelligence (AI), along with cloud computing, Hadoop, and other data mining tools can be more than handy. Big data can provide data repositories, computing infrastructure, and efficient data manipulation tools for investigators to gather and analyze biological information.
According to IBM, medical images currently account for at least 90% of all medical data. This becomes an overwhelming amount on a human scale when you consider that radiologists in some hospital emergency rooms are presented with thousands of images daily.
Big Data Analytics in Imaging Informatics
Imaging informatics is the study of methods for generating, managing and representing imaging information in various biomedical applications. It deals with how medical images are exchanged and analyzed throughout complex health-care systems.
Advances in both imaging and computers have synergistically led to a rapid rise in the potential use of artificial intelligence in various radiological imaging tasks, such as risk assessment, detection, diagnosis, prognosis, and therapy response, as well as in multi-omics disease discovery.
Machine learning and deep learning can deal with tons of images with superior accuracy to detect tumors & monitor tumor development, and many other diseases. The ultimate benefits are generating predictive image-based phenotypes of disease for personalized medicine. Additionally, it would also yield quantitative image-based phenotypes for data mining with other omics for discovery (ie, imaging genomics). Read more about applications of big data and machine learning in clinical medicine.
Big Data Analytics in Drug Discovery
The drug discovery process has always been a challenge for the biomedical and pharmaceutical industry. It takes approximately 13 years and more than $1 billion to successfully launch a safe and effective drug.
Pre-Clinical Studies: Drug Target Selection & Lead Identification
The drug discovery process starts with the classification and understanding of disease processes, followed by target identification and lead compound discovery. Nowadays, disease classification in drug discovery is moving from a symptom‐based disease classification system to a system of precision medicine based on molecular states.
Big data is used to select drug targets for preclinical studies. The process starts with the identification of molecular changes between disease samples and healthy samples. The molecular changes are implicated in gene expression change, genetic variation, or other features, and are furthermore used to inform target discovery.
A particular tumor type that possesses a genomic alteration could be treated by a drug that targets this alteration, even though this drug was not originally discovered for this tumor type. For instance, KIT was discovered as a target for chronic myelogenous leukemia and later it was discovered as a target in gastrointestinal stromal tumors, leading to the repositioning of the KIT inhibitor, Imatinib, for treating patients with KIT‐positive gastrointestinal stromal tumors.
In the recent past, data-sharing arrangements between the pharmaceutical giants have led to breakthroughs such as the discovery that desipramine, commonly used as an anti-depressant, has potential uses in curing types of lung cancer.
Identifying a set of drug candidates from a library of thousands (or even millions) for testing against a specific disease would be similar to finding the needle in the haystack. Hence, implementing big data analytics to screen large databases containing biological, chemical and clinical information can help whittle down the probable for testing. Moreover, advanced analytics can help analyze the clinical imperative and candidate profiles to create a product pipeline for the pharmaceutical companies. Learn more about leveraging big data to transform target selection and drug discovery.
Clinical trials involve testing of the candidate drugs in large populations with optimum diversity and across multiple study sites (locations). Activities include study design, patient (and healthy volunteer) enrolment, data collection, data analysis & interpretation. So, it’s an extremely tedious task. Conventional data management and analytical tools would not be sufficient to scale up studies and to make sense of diverse streams of data. Then there will be bias and human errors as well.
Big data and cloud platforms allow clinicians and authorities to include terabytes of unstructured data from various real-world data sources (EMRs, genetic profiles, phenotypic data, biomedical devices, mobile apps etc.) along with the data from the clinical trials. Big data analytics has got the capability to take care of such huge amount of structured and unstructured data.
“For instance, the National Cancer Institute (NCI) had set up a prototype project, to gain more insights into the relationship between genes and cancer. NCI was able to search a 4.5 million cell matrix in 28 seconds. In this search, NCI cross-referenced the relationships between 15,000 genes and five major cancer types, across 20 million medical publication abstracts. It also cross-referenced genes from 60 million patients. This helped NCI to gain a deeper understanding of the network of gene-cancer interactions and the state of research in relation to cohort groups treated.” – Excerpt from the Q&A with James Streeter, Global Vice President, Life Sciences Strategy, for Oracle Health Sciences Business Unit.
Drug Discovery for Personalized or Precision Medicine
A drug candidate might not exhibit statistical significance to be effective in a particular clinical trial. But, the drug candidate could be extremely effective in particular subpopulations in terms of genetics, demographics, secondary disease phenotype or other commonalities.
Big Data analytics can help identify specific subpopulations for which a “failed” drug can still be a success. This practice of drug repurposing will be especially beneficial to patients suffering from rare or genetic diseases that may not by commercial attractive for dedicated product development, but clinically relevant and imperative – that is precision or personalized medicine for you.
Big Data Analytics in Controlling Epidemics, Disease Mapping & Emergency Situations
During the Ebola fiasco in Africa, mobile phone location data is proving highly valuable in efforts to track population movements, which helps to predict the spread of the Ebola virus (Source: BBC). Thus, Big Data can also help in the fight against the spread of epidemics.
The use of various platforms, apps, and services to collect data and communication and provide real-time information about Ebola outbreak situations and developments was a great success. Another example is UNICEF’s development and utilization of Edutrac, a mobile mobile-based data-collection system that was able to collect real-time data for ensuring hygiene equipment had been delivered to schools in Sierra Leone.
In Europe, a surveillance system called Influenzanet uses data available online to gather self-reported volunteer info about symptoms related to not only influenza but also Salmonella, E. Coli, and the Zika virus. Scientists are able to then map out geographical risk charts in order to better predict future outbreaks, should they begin to accumulate. All these cases clearly show the impact of big data on disease mapping.
Hospitals in Paris have been using data and machine learning from a variety of sources to come up with daily and hourly predictions of how many patients are expected to be at each hospital (Source: Forbes).
Education & Training in Health Data Science & Analytics
Although it is possible to make a career in healthcare analytics via multiple backgrounds (biomedical science, biotechnology, bioinformatics, computational biology, pharmacy, computer science, mathematics, statistics, and biomedical engineering), a Master’s degree in Health Analytics (or related program) and/or a research program can give you an edge.
But, make no mistake. Relevant skills and understanding of both healthcare & life sciences and analytics will give you more brownie points than a specialized Master’s degree in this domain. The following list is not in any particular order for rankings. Use this list if you want a boost to make a career in big data analytics in healthcare and medicine.