Conventional qualitative and subjective methods of interpreting medical images have a bad reputation due to their observer-dependent and highly variable nature, which can affect the accuracy of diagnosis and patient outcomes. To reduce variability and maintain higher patient outcomes, the goal, quantitative image biomarkers (QIB) are being introduced medical image analysis.

Quantitative imaging biomarkers are objective physical and biological characteristics calculated and measured from any type of medical imaging. QIBs can help identify normal biological processes, pathogenic processes, or response to treatment and intervention.

3 applications of quantitative image biomarkers in the healthcare field

Connect medical images with biological events

QIBs show how the disease affects a patient’s organs and systems. These details can be crucial in making an accurate diagnosis, defining the accurate treatment plan, and evaluating the response to treatment.

Take the example of non-alcoholic steatohepatitis. This condition involves steatosis and inflammation of the liver, which leads to chronic liver disease and cirrhosis. Steatosis and liver inflammation should be measured together, as both QIBs are synergistic toward the progression of parenchymal fibrosis.

In this case, magnetic resonance imaging in multiple sequences allows the necessary quantitative biomarkers to be extracted. T1-T2 sequences can help estimate fat deposition, and a diffusion-weighted MRI image with the intravoxel inconsistent motion model provides an estimate of the inflammation state. In combination, these sequences create a multivariate prognostic biomarker for the early diagnosis and classification of steatohepatitis.

Support precision medicine and bedside decision making

Image biomarkers, comparable to specimen biomarkers, provide information about the molecular processes in a patient’s body. However, specimen biomarkers have some limitations, which image biomarkers can overcome.

Biological samples are usually acquired by biopsy or from body fluids. The main disadvantage is that the collection of specimens involves taking only a small part of the normal or damaged tissue, which causes a spatial sampling bias. Imaging scans, in turn, typically show a broad tissue segment or even the entire patient, thus allowing for more complete heterogeneous data.

In addition, image data can be acquired dynamically and longitudinally without additional invasive procedures, which is impossible for specimen biomarkers when frequent and continuous analysis of tissues or liquids is needed.

Consequently, imaging biomarkers can help providers close gaps related to specimen biomarkers and improve therapeutic development, treatment follow-up, and general bedside decision-making. In addition, current computing capabilities and image processing methods are able to extract QIBs at little or no cost, which facilitates the integration of quantitative imaging into the radiology department’s workflows.

Rethink medical imaging technology

In accordance with medicalphysicsweb, with advances in imaging technology in recent years, we are approaching the practical limits of spatial resolution for almost all modes. Therefore, the focus of imaging technology improvement gradually shifts from image quality to the value of image information.

Although a medical image allows you to recognize, measure, and map visible changes in the tissues, blood flow, bones, and other structures within a patient’s body, there is still a gap between imaging measurements and conditions. underlying. For example, the functional architecture of the brain can be easily studied using functional magnetic resonance imaging, but connecting the map of neuronal activity and the underlying neurochemical and electrophysiological processes is still problematic.

Quantitative image biomarkers can reflect specific physiological phenomena or biophysical properties, providing the necessary connections between image and state. Developing them further is the next step in the evolution of imaging technology, shifting the goal of medical imaging from showing the “best image possible” to extracting valuable knowledge.

Quantitative image biomarkers in different modalities

Because current medical imaging technology provides high-resolution vision of a patient’s body, it is possible to extract several quantitative biomarkers of imaging within most modalities:

  • A CT The signal has a high spatial resolution, so CT scans can provide accurate distance measurements, which include: basic tissue characterization, tumor size or volume, quantitative physiological information related to perfusion or necrosis, angiographic information, and functional information. using dynamic contrast techniques.
  • With proper calibration and standardization, an magnetic resonance imaging The signal response can be translated into quantitative biomarkers of tissue structure and function.
  • A PET the signal has a high sensitivity. PET radiopharmaceuticals are able to test a wide variety of physiological or molecular functions.
  • Ultrasound the signals are suitable for QIBs. They allow to extract quantitative information on reflection, attenuation, refraction, properties of bulk tissue and shear wave velocity. In addition, quantitative measures of distance, elasticity, and Doppler flux can be calculated.

2 models of image biomarkers

There are two main models of image biomarkers: the static anatomical model and the biological dynamic model.

2 models of image biomarkers

He static anatomy The model estimates tissue aspects related to tissue volume and shape, topology, and co-occurrence matrix characteristics for texture classification. For example, analyzes of cortical thickness and pulmonary emphysema are considered static methods.

He dynamic biological The model evaluates different physical, chemical, and biological characteristics, such as measurements of fat and iron within the pancreas. These biomarkers are obtained after the dynamic biological modeling of the acquired data.

Challenges to QIB implementation

The main challenge that hinders the widespread adoption of QIB is rooted in the complexity of QIB development and implementation processes. These processes have several consecutive steps, including the definition of the target distinctive, the source images, the analytical methodology, and the type of measures.

To validate and integrate quantitative image biomarkers into clinical practice, they must meet requirements such as conceptual coherence, technical reproducibility, appropriate accuracy, and meaningful adequacy. Ensuring this level of standardization in QIBs requires a coordinated effort between various parties: imaging and software device manufacturers, regulatory organizations, vendors, academic institutions, and more.

In addition, continuous technological advancement in medical imaging hardware and software requires a constant reassessment of the quantitative accuracy of medical imaging and periodic updates of standardization requirements.

The solution to implementation challenges

To drive the development, validation, and integration of QIBs, the Radiological Society of North America (RSNA) organized the Quantitative Imaging Biomarkers Alliance.QIBA). The mission of this collaboration is to improve the value and practicality of quantitative imaging biomarkers by reducing variability between devices, patients, and time.

To achieve this, they created the so-called QIBA profile. This document describes in detail a quantitative image biomarker that needs to be validated, explaining properties such as:

  • Intended use of QIB or clinical context
  • An image acquisition protocol
  • Compliance elements
  • A claim to the minimum achievable variability / bias

According to the QIBs described, each profile is designed for different audiences, including imaging device manufacturers, pharmaceutical companies, researchers, technologists, physicians, regulatory and accreditation authorities, among others.

Distant perspective of QIB adoption

From what we are seeing now, medical imaging technology is evolving in the direction that is beneficial for the development of quantitative imaging biomarkers. However, there are multiple challenges on the path to their integration into real life.

In addition to the fact that QIB implementation may be hampered by “inconsistent or incorrect use of terminology or methods for technical performance and statistical concepts,” according to QIBA, there are also infrastructure-related issues.

To enable the use of QIBs in clinical and research environments, a whole network of image acquisition protocols, visualization methods, image analysis guides, and developed, standardized, and optimized reporting structures need to emerge. . It will take a long time, even with QIBA on board to facilitate the big challenges.

However, with a high potential for quantitative imaging biomarkers for precision medicine, diagnosis, and drug research, this effort will ultimately be rewarding.

Medical image analysis by ScienceSoft

Innovates in the prevention, diagnosis and treatment of diseases with an efficient analysis of medical images.


Source link