Ultrasound image texture analysis for characterizing intramuscular fat content of live beef cattle

Identification of EIMD Level Differences Between Long- and Short Head of Biceps Brachii Using Echo Intensity and GLCM Texture Features

Purpose: This study aimed to compare the time-course changes of exercise-induced muscle damage (EIMD) levels in the long head of biceps brachii (LHB) and short head of the biceps brachii (SHB) using echo intensity (EI) and to determine the efficiency of the gray level co-occurrence matrix (GLCM) texture parameters. Methods: The participants performed 30 maximal eccentric contractions of the elbow flexor. Along with muscle damage indicators, including circumference, range of motion, muscle soreness, and maximal voluntary isometric contraction (MVIC), the EI and GLCM texture features of the LHB and SHB was also assessed using B-mode ultrasonography. All measurements were assessed pre- and immediately post-exercise and after 24, 48, 72, and 96 h. Results: The muscle damage indicators indicated significant changes after the eccentric contractions (p < 0.01 for circumference, range of motion, muscle soreness, and MVIC). The EI of LHB significantly increased following the contractions (p < 0.01), but that of SHB did not (p > 0.05). In contrast, for the GLCM texture parameters, there were significant changes in the SHB (p < 0.01 for homogeneity, energy, and entropy). Conclusion: Thus, this study demonstrated that EIMD severity is different between LHB and SHB even within the same muscle. In the GLCM features, the time course of SHB after eccentric contraction revealed different patterns compared with those of LHB. Therefore, even if there are no changes in EI within a target muscle following muscle contractions, new information on muscle quality can be obtained through GLCM analysis.

Quantitative Texture Analysis of Parotid Gland Ultrasound Images Yield Higher Correlation with Scintigraphy than Semiquantitative Scoring in Primary Sjögren's Syndrome Patients

Background

Ultrasound (US) can detect salivary gland abnormalities in primary Sjögren's syndrome (SS). This study aimed to compare the correlation among the semiquantitative US scores, texture features, and the quantitative salivary gland scintigraphy (SGS) results.

Methods

This retrospective study included 11 patients who were diagnosed with primary SS and underwent US examinations of the parotid glands and SGS simultaneously. We evaluated SGS quantitatively based on the calculation of maximum accumulation ratio (MAR) and stimulated excretion fraction (EF). The US findings were accessed through the semiquantitative Outcome Measures in Rheumatology scoring system and by gray-level co-occurrence matrix (GLCM) texture analysis. Spearman's rank correlation tests were performed.

Results

A significant moderate negative correlation was noted between the semiquantitative US score and MAR (rho = -0.57, P = 0.006), but not with EF (rho = -0.11, P = 0.613). The GLCM texture metrics, including contrast, dissimilarity, and homogeneity, were all determined to be significantly associated with both MAR and EF. The GLCM contrast correlated moderately to MAR (rho = -0.66, P = 0.001). The GLCM homogeneity highly correlated to EF (rho = 0.74, P < 0.001). The contrast and homogeneity can still discriminate the changes in MAR and EF in the subgroups with the same semiquantitative US scores.

Conclusion

US findings on parotid gland can correlate with SGS results when analyzed based on GLCM texture features. With the GLCM texture metrics, US appears to be an excellent imaging tool for the assessment of the parotid glands in primary SS patients.

Explainable Ensemble Machine Learning for Breast Cancer Diagnosis Based on Ultrasound Image Texture Features

Image classification is widely used to build predictive models for breast cancer diagnosis. Most existing approaches overwhelmingly rely on deep convolutional networks to build such diagnosis pipelines. These model architectures, although remarkable in performance, are black-box systems that provide minimal insight into the inner logic behind their predictions. This is a major drawback as the explainability of prediction is vital for applications such as cancer diagnosis. In this paper, we address this issue by proposing an explainable machine learning pipeline for breast cancer diagnosis based on ultrasound images. We extract first- and second-order texture features of the ultrasound images and use them to build a probabilistic ensemble of decision tree classifiers. Each decision tree learns to classify the input ultrasound image by learning a set of robust decision thresholds for texture features of the image. The decision path of the model predictions can then be interpreted by decomposing the learned decision trees. Our results show that our proposed framework achieves high predictive performance while being explainable.

Correlations among Ultrasonographic, Physicochemical and Sensory Characteristics of Pectoralis Major Muscles in Turkeys Reared in a Sustainable Farming System

Simple Summary

One of the challenges of the contemporary poultry industry is to obtain reliable information on meat quality throughout the entire production cycle. Previous studies have shown that computerized analysis of the ultrasonographic images in live birds is a promising method to predict certain characteristics of skeletal muscles. In the present study, the left pectoralis major muscle was scanned just before slaughter in forty-five meat-type turkeys reared in an organic farm. Physicochemical and sensory attributes of the pectoral muscles were determined after slaughter using validated laboratory and analytical methods. There were several significant correlations among ultrasonographic image attributes and physical/sensory characteristics of pectoralis major muscles in the turkeys of the present study, but the moisture content was the only chemical trait associated with ultrasound images. The strongest overall correlation was between pixel heterogeneity obtained in the muscles examined in an oblique plane and aroma. The occurrence and strength of quantitative correlations among ultrasound image characteristics in situ and post-mortem traits of turkeys’ breast are clearly affected by a scanning plane. Computerized analysis of pectoral muscle ultrasonograms provides information on several characteristics that are indicative of meat quality and hence could be potentially used in commercial settings and breed development programs.

Abstract

This study set out to examine associations among echotextural, physicochemical and sensory attributes of the pectoralis major muscles in 17-week-old organic turkeys (B.U.T. Big-6) varying in the amount of wheat and oat grain in daily feed rations (Group C: complete feed only; Group Exp1: 5–30% of wheat and 0–20% of oat; and Group Exp2: 5–50% of wheat and 0–50% of oat; n = 15 turkeys/group). Digital ultrasonograms of the left pectoral muscle in four different planes (longitudinal-L, transverse-T, and two oblique planes-O1 and O2) were obtained with a 5.0-MHz linear-array transducer just before slaughter. Mean numerical pixel intensity (MPI) and pixel heterogeneity (MPH) of the muscle parenchyma were computed using the ImageProPlus^® analytical software. Ten significant correlations between echotextural attributes and various meat characteristics were recorded in Group C, one in Group Exp1, and eight in Group Exp2. When data were pooled for all birds studied, there were twelve significant correlations (p < 0.05); all but one correlation (between MPH and moisture) were for physical and sensory characteristics of meat samples. Computer-assisted analysis is a potential method to determine moisture as well as physical (e.g., coloration) and sensory (e.g., aroma) characteristics of pectoralis major muscles in organic turkeys.

Correlations between first order echotextural characteristics and chemical composition of pectoralis major muscles in broiler chickens receiving different dietary fat supplements

This study examined the quantitative relationships among ultrasonographic image attributes and chemical composition of the pectoralis major muscles in broiler chickens that received four different dietary fat supplements (Group SO: soybean oil; Group FO: flaxseed oil; Group SO+FO: soybean oil+flaxseed oil; and Group BT: beef tallow; n=10 birds/group). Ultrasonograms of birds’ pectoral muscles, in a transverse (T) and longitudinal (L) plane, were obtained just before slaughter at 6 weeks of age and were subjected to digital image analyses to determine mean pixel intensity (MPI) and pixel heterogeneity values (standard deviation of numerical pixel values; MPH; a.k.a first order echotextural characteristics). Thirty-eight chemical characteristics of the muscles were determined post-mortem (crude fat, protein, and dry matter as well as fatty acid profiles) and were analyzed for correlations with the echotextural variables. A total of 12 (L-MPI: 7; L-MPH: 4; and T-MPH: 1 correlation), 5 (L-MPI: 2; L-MPH: 2; and T-MPI: 1 correlation), 15 (L-MPI: 10; T-MPI: 4; and T-MPH: 1 correlation) and 8 (L-MPI: 2; L-MPH: 1; and TMPH: 5 correlations) significant correlations were recorded in Groups SO, FO, SO+FO and BT, respectively. When the data were pooled for all 40 birds studied, significant correlations with echotextural attributes were recorded for eighteen different chemical constituents, with the strongest overall correlation found between crude fat content and T-MPI (r=0.52, P=0.0005). In conclusion, there exists a potential application for ultrasonographic imaging in situ combined with computerized image analysis to estimate certain chemical constituents of pectoralis major muscles in broiler chickens. However, the existence and strength of correlations among ultrasonographic image attributes and muscle composition are affected by the source of dietary fat and relative abundance (“threshold concentrations”) of individual chemical components.

Grading Bleomycin-Induced Pulmonary Fibrosis in ex vivo Mouse Lungs Using Ultrasound Image Analysis

OBJECTIVES

The aim of this study was to assess bleomycin-induced pulmonary fibrosis on ex vivo mouse lungs using ultrasound image grading and texture analysis.

METHODS

Excised mouse lungs were divided into 3 groups: control, mild fibrosis, and severe fibrosis based on the monitored indicators of health. B-mode ultrasound images were obtained via scanning the mouse lungs ex vivo. The surface smoothness, echo density, and angle of lesions or the lung margin were graded, and the imaging contrast, correlation, homogeneity, and entropy were assessed via texture analysis.

RESULTS

The grades of surface smoothness, echo density, and angle were statistically higher for the severe fibrosis group compared with those of the control and mild fibrosis groups (P < .05). In addition, statistically significant differences in the contrast, correlation, and homogeneity between mild and severe fibrosis groups were observed (P < .05).

CONCLUSIONS

The results obtained in this study suggest that ultrasound image grading and texture analysis are valuable and meaningful methods for assessing pulmonary fibrosis in a bleomycin mouse model.

Muscle Composition Analysis of Ultrasound Images: A Narrative Review of Texture Analysis

Skeletal muscle composition, often characterized by the degree of intramuscular adipose tissue, deteriorates with aging and disease and contributes to impairments in function and metabolism. Ultrasound can provide surrogate measures of muscle composition through measurement of echo intensity; however, there are several limitations associated with its analysis. More complex image processing features, broadly known as texture analysis, can also provide surrogates of muscle composition and may circumvent some of the limitations associated with muscle echo intensity. Here, texture features from the intensity histogram, gray-level co-occurrence matrix, run-length matrix, local binary pattern, blob analysis, texture anisotropy index and wavelet analysis are discussed. The purpose of this review was to provide a conceptual understanding of texture analysis as it pertains to muscle composition of ultrasound images.

The size and echogenicity of the tibialis anterior muscle is preserved in both limbs in young children with unilateral spastic cerebral palsy

PURPOSE

The primary of this study was to compare the volume, length, echo intensity, and growth rate of the medial gastrocnemius (MG) and tibialis anterior (TA) muscle of both limbs (more-involved and less-involved) in children with unilateral spastic cerebral palsy (USCP), with those of an age-matched typically developing (TD) group. A secondary aim in the USCP group was to explore the associations between these muscle parameters and discrete ankle positions during phase of gait.

METHODS

Muscle parameters were assessed using 3D ultrasound. Maximal ankle dorsiflexion in stance and swing during walking were determined from 2D video analysis. Group differences in muscle size and echo intensity were assessed using a two-way analysis of covariance (age-by-group), with the interaction term used to compare muscle growth rates. Associations between muscle parameters and maximal ankle dorsiflexion in stance and swing were assessed using backwards multiple linear regression analyses.

RESULTS

The MG of both limbs in children with USCP had signs of impaired muscle development (smaller volume and length, higher echo intensity and lower growth rate). There was no evidence of impaired muscle development of TA between limbs or compared the TD children. Tibialis anterior volume, length, echo intensity and MG volume explained 66% and 83% of the variance in maximal ankle dorsiflexion position in the stance and swing phases of walking, respectively.

CONCLUSIONS

Unlike the MG, the TA volume and growth rate in children with USCP are equivalent between limbs and compared to TD children. For the more-involved limb only, TA volume, length, and echo intensity appear associated with maximal ankle dorsiflexion during walking and represent important muscle parameters that could be targeted in with early exercise therapy.Implications for rehabilitationTibialis anterior (TA) size and echogenicity appear normal in both limbs in young children with unilateral spastic cerebral palsy (USCP); findings that could indicate sufficient mechanical stimulus and muscle anabolism to maintain normal muscle growth.Tibialis anterior size and echogenicity are associated with maximal ankle dorsiflexion in both stance and swing phase of walking in young children with USCP; though such relations appear isolated to the more-involved limb.Early therapeutic interventions that target TA are likely to be successful in maintaining muscle size and may offset the negative effects of medial gastrocnemius atrophy in the development of fixed ankle equinus of the more-involved limb and improve ankle positioning during gait.

Gain and loss of subcutaneous and abdominal adipose tissue depot mass of German Holstein dairy cows with different body conditions during the transition period

Subcutaneous adipose tissue (SCAT) and abdominal adipose tissue (AAT) depots are mobilized during the fresh cow period (FCP) and early lactation period (ELP) to counteract the negative energy balance (NEB). Earlier studies suggested that fat depots contribute differently to lipomobilization and may vary in functionality. Differences between the adipose depots might influence the development of metabolic disorders. Thus, the gain and loss of subcutaneous and abdominal adipose depot masses in Holstein cows with lower and higher body condition (mean body condition scores: 3.48 and 3.87, respectively) were compared in the period from d -42 to d 70 relative to parturition in this study. Animals of the 2 experimental groups represented adequately conditioned and overconditioned cows. Estimated depot mass (eDM) of SCAT, AAT, retroperitoneal, omental, and mesenteric adipose depots of 31 pluriparous German Holstein cows were determined via ultrasonography at d -42, 7, 28, and 70 relative to parturition. The cows were grouped according to the eDM of SCAT on d -42 [low body condition (LBC) group: n = 16, mean eDM 8.6 kg; high body condition (HBC) group: n = 15, mean eDM 15.6 kg]. Average daily change (prepartum gain and postpartum loss) in depot masses during dry period (DP; from d -42 to d 7), FCP (d 7 to d 28), and ELP (d 28 to d 70) were calculated and daily dry matter intake and lactation performance recorded. Cows of this study stored about 2 to 3 times more fat in AAT than in SCAT depots. After parturition, on average more adipose tissue mass was lost from the AAT than the SCAT depot (0.23 kg/d vs. 0.14 kg/d). Cows with high compared with low body condition had similar gains in AAT (0.33 kg/d) and SCAT (0.14 kg/d) masses during the DP but mobilized significantly more adipose tissue mass from both depots after calving (AAT, HBC vs. LBC: 0.30 vs. 0.17 kg/d; SCAT, HBC vs. LBC: 0.19 vs. 0.10 kg/d). Correlation analysis indicated a functional disparity between AAT and SCAT. In the case of AAT (R² = 0.36), the higher the gain in adipose mass during DP, the higher the loss in FCP, but this was not the case for SCAT. During FCP, a greater NEB resulted in greater loss of mass from SCAT (R² = 0.18). In turn, greater mobilization of SCAT mass led to a higher calculated feed efficiency (R² = 0.18). However, AAT showed no such correlations. On the other hand, during ELP, loss of both SCAT and AAT mass correlated positively with feed efficiency (R² = 0.35 and 0.33, respectively). The results indicate that feed efficiency may not be an adequate criterion for performance evaluation in cows during NEB. Greater knowledge of functional disparities between AAT and SCAT depots may improve our understanding of excessive lipomobilization and its consequences for metabolic health and performance of dairy cows during the transition period.

Quantification of Muscle Tissue Properties by Modeling the Statistics of Ultrasound Image Intensities Using a Mixture of Gamma Distributions in Children With and Without Cerebral Palsy

OBJECTIVES

To investigate whether quantitative ultrasound (US) imaging, based on the envelope statistics of the backscattered US signal, can describe muscle properties in typically developing children and those with cerebral palsy (CP).

METHODS

Radiofrequency US data were acquired from the rectus femoris muscle of children with CP (n = 22) and an age-matched cohort without CP (n = 14) at rest and during maximal voluntary isometric contraction. A mixture of gamma distributions was used to model the histogram of the echo intensities within a region of interest in the muscle.

RESULTS

Muscle in CP had a heterogeneous echo texture that was significantly different from that in healthy controls (P < .001), with larger deviations from Rayleigh scattering. A mixture of 2 gamma distributions showed an excellent fit to the US intensity, and the shape and rate parameters were significantly different between CP and control groups (P < .05). The rate parameters for both the single gamma distribution and mixture of gamma distributions were significantly higher for contracted muscles compared to resting muscles, but there was no significant interaction between these factors (CP and muscle contraction) for a mixed-model analysis of variance.

CONCLUSIONS

Ultrasound tissue characterization indicates a more disorganized architecture and increased echogenicity in muscles in CP, consistent with previously documented increases in fibrous infiltration and connective tissue changes in this population. Our results indicate that quantitative US can be used to objectively differentiate muscle architecture and tissue properties.

Quantitative 3-D Ultrasound of the Medial Gastrocnemius Muscle in Children with Unilateral Spastic Cerebral Palsy

Three-dimensional ultrasound (3-DUS) was used to examine the size and appearance of the medial gastrocnemius (MG) muscle in children with unilateral cerebral palsy (CP). Twenty-six children with CP and 10 typically developing (TD) children participated. Three-dimensional US images of both limbs in children with CP and the right limb in TD children were analysed using quantitative methods to determine muscle volume, global echo intensity, global echo pattern and regional echo intensity. Significant differences in MG volume and all echo parameters were found between TD and CP children. The more involved limb was smaller and had higher echo intensity and a more heterogenous echo pattern compared with the TD group. Compared with that of the more involved limb, the MG of the less involved limb was larger but had a similar echo appearance. The MG of both limbs in children with unilateral spastic CP is smaller and, based on quantitative ultrasound, structurally different from that of TD children.

Advances in quantitative muscle ultrasonography using texture analysis of ultrasound images

Musculoskeletal ultrasound imaging can be used to investigate the skeletal muscle structure in terms of architecture (thickness, cross-sectional area, fascicle length and fascicle pennation angle) and texture. Gray-scale analysis is commonly used to characterize transverse scans of the muscle. Gray mean value is used to distinguish between normal and pathologic muscles, but it depends on the image acquisition system and its settings. In this study, quantitative ultrasonography was performed on five muscles (biceps brachii, vastus lateralis, rectus femoris, medial gastrocnemius and tibialis anterior) of 20 healthy patients (10 women, 10 men) to assess the characterization performance of higher-order texture descriptors to differentiate genders and muscle types. A total of 53 features (7 first-order descriptors, 24 Haralick features, 20 Galloway features and 2 local binary pattern features) were extracted from each muscle region of interest (ROI) and were used to perform the multivariate linear regression analysis (MANOVA). Our results show that first-order descriptors, Haralick features (energy, entropy and correlation measured along different angles) and local binary pattern (LBP) energy and entropy were highly linked to the gender, whereas Haralick entropy and symmetry, Galloway texture descriptors and LBP entropy helped to distinguish muscle types. Hence, the combination of first-order and higher-order texture descriptors (Haralick, Galloway and LBP) can be used to discriminate gender and muscle types. Therefore, multi-texture analysis may be useful to investigate muscle damage and myopathic disorders.

Development of a hybrid image processing algorithm for automatic evaluation of intramuscular fat content in beef M. longissimus dorsi

An automatic method for estimating the content of intramuscular fat (IMF) in beef M. longissimus dorsi (LD) was developed using a sequence of image processing algorithm. To extract IMF particles within the LD muscle from structural features of intermuscular fat surrounding the muscle, three steps of image processing algorithm were developed, i.e. bilateral filter for noise removal, kernel fuzzy c-means clustering (KFCM) for segmentation, and vector confidence connected and flood fill for IMF extraction. The technique of bilateral filtering was firstly applied to reduce the noise and enhance the contrast of the beef image. KFCM was then used to segment the filtered beef image into lean, fat, and background. The IMF was finally extracted from the original beef image by using the techniques of vector confidence connected and flood filling. The performance of the algorithm developed was verified by correlation analysis between the IMF characteristics and the percentage of chemically extractable IMF content (P<0.05). Five IMF features are very significantly correlated with the fat content (P<0.001), including count densities of middle (CDMiddle) and large (CDLarge) fat particles, area densities of middle and large fat particles, and total fat area per unit LD area. The highest coefficient is 0.852 for CDLarge.

An incremental neural network for tissue segmentation in ultrasound images

This paper presents an incremental neural network (INeN) for the segmentation of tissues in ultrasound images. The performances of the INeN and the Kohonen network are investigated for ultrasound image segmentation. The elements of the feature vectors are individually formed by using discrete Fourier transform (DFT) and discrete cosine transform (DCT). The training set formed from blocks of 4x4 pixels (regions of interest, ROIs) on five different tissues designated by an expert is used for the training of the Kohonen network. The training set of the INeN is formed from randomly selected ROIs of 4x4 pixels in the image. Performances of both 2D-DFT and 2D-DCT are comparatively examined for the segmentation of ultrasound images.

Automatic measurement of dermal thickness from B-scan ultrasound images using active contours

BACKGROUND/PURPOSE

Measurement of dermal thickness is useful in the evaluation of dermo-cosmetics for assessing not only morphological changes but also mechanical properties of this layer. Our aim was first to standardise the manual dermal thickness measurement procedure on B-scan ultrasound images, then to develop an automatic operator independent method to detect the boundaries of the dermis.

MATERIAL AND METHODS

The Dermcup 20 MHz B-scan ultrasound system was used. The method used for detecting the boundaries was adapted from active contour algorithms. The innovative aspect of the method consists in an automatic initialization of the first step of the algorithm. To validate the method, we correlated measurements obtained by the manual and automatic approaches from a set of images from different anatomical sites.

RESULTS AND CONCLUSION

The results showed for the two measurement methods, 72% of the images were perfectly correlated. The remaining images required manual initialization of the boundaries by a non-expert operator before the active contour process could be used. Subsequent to this semi-automatic procedure, the correlation was very high.

Review of Methods for Texture Analysis of Myocardium From Echocardiographic Images: A Means of Tissue Characterization

This review discusses the definition of texture and identifies its utility in echocardiography for characterization of tissue. Methods used for quantification of texture in echocardiography and other disciplines are discussed. Several methodologies, particularly the wavelet method of texture quantification, seem to be promising. Image texture analysis appears to be a fertile area for research in echocardiography.

Echocardiographic texture analysis using the wavelet transform: differentiation of early heart muscle disease

Echocardiographic quantitation of myocardial texture for diagnosis of early cardiomyopathy (CMP) remains problematic. Conventional statistical methods are limited, contributed by a small image region-of-interest (ROI) and difficulty in discrimination from noise. This study was performed to evaluate the 2-D Haar wavelet decomposition method as a tool to identify textural changes in a rat model of early CMP, focusing on changes that occur before development of M-mode structural abnormalities. Early diabetic CMP, ethanol CMP and diabetic-ethanol CMP rat models were evaluated. Echocardiography was performed on two groups of rats. Group I (5 week cohort, n = 4 per subgroup) included controls, rats on 12% ethanol, a diabetic subgroup, and diabetic rats on 4% ethanol. Group II (10 week cohort, n = 5 per subgroup) included the same categories as group I with an additional subgroup taking 4% ethanol was also studied. M-mode left ventricular measurements were comparable in all subgroups of group I. However, diabetic rats in group II had an increased left ventricular dimension (LVD) compared to all others and an increased septal dimension (IVSD) and posterior wall dimension (PWD) were noted in the 4% and 12% ethanol groups. End-diastolic digital images of all rats in the parasternal short axis view, at the papillary muscle level, were downloaded to a computer. A 16 x 16 (ROI) was selected from the anterior interventricular septum. Although standard statistical methods could not differentiate any of the groups, calculation of textural energy and normalized textural energy with the 2-D Haar wavelet decomposition method found at 5 weeks increased normalized texture energy in diabetics compared to all others. At 10 weeks increased texture energy was noted in diabetics. Diabetic-ethanol rats at both 5 and 10 weeks revealed a blunted textural energy compared to diabetic rats. In a rat model of diabetic cardiomyopathy, the 2-D wavelet decomposition method identified textural energy changes before development of echocardiographic structural changes. Ethanol-associated blunting of textural changes in diabetic rats was also noted. This method for quantitation of ventricular texture may be relevant for diagnosis of early cardiomyopathy.

Breast cancer diagnosis using self-organizing map for sonography

The purpose of this study was to evaluate the performance of neural network model self-organizing maps (SOM) in the classification of benign and malignant sonographic breast lesions. A total of 243 breast tumors (82 malignant and 161 benign) were retrospectively evaluated. When a sonogram was performed, the analog video signal was captured to obtain a digitized sonographic image. The physician selected the region of interest in the sonography. An SOM model using 24 autocorrelation texture features classified the tumor as benign or malignant. In the experiment, cases were sampled with k-fold cross-validation (k = 10) to evaluate the performance using receiver operating characteristic (ROC) curves. The ROC area index for the proposed SOM system is 0.9357 +/- 0.0152, the accuracy is 85. 6%, the sensitivity is 97.6%, the specificity is 79.5%, the positive predictive value is 70.8%, and the negative predictive value is 98. 5%. This computer-aided diagnosis system can provide a useful tool and its high negative predictive value could potentially help avert benign biopsies.