Datameer Case Studies Using Big Data Analytics to Create Better Outcomes for Cancer Patients

Edit This Case Study Record

	Using Big Data Analytics to Create Better Outcomes for Cancer Patients Datameer

Using Big Data Analytics to Create Better Outcomes for Cancer Patients

Datameer

Technology Category	Analytics & Modeling - Big Data Analytics Infrastructure as a Service (IaaS) - Cloud Computing
Applicable Industries	Healthcare & Hospitals
Applicable Functions	Product Research & Development
Use Cases	Clinical Image Analysis Predictive Maintenance
Services	Data Science Services
Challenge	Cancer diagnosis is complicated due to the uniqueness of each case, and treatment outcomes vary greatly from patient to patient. DKFZ, the largest biomedical research Institute in Germany, is working to understand the mechanisms of cancer, identify risk factors, and find new ways to prevent people from getting cancer. A key focus of DKFZ’s medical researchers is genomic data research. However, due to the massive volumes of genomic data involved in this research, DKFZ faced huge challenges on the data and analytics front. Their analytic systems were overwhelmed by many petabytes of data, and analyzing an entire patient data set took weeks and even months to complete. These huge bottlenecks greatly slowed research and frustrated staff. Read More
About Customer	DKFZ is the largest biomedical research Institute in Germany. They are working to understand the mechanisms of cancer, identify risk factors, and find new ways to prevent people from getting cancer. A key focus of DKFZ’s medical researchers is genomic data research. By analyzing human genomes, they can identify DNA problems which are the root cause of cancer in an individual — information that can be used to personalize cancer treatment — as well as understand the genetic evolution of disease and monitor patient responses to different treatments to better understand their efficacy. Read More
Solution	DKFZ collaborated with Fujitsu to deploy the Fujitsu Prime Flex Integrated System for Hadoop, a powerful and scalable server cluster that uses Datameer’s analytic platform for parallel processing power and incredibly fast, actionable analytics. With Datameer, researchers can analyze the complete, raw genomic data sets of multiple patients in parallel, along with patient data records, detail selection data, and reference genome data. Analysis on the complete, unreduced data set can be performed by block of data and by patient. Using this new parallelization technique the analysis on the complete data set can be completed between five and twenty minutes. By comparisons, the analysis on the greatly reduced data set used to take 24-48 hours to complete. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Big Data Analytics

Infrastructure as a Service (IaaS) - Cloud Computing

Applicable Industries

Healthcare & Hospitals

Applicable Functions

Product Research & Development

Use Cases

Clinical Image Analysis

Predictive Maintenance

Services

Data Science Services

Challenge

Cancer diagnosis is complicated due to the uniqueness of each case, and treatment outcomes vary greatly from patient to patient. DKFZ, the largest biomedical research Institute in Germany, is working to understand the mechanisms of cancer, identify risk factors, and find new ways to prevent people from getting cancer. A key focus of DKFZ’s medical researchers is genomic data research. However, due to the massive volumes of genomic data involved in this research, DKFZ faced huge challenges on the data and analytics front. Their analytic systems were overwhelmed by many petabytes of data, and analyzing an entire patient data set took weeks and even months to complete. These huge bottlenecks greatly slowed research and frustrated staff.

About Customer

DKFZ is the largest biomedical research Institute in Germany. They are working to understand the mechanisms of cancer, identify risk factors, and find new ways to prevent people from getting cancer. A key focus of DKFZ’s medical researchers is genomic data research. By analyzing human genomes, they can identify DNA problems which are the root cause of cancer in an individual — information that can be used to personalize cancer treatment — as well as understand the genetic evolution of disease and monitor patient responses to different treatments to better understand their efficacy.

Solution

DKFZ collaborated with Fujitsu to deploy the Fujitsu Prime Flex Integrated System for Hadoop, a powerful and scalable server cluster that uses Datameer’s analytic platform for parallel processing power and incredibly fast, actionable analytics. With Datameer, researchers can analyze the complete, raw genomic data sets of multiple patients in parallel, along with patient data records, detail selection data, and reference genome data. Analysis on the complete, unreduced data set can be performed by block of data and by patient. Using this new parallelization technique the analysis on the complete data set can be completed between five and twenty minutes. By comparisons, the analysis on the greatly reduced data set used to take 24-48 hours to complete.

Impact #1	DKFZ can now analyze 10 TB of raw data per day – the equivalent of 140 billion records looking at 900,000 thousand positions in each genome.
Impact #2	They can analyze complete data sets in minutes, eliminating the need to reduce data and risk missing out on key insights.
Impact #3	Vastly faster processing enables the DKFZ to more quickly identify specific, optimal cancer therapies for each patient, as well as further their overall research on correlations between cancer and genetics.

Impact #1

DKFZ can now analyze 10 TB of raw data per day – the equivalent of 140 billion records looking at 900,000 thousand positions in each genome.

Impact #2

They can analyze complete data sets in minutes, eliminating the need to reduce data and risk missing out on key insights.

Impact #3

Vastly faster processing enables the DKFZ to more quickly identify specific, optimal cancer therapies for each patient, as well as further their overall research on correlations between cancer and genetics.

Benefit #1	Analysis on the complete data set can be completed between five and twenty minutes, compared to the 24-48 hours it used to take.
Benefit #2	DKFZ can now analyze 10 TB of raw data per day – the equivalent of 140 billion records looking at 900,000 thousand positions in each genome.

Benefit #1

Analysis on the complete data set can be completed between five and twenty minutes, compared to the 24-48 hours it used to take.

Benefit #2

DKFZ can now analyze 10 TB of raw data per day – the equivalent of 140 billion records looking at 900,000 thousand positions in each genome.

Download PDF Version

Overview

Using Big Data Analytics to Create Better Outcomes for Cancer Patients

Operational Impact

Quantitative Benefit