Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Tissue characterization from ultrasound B-scan data.

D Nicholas, D K Nassiri, P Garbutt

    Ultrasound in Medicine & Biology
    |February 1, 1986
    PubMed
    Summary
    This summary is machine-generated.

    Related Concept Videos

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    A model of well-being for children with neurodevelopmental disorders: Parental perceptions of functioning, services, and support.

    Child: care, health and development·2017
    Same author

    International outbreak of multiple Salmonella serotype infections linked to sprouted chia seed powder - USA and Canada, 2013-2014.

    Epidemiology and infection·2017
    Same author

    The impact of autism services on mothers' psychological wellbeing.

    Child: care, health and development·2016
    Same author

    Altered breathing mechanics and ventilatory response during exercise in children born extremely preterm.

    Thorax·2016
    Same author

    About my Child: measuring 'Complexity' in neurodisability. Evidence of reliability and validity.

    Child: care, health and development·2016
    Same author

    Biology and Genetics of New Autosomal STR Loci Useful for Forensic DNA Analysis.

    Forensic science review·2015
    Same journal

    Multidimensional Safety Assessment of a Low-Intensity Scanning Ultrasound (SUS) Protocol in Sheep.

    Ultrasound in medicine & biology·2026
    Same journal

    Acoustic Characterization of a Modified IEC Agar-Based Tissue-Mimicking Material Across the 3.5-50 MHz Frequency Range.

    Ultrasound in medicine & biology·2026
    Same journal

    Deep Learning-Based Standard Section Recognition and Multi-Organ Segmentation in Upper Abdominal Ultrasound.

    Ultrasound in medicine & biology·2026
    Same journal

    Cardiac Natural Mechanical Wave Detection and Speed Estimation Using Deep Learning-Based 2-D Ultrasound Imaging: A Feasibility Study.

    Ultrasound in medicine & biology·2026
    Same journal

    Region-Specific Evaluation of Plaque Segmentation in Cross-sectional Projections of Carotid Ultrasound Images Using Deep Learning Models in a Sub-clinical Atherosclerosis Cohort.

    Ultrasound in medicine & biology·2026
    Same journal

    Simulating the Dedifferentiation Process of Thyroid Cancer: Insights from Mouse Models and Ultrasound Imaging.

    Ultrasound in medicine & biology·2026
    See all related articles

    This article presents a computational method to identify different body tissues by analyzing the visual patterns, or textures, found in ultrasound images. By extracting 93 distinct features from image segments, the researchers developed a system to distinguish between healthy liver and spleen tissues. Their findings show that this automated analysis can accurately classify these organs, suggesting that subtle changes in tissue appearance can be detected through this quantitative approach.

    Area of Science:

    • Medical imaging diagnostics within ultrasound B-scan tissue characterization
    • Biomedical signal processing and computational pattern recognition

    Background:

    No prior work had resolved how to quantify subtle visual variations within standard medical sonography images for diagnostic purposes. That uncertainty drove the need for automated pattern recognition tools. Prior research has shown that human interpretation of organ appearance often suffers from subjective variability. This gap motivated the development of objective metrics for analyzing image data. Existing diagnostic protocols frequently rely on qualitative assessments rather than precise mathematical descriptors. Researchers have long sought to transform raw visual data into actionable clinical insights. That challenge persisted because biological structures often exhibit overlapping visual characteristics. This study addresses the requirement for standardized, reproducible methods to evaluate parenchymal integrity using non-invasive imaging techniques.

    Purpose Of The Study:

    The aim of this study is to describe a systematic approach for characterizing biological tissues using textural features derived from ultrasound B-scan images. Researchers sought to address the limitations of traditional visual interpretation by developing a quantitative framework. The primary motivation was to create an objective method for identifying subtle changes in organ parenchyma. This problem is significant because standard sonography often relies on subjective assessment, which may lead to diagnostic inconsistencies. The authors aimed to define a comprehensive set of descriptors that could reliably distinguish between different tissue types. They also intended to establish criteria for selecting the most effective combinations of these features to maximize diagnostic accuracy. By applying this method to human liver and spleen data, the team tested the practical utility of their computational model. This work serves to advance the field of medical imaging by introducing standardized, reproducible techniques for tissue evaluation.

    Keywords:
    medical imagingpattern recognitiondiagnostic sonographytissue classification

    Frequently Asked Questions

    The researchers utilize a set of 93 distinct textural features extracted from 64x64 pixel image segments. These descriptors allow the system to mathematically quantify visual patterns, which are then analyzed to distinguish between different biological structures.

    The study employs a computer-based analysis framework to process ultrasound B-scan data. This approach involves selecting optimal combinations of textural descriptors to maximize classification performance, moving beyond simple visual interpretation of the images.

    A regional analysis is necessary because the technique relies on identifying uniform changes within the organ parenchyma. By focusing on specific regions of interest, the researchers can isolate the textural signatures required for accurate classification.

    The researchers use B-scan image data, specifically 8-bit deep segments, as the primary input for their classification models. This quantitative data type enables the extraction of the 93 textural features required for the study.

    Related Experiment Videos

    Main Methods:

    The review approach involved acquiring 64 by 64 pixel segments from specific regions of interest within ultrasound images. Each segment contained 8-bit depth information to ensure sufficient data resolution for subsequent processing. The researchers systematically defined a comprehensive set of 93 textural descriptors to characterize the visual patterns. They discussed rigorous criteria for selecting the most effective combinations of these descriptors to optimize classification performance. The team validated the power of their framework by discriminating between liver and spleen textures from healthy human subjects. They evaluated the performance using both a training set alone and a combined training and test set strategy. This design ensured that the classification models were tested for reliability and generalizability. The methodology focused on transforming raw visual information into quantifiable metrics for objective tissue assessment.

    Main Results:

    Key findings from the literature demonstrate that the proposed approach achieves an overall test probability of success of 82% when analyzing a single image. The accuracy increases to 94% when the system evaluates multiple images from a single subject. These results confirm the utility of the textural features in distinguishing between liver and spleen tissues. The researchers successfully quantified various measures of success across different testing configurations. Their data reveal that the combination of multiple image frames significantly enhances the classification reliability. The findings suggest that the technique is capable of identifying subtle variations in parenchymal texture. This quantitative performance indicates that the approach provides a robust alternative to qualitative visual assessment. The results highlight the potential for automated systems to improve diagnostic outcomes in clinical sonography.

    Conclusions:

    The authors propose that their computational framework effectively differentiates between distinct organ types based on textural patterns. Synthesis and implications suggest that this methodology holds promise for identifying uniform parenchymal alterations. The researchers demonstrate that analyzing multiple images per patient significantly improves diagnostic accuracy compared to single-frame evaluations. Their data indicate that specific feature combinations provide a robust basis for automated tissue classification. This study highlights the potential for quantitative sonography to augment traditional diagnostic workflows. The authors emphasize that their approach remains sensitive to subtle changes that might otherwise escape visual inspection. These findings support the integration of advanced signal processing into routine clinical ultrasound examinations. The work provides a foundation for future investigations into characterizing various pathological states using standardized image descriptors.

    The researchers measured success rates of 82% for single images and 94% for subjects with multiple images. These metrics quantify the power of the approach in discriminating between human liver and spleen textures.

    The authors propose that this methodology could be applied to conditions where subtle, uniform parenchymal changes are present. They suggest that such quantitative techniques offer a viable path for improving diagnostic precision in clinical settings.