Digital health records, blood analysis results, digital images: how to use the entire information for the early diagnosis or prognosis prediction of impairing neurodegenerative diseases and cancers?
How we do it
Digital images can be expressed with hundreds of “textural features”, numbers, which identify specific patterns. The values of blood analysis are also numbers. Finally, digital health records annotated by medical doctors can be also represented as numbers in a database. A patient is therefore uniquely identified by a “chain” of these numbers. This is the Radiolytx digital Patient ID, representing the frontier of “radiomics”. Machine learning algorithms compare the patient with the categories in the rich database and classifies it according to the risk of developing the target illness.
From any device, even a Smartphone, all you need is just a simple internet connection
The user uploads the medical image (PET, CT or MRI) in DICOM format, the blood analysis as data format and the additional annotations as free keywords. Tools are available for image segmentation in order to define a proper region of interest. ROIs are either selected using seeds defined by the user, or with pre-existing templates as in case of the brain. Radiomics parameters are calculated in the defined region of interest and analyzed in view of the clinical test selected by the user among the available ones.
Time is success, for patients and for doctors
The assessment of cancer outcome is a key-problem in diagnosis and therapy. If the aggressivity of a cancer is verified early enough, personalized diagnostic paths can be foreseen. Similarly, neurological diseases rely on an early identification of the degradation of brain functionality before the appearance of clinical symptoms.
Who we are
Founded in 2017, we are a Berlin based Health-Tech Company
We are a multidisciplinary team composed of mathematicians, physicists, engineers, software designers and business developers. Our creative community is responding with digital innovations to the demanding needs of clinicians, mainly in oncology and neurology.
What we do
Utilizing advanced technologies to redefine medicine
We developed and we continuously extend our all-inclusive platform solution with a series of analytic models for functional medical images. Working in close cooperation with neurologist and nuclear medicine clinicians we write quantitative algorithms for the evaluation of cancer therapies, for early diagnosis of neurological and oncological diseases, for prevention.
Despite the request of new analytical tools accessible to the medical community, there was no quantitative and tested platform available until now.
Innovative & Unique
Radyolitx patient ID is a unique combination of three-dimensional
radiomic features, digital health records and blood analysis values.
Thanks to the rich information, it can be associated to certain
pathologies, with a successful patient classification.
Our unconventional and new approach considers not only the averages of
quantities, but their spatial and temporal distribution in an
all-inclusive model and software. We propose thus a 4-dimensional
revolution in digital healthcare.
Our unconventional and new approach considers not only the averages of quantities, but their spatial and temporal distribution in an all-inclusive model and software. We propose thus a 4-dimensional revolution in digital healthcare.
Easy upload and analysis:
DICOM data can be imported from any device and are analyzed centrally in our Memory Driven Computing based cluster with HP technology.
„Simple & advanced 3D analysis of cancer tissues and functionality"
The conventional analysis of functional medical images is based on the concept of standard uptake value. In few words, the medical doctor identifies often just manually the volume of the cancer and an average quantity of radiation is measured in such a volume. The higher is the radiation, the stronger is the metabolism of the cells in the volume, the higher is the activity of the cancer. It is well known that a tumoral mass cannot be considered as a uniform agglomerate of cells, but it is a heterogeneous distribution of cancer cells.
Radiolytx able to individuate specific features of cancers, which are correlated to a diagnostic pattern. Working together with nuclear medical doctors, can identify from medical images quantitative biomarkers related to prognosis, diagnosis and validation of therapies.
„From neurons to pixels & back: digital brain“
Functional medical imaging represents the basic diagnostic tool for the investigation of brain functionality in a large number of neurological diseases, such as AD, Parkinson among others. The traditional analysis of functional brain medical images is based on the measurement of the average standard uptake value in a given region of the brain. It is well known that such approach is well suited to the diagnosis of neurological diseases, but does not allow an early detection.„From neurons to pixels & back: digital brain“
Working with neurologists and biologists, RADIOLYTX is able to identify specific textures in the different brain regions. Through the correlation of the behavior of the interconnected regions, our algorithms are able to identify new personalized biomarkers for neurological diseases, supporting the neurologists and scientists in the early diagnosis and therapy.
Benefiting the community - locally and globally
We think that our power, and our strategy, is the availability of a cross-disciplinary team, involving biologists, mathematicians, physicists, clinicians, software developers and engineers, with previous experience in the academia and in the industry.
The keyword of our Team is creativity.
We apply to medicine disruptive data modeling strategies, often adapted from other fields such as particle physics, in order to provide a novel perspective of medical images data. While the data modeling and analysis algorithms are embedded in our advanced Radiolytx platform and maintained by our teams, users are given an easy-to-go time-saving solution.