With nearly 3,000 employees, Finnish industrial R&D firm VTT has a strong presence, from Northern Europe to Silicon Valley. Now, with the upcoming launch of megNet — a software tool designed to integrate omics data with higher level images of cells and tissues — that presence could become even stronger, especially in the cardiology sector, where the company sees keen demand for such software.
MegNet’s chief research scientist, Matej Oresic of VTT’s Technical Research Center in Finland, said the software “is of particular interest to large healthcare companies like GE Healthcare and Phillips.”
The new software, which will be commercially available in the fall, follows a series of VTT entrees into the biotech domain. “Traditionally, the company has served the metabolic engineering domains such as bioethanol, nutrition, and protein production from microbes and engineering plants for production of high-value chemicals,” Oresic said.
The company also has a presence in IT, developing software for Nokia and simulation software for Rolls Royce airplane engines. VTT has also entered the systems biology field, combining its expertise in both IT and biotech.
Now with 3D imaging tool megNet, VTT seeks to marry omics data and magnetic imaging.
The software is designed to do this in a variety of ways, such as by quantifying image data and serum metabolic profiles for metabolic fingerprinting. “The data will be combined within the megNet framework, both for potential diagnostic purposes as well as interpretation,” said Oresic.
In proof-of-concept studies, Oresic said VTT used the software to integrate data from several sources and reconstruct biological pathways in order to successfully identify two pathways leading to the production of ceramides, toxic lipids known to affect insulin sensitivity. “Interestingly, our findings were recently confirmed independently since pharmacological inhibition of one of the pathways led to an increase in insulin sensitivity,” he said.
While megNet’s debut is a few months off, VTT is currently working with Finnish and other European healthcare providers, including GE Healthcare in combining medical images with metabolomic data, hoping to pinpoint new biomarkers and other precursors for disease. The software is designed to complement existing tools.
Oresic said the mindset behind megNet is the disparity between what is known scientifically at the molecular level and what is employed in practice. “We are struggling today to apply the rapidly increasing basic knowledge in medical sciences towards better healthcare products [and/or] drugs as well as diagnostics,” he said.
The evidence is overwhelming, according to Oresic, that “we need better predictive tools at the human level; not just for the rats and mice.”
Oresic touted megNet’s ability to enable translational research. He added that the software is in line with the goals of the FDA’s 2004 Critical Path initiative for speedier drug approval.
Oresic said that drug development is still “based on fairly old methods,” and that not much has been gained from the molecular-level knowledge generated over the past few decades. “It’s surprisingly slow [bringing new technologies and drugs to market], even though you would think we are in the middle of a massive revolution. Really, not that much has been translated into success stories in this area.”
Cardiologist Tiina Helio with Helsinki University Central Hospital has been a clinical partner in developing megNet. She said that with the software, ideally a doctor could better screen possible heart patients because it could be offered in addition to conventional methods such as the electrocardiogram, which is sometimes inconclusive.
“The clinical problems we face include seeing patients who are suspected of having a cardiomyopathy and some abnormal readings. We refer them for further evaluation because perhaps they don’t meet all the criteria of a myopathy. But the question is whether they could develop [a myopathy] in the future. MegNet would be very useful in this regard; and later it might be possible to phenotype even larger subgroups,” she said.
“The data will be combined within the megNet framework, both for potential diagnostic purposes as well as interpretation.”
Jerki Lotjonen, a senior research scientist with VTT’s signal and image processing group, has been working on the imaging part of the project, particularly with regard to modeling the heart. “How I see this,” he said, “is that there are two commercial uses. One is data mining and the other is as a diagnostic tool — such as when a patient enters a hospital, the doctor could use this to provide information [about the patient’s condition] … such as through medical images.”
Now, the challenge is the sheer volume of data that must be dealt with, said Lotjonen. With megNet, “We can say, [for example], that ‘this patient belongs to this population; or [determine] which patient category that patient belongs to.”
Helio is confident that the technology “would be worthwhile in treating diseases for the myocardium” and reckons that while megNet is a tool for other disciplines, cardiology could benefit most. “Because [here in cardiology] imaging is most important, we try to avoid complications and invasive [procedures]” whenever possible, she explained.
Certain target customers — such as Hitachi or Siemens — may be interested in incorporating megNet into their own imaging software applications, according to Oresic. “MegNet is rather flexible concerning the underlying database setup, but of course each configuration requires a certain degree of customization.”
Oresic said that the technology is designed to quantify both molecular-scale and organ-scale information.
“The heart image can be described by multiple measurable parameters such as wall motion or wall thickness in different sections. The image can be quantitated the same way as, for example, mass spectrometric molecular profiles can be represented by a set of numbers, such as those corresponding to metabolite concentrations.”