Article

Explainable Ensemble Machine Learning for Breast

Cancer Diagnosis based on Ultrasound Image Texture
Features
Alireza Rezazadeh 1 , Yasamin Jafarian 2 and Ali Kord MD, MPH 3
1 Department of Electrical and Computer Engineering, University of Minnesota; rezaz003@[Link]
2 Department Computer Science and Engineering, University of Minnesota; yasamin@[Link]
3 Division of Interventional Radiology, Department of Radiology, University of Cincinnati; kordai@[Link]
arXiv:2201.07227v1 [[Link]] 17 Jan 2022

Abstract: 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.

1. Introduction
Ultrasound imaging is an effective method for breast cancer diagnosis [1,2] that, compared to
alternative modalities, is more accessible and less costly. Several recent studies have explored building
data-driven automated breast cancer diagnosis machine learning pipelines to detect the malignancy of
tumors observed in ultrasound images [3–6]. These studies dominantly rely on deep convolutional neural
network architectures to classify tumor images. Convolutional neural networks essentially learn to map
the input image pixel information to a lower-dimensional feature space through a series of hidden layers.
Although notable in prediction performance, convolutional networks are largely black-box machine
learning models that provide little to no insight into the logic behind their predictions [7,8]. In general,
humans tend to be unwilling to rely on procedures that are not interpretable, explainable and transparent,
especially for making critical predictions such as cancer diagnosis [8,9]. Yet, the explainability of machine
learning models for cancer diagnosis falls short of the increasing demand for interpretable and reliable
artificial intelligence [10].
In a recent study, Moon et al. [3] adopted standard deep convolutional neural network architectures
(including VGG, ResNet, and DenseNet) to classify breast ultrasound images and detect the malignancy of
tumors. The study reports a high predictive performance for these standard architectures. In a similar
study, Masud et al. [11] evaluated pretrained convolutional models for ultrasound image classification.
Other studies, such as [12,13], focused on semantic segmentation of the breast tumor from ultrasound
images.
In this paper, we propose an explainable machine learning pipeline for probabilistic breast cancer
diagnosis based on ultrasound images. We formulate this as a binary classification problem. First, through
 2 of 13

a comprehensive texture analysis, we extract first- and second-order texture features of the region of
interest in the ultrasound image. We then use these features to train an ensemble of decision trees. Each
decision tree in the model learns to classify the input image through a set of test conjunctions where each
test compares a texture feature with a robust numerical decision threshold. We show that our proposed
pipeline achieves a high predictive performance that is comparable to the existing black box convolution
neural network architectures. More importantly, we demonstrate that our proposed model can be probed
to accurately track and explain the decision path behind its prediction.

2. Materials and Methods

2.1. Data
We use a public dataset of breast ultrasound images [2]. In this dataset, a total number of 780 images
are obtained from 600 female patients (age of 25-75 years old). This includes 133 normal cases with no
mass, 210 cases with a benign mass, and 487 cases with a malignant mass. Images are obtained using
LOGIQ E9 ultrasound and LOGIQ E9 Agile ultrasound systems. These instruments produce DICOM
images with 1280 × 1024 resolution using 1-5 MHz transducers on ML6-15-D Matrix linear probe. The raw
DICOM images are cropped, preprocessed, and converted to PNG format with an average resolution of
500 × 500 pixels. For each case with a mass, a ground-truth binary mask of the region of interest (ROI) is
manually created. Dataset is split into 80% train set and 20% test set.

2.2. Texture Analysis

Texture features are important quantifiable metrics to characterize and describe a region of interest
in an image [14–16]. Texture feature analysis is typically measured using first-order and second-order
statistical metrics [17].

2.2.1. First-Order Statistics Texture Features

First-order texture features are computed based on first-order statistics of the one-dimensional gray
level histogram of the image. Therefore, a first-order texture feature does not take into account the pixel
neighborhood information [17,18]. In this study, we compute eight common first-order statics of the ROI
pixels: Mean, Variance, RMS, Energy, Entropy, Kurtosis, Skewness, and Uniformity. These statistical
metrics described in Table A1.

2.2.2. Second-Order Statistics Texture Features

The second-order statistical texture features are computed based on the gray-level co-occurrence
matrix (GLCM). GLCM elements are an estimation of the probability of transition from one gray level to
another along a certain pixel distance and direction [19–21]. We measure five common statistics based
on the GLCM computed for the ROI: Contrast, Dissimilarity, Homogeneity, Energy, and Correlation. A
description of these statistics are included in Table A2.

2.3. Decision Tree Models

A decision tree (DT) is a predictive model that consists of a set of test conjunctions, where each test
compares a feature of data with a numerical threshold [22]. Decision tree classification models are learned
by recursively partitioning the feature space to discover a set of robust decision rules [23,24]. One major
advantage of decision tree modeling is their interpretability. The set decision rules learned in a decision
tree model can be directly used to explain the logic behind the prediction of the model. Explainable
 3 of 13

predictions, as opposed to black-box predictions, can be used more reliably in applications such as medical
diagnosis.

2.3.1. Gradient Boosting Decision Tree

Gradient boosting decision tree (GBDT) is an ensemble of sequential decision tree models. GBDT
is frequently used in a variety of machine learning tasks due to its accuracy and efficiency. At each
boosting iteration, the ensemble learns a decision tree to predict the residual errors [22,25]. Specifically,
for a classification problem, let {( Xi , yi )}i=1...N be a dataset where for each entry, X and y correspond to
feature vector and the class label the entry belongs to. The goal is to approximate the function F̂ ( X ) = y
that learns to map feature vectors to their corresponding class label under an arbitrary differentiable loss
function L(yi , F ( Xi )), cross-entropy loss in our case.
With gradient boosting, at first the model is initialized with a constant,
n
F0 ( x ) = min ∑ L(yi , λ) (1)
λ i =1

Next, at each boosting iteration m, for each entry the residuals are computed as

ri,m = −∂L(y, F )/∂F | F= Fm−1 (2)

A decision tree with J terminal nodes is fitted to the residuals where

λ j,m = min ∑ L(y, Fm−1 ( x ) + λ) (3)

λ j

Finally, the boosted model is updated as:

Fm ( x ) = Fm−1 ( x ) + ν ∑ λ j,m I ( x ) (4)

J

Through grading boosting, GBDT combines multiple "weak" learner classifiers into an ensemble of
strong classification model.

2.3.2. LightGBM
LightGBM is an open-source GBDT framework [25]. LightGBM is based on a gradient-based one-side
sampling to filter data instances and an exclusive feature bundling to encode features into less dense space.
Specifically, LightGBM discretizes continuous features using a histogram-based algorithm for a faster
training process and reduced memory consumption. Also, LightGBM uses a leaf-wise strategy of growing
decision trees by discovering a leaf with the highest gain of variance. This enables LightGBM to achieve
state-of-the-art performance in a variety of applications [26–28].

2.4. Machine Learning Diagnosis Pipeline

In this paper, we propose an interpretable machine learning pipeline for breast cancer diagnosis based
on ultrasound images. We formulate this as a binary classification problem where class labels belong to
{benign, malignant}. As shown in Figure 1, the pipeline takes as input the ultrasound image of the mass
and a binary mask of the ROI. We use this binary mask to crop and compute the GLCM of the ROI. This
yields a total number of 60 texture features (5 GLCM statistics ×3 distances ×4 angles). These texture
 4 of 13

Figure 1. Our proposed pipeline uses the ultrasound image and ROI mask to extract GLCM texture features
and learns a boosted decision tree model to make probabilistic diagnosis based on explainable decision
trees.

features are denoted as {statistic}_d{distance}_a{angle} (e.g., energy_d3_a135 refers to the GLCM energy
computed within a distance of 3 pixels and along a 135◦ direction).
We then use a LightGBM classification model with a gradient boosting decision tree strategy, 10
leaves per tree, and a maximum feature bin size of 512. The classification model is trained by minimizing a
binary log loss, with a learning rate of 0.05, and for a total number of 500 boosting iterations. For a given
ultrasound image input, the model outputs the probability of the mass in ROI being benign or malignant.
Importantly, the decision tree ensemble can be decomposed to explain how the model comes up with a
prediction. The learned model is a set of decision trees with multiple test conjunctions that compare the
texture features of the ROI with numerical thresholds inferred from the data.

3. Results
In this section, we first summarize the texture analysis results and then evaluate the performance
of our purposed pipeline. Lastly, we highlight how our pipeline can be used as an explainable machine
learning framework to understand the logic behind each of its diagnostic predictions.

3.1. Texture Analysis Statistical Analysis

We perform texture analysis by computing the ROI first- and second-order statistics (see Table A1
and A2 for mathematical descriptions). A standard t-test is used to compare each texture feature between
the two groups of benign and malignant masses.
Most first-order statistic texture features are significantly different for the two sets of benign and
malignant masses (p < 0.001). Mean (p = 0.43) and RMS (p = 0.28) first-order statistics, however, are not
significantly different across the groups. Figure 2 demonstrates the first-order texture features comparison.
All t-test results of this comparison are thoroughly reported in Table A1.
Second-order statics based on GLCM are computed across 3 distances {1, 3, 5} pixels and 4 angles
{0 , 45◦ , 90◦ , 135◦ }. All GLCM features are significantly different between benign and malignant groups
◦

(p < 0.001). See Tables A2-A7 for the complete t-test results. Interestingly, the difference between the
two groups is consistent for various distances and angles. This further indicates that GLCM features are
consistent for different orientations of the ROI. Figure 3 exemplifies the persistence of the difference in
each GLCM features across all angles for d = 3 pixels.
 5 of 13

Figure 2. First-order texture features. All measured statistics, except Mean and RMS, are significantly
different between the two groups of benign and malignant masses. Refer to Table A1 for a description of
the metrics and t-test results.

Figure 3. Second-order GLCM texture features. All measure statistics are significantly different between
the benign and malignant groups. This significant difference is consistent across various pixel distances
(d = {1, 3, 5}, only d=3 results are demonstrated here) and angles (θ = {0◦ , 45◦ , 90◦ , 135◦ }). Refer to Tables
A2-A7 for the deception of each metric and detailed t-test results.
 6 of 13

Table 1. Model evaluation. Standard classification performance metrics measured on the test set. The best
performance is achieved with LightGBM model with 500 gradient boosting iterations.

Model Precision Recall F1-score AUC Accuracy

DT 0.85 0.82 0.83 0.82 0.86
LightGBM-10 0.90 0.80 0.83 0.79 0.87
LightGBM-50 0.88 0.87 0.88 0.90 0.87
LightGBM-100 0.93 0.91 0.92 0.92 0.90
LightGBM-500 0.94 0.93 0.93 0.93 0.91

Table 2. Model comparison with convolutional architectures. Standard classification performance metrics
measured on the test set. Our explainable model based on decision trees achieves high predictive
performance that is comparable to existing black box convolutional neural network architectures.

Model Precision Recall F1-score AUC Accuracy

VGG 0.75 0.76 0.76 0.87 0.85
ResNet 0.89 0.89 0.89 0.96 0.91
DenseNet 0.90 0.92 0.91 0.97 0.94
Decision Tree Ensemble (ours) 0.94 0.93 0.93 0.93 0.91

3.2. Machine Learning Diagnosis Pipeline Performance

In this section, we compare multiple variations of our framework. The comparison is made using
common standard statistical metrics for evaluating classification performance on the test set. Specifically,
we report: Accuracy, Precision, Recall, F1-score, and Area under the ROC Curve (AUC) (see [29] for
mathematical descriptions of these metrics).
Using LightGBM algorithm with 500 iterations of gradient boosting, our pipeline reaches its best
performance with 0.91 accuracy, 0.94 precision, 0.93 recall, and 0.93 F1-score (Table 1). With smaller number
of boosting iterations, the accuracy slightly drops down. Also, compared to a simple DT, using LightGBM
significantly improved the performance of our pipeline. This emphasizes the importance of gradient
boosting approaches in our pipeline.
We also compare our model’s performance with the standard convolutional neural network
architectures from Moon et al. [3] which is also using a similar dataset to ours [2]. Our pipeline, based
on decision tree ensembles, achieves comparable results to the convolutional network models (Table 2)
while being explainable. Our model can be probed to trace the logic behind its predictions while the
convolutional models do not provide any insight into the process behind their predictions. In comparison,
our model achieves higher precision, recall, and F1-score. Note that F1-score is more suitable to assess
models for their classification performance on imbalanced datasets.
To further identify the most important texture features for the model, we quantify the feature
importance with SHAP values [30]. The SHAP values are Shapley values from coalitional game theory
and correspond to the magnitude of each feature’s attribution on the output of the model. With our
dataset, the most important features are: GLCM correlation within 3 pixels along the 90◦ direction, GLCM
energy within a 5 pixels distance along the 90◦ direction, GLCM energy within a 3 pixels distance along
the 90◦ direction, and GLCM correlation within a 5 pixels distance along the 90◦ direction (see Figure 4).
Interestingly, all top 4 important features are statistics measured in the 90◦ direction.
 7 of 13

Figure 4. Feature importance. SHAP value quantifies the contribution of each feature to the prediction of
each class. The features are denoted as {statistic}_d{distance}_a{angle}.

3.3. Explainability of the Machine Learning Diagnosis Pipeline

In this section, we give an overview of how the explainability of our pipeline enables tracking down
the logic behind its predictions. First, we evaluate a benign case from the test set (Figure 5). Given the
ultrasound image and the cropped ROI, the learned model predicts the mass in ROI is a benign with a
likelihood of 0.97. We can further decompose the learned model to understand the decision path that leads
to this prediction. The learned model is an ensemble of multiple decision trees (T = 500).
In this example (Figure 5), two decision trees from the ensemble are depicted. In Tree Classifier 1, the
first nodes splits the data based on correlation_d3_a90. This texture feature is measured as 0.36 in the
cropped ROI which, in comparison with the learned split threshold 0.78, determines the outcome of the split
at this node. The histogram of nodes also compares the split threshold with the distribution of the texture
feature for both malignant and benign cases in the train set. At the next node energy_d5_a0 = 0.23 < 0.26
which leads to the final node dissimilarity_d1_a0 = 4.09 < 5.88. Based on this decision path the model
predicts the ROI belongs to the benign class. Other decision trees in the ensemble can also be interpreted
in a similar way.
Next, we evaluate a malignant case from the test set (Figure 6). For this input, the model predicts
the mass in ROI is malignant with a likelihood of 0.93. In Tree Classifier 1, the model’s decision path
begins with a split based on correlation_d3_a90 = 0.86 > 0.78. This is followed by the final node
energy_d5_a90 = 0.31 > 0.30 which predicts the ROI mass is malignant.

4. Conclusions
Ultrasound imaging is an accessible and cost-effective imaging modality to diagnose breast cancer.
Most of recent work on building machine learning models for breast cancer diagnosis depend on
convolutional neural network architectures. Although accurate in performance, convolutional networks
are black-box models and cannot be interpreted in terms of the logic behind their predictions.
In this paper, we proposed a novel explainable machine learning pipeline for breast cancer diagnosis
based on ultrasound images. Our pipeline uses texture analysis of the ultrasound images as its basis to
learn an ensemble of decision trees to predict the likelihood of malignancy of breast tumors. Importantly,
our model can be decomposed into its underlying decision trees to fully interpret the decision path behind
its outputs by following test conjunctions in each decision tree.
 8 of 13

Figure 5. Qualitative results of a benign case. Our pipeline learns to infer a probabilistic diagnosis of the
breast ultrasound images. The learned ensemble can be probed to obtain explainable decision paths in a
set of learned decision trees. In each learned tree classifier, the orange arrows highlight the decision path.
The model learns to compare the texture features obtained from the input image (orange numbers at the
bottom of each dashed box) with the learned thresholds (black triangle on each histogram) at each node of
the decision tree.
 9 of 13

Figure 6. Qualitative results of a malignant case. Refer to Figure 5 caption for more details.
 10 of 13

We believe our work is a step towards building more practical and comprehensible machine learning
tools for cancer diagnosis by increasing the transparency of the prediction process. An interesting future
work is to combine convolutional networks with decision trees. Finally, we hope our approach in this
work to inspire future research on data-driven medical diagnosis to devote more attention into increasing
the explainability of their solutions.

Appendix A
This section gives an overview of the first- and second-order GLCM texture feature statistics along
with a detailed report of the t-test of each statistic between the two group of benign and malignant masses.

Table A1. First-order Texture Features. The metrics are computed for a set of N pixels inside of the ROI
denoted as X. Benign and malignant groups are compared using a t-test.

Feature Description µbenign µmalignant p-value

energy ∑iN=1 (X( i ) + c )2 6.66e+07 1.17e+08 <0.001
N
− ∑ p(i ) log2 p(i ) + e


entropy 1.98 2.23 <0.001
i =1
1 N
N ∑i =1 (X(i )− X̄ )
4
kurtosis 1 N 2 6.50 4.27 <0.001
( N ∑i =1 (X(i )− X̄ ) )
2

N
mean 1
N ∑ X( i ) 6.40e+01 6.59e+01 0.43
q i =1
1 N
rms ∑
N i =1 (X ( i ) + c )
2 7.26e+01 7.51e+01 0.28
1 N
∑i=1 (X(i )− X̄ ) 3
skewness qN 3 1.45 9.92e-01 <0.001
1 N
N ∑i =1 (X(i )− X̄ )
2

N
uniformity ∑ p ( i )2 3.40e-01 2.69e-01 <0.001
i =1
N
variance 1
N ∑ (X(i) − X̄ )2 1.04e+03 1.22e+03 <0.001
i =1

Table A2. Second-order Texture Features. The metrics are computed based on the gray-level co-occurrence
matrix for a set of N gray level of pixels inside of the ROI.

GLCM Feature Description

N N
contrast ∑ ∑ (i − j)2 p(i, j)
i =1 j =1
energy ∑iN=1 (X(i ) + c)2
N N
∑i=1 ∑ j=1 p(i,j)ij−µ x µy
correlation σx (i )σy ( j)
N N
dissimilarity ∑ ∑ |i − j| p(i, j)
i =1 j =1
N N
2
energy ∑∑ p(i, j)
i =1 j =1
N N
2
energy ∑∑ p(i, j)
i =1 j =1
N N
p(i, j)
homogeneity ∑ ∑ 1 + |i − j |
i =1 j =1
 11 of 13

Table A3. Gray level co-occurrence matrix contrast feature measured across three pixel distances {1, 3, 5}
and four angles {0◦ , 45◦ , 90◦ , 135◦ }.
GLCM contrast
distance angle µbenign µmalignant p-value
1 0° 184.86 114.20 <0.001
45° 433.20 238.06 <0.001
90° 304.20 154.26 <0.001
135° 438.01 238.07 <0.001
3 0° 730.45 493.19 <0.001
45° 1032.13 582.21 <0.001
90° 1229.20 660.83 <0.001
135° 1049.76 582.42 <0.001
5 0° 1136.95 830.26 <0.001
45° 1768.76 1132.49 <0.001
90° 1681.63 1040.57 <0.001
135° 1801.47 1130.44 <0.001

Table A4. Gray level co-occurrence matrix correlation feature measured across three pixel distances {1, 3, 5}
and four angles {0◦ , 45◦ , 90◦ , 135◦ }.
GLCM correlation
distance angle µbenign µmalignant p-value
1 0° 0.94 0.97 <0.001
45° 0.87 0.94 <0.001
90° 0.91 0.96 <0.001
135° 0.86 0.94 <0.001
3 0° 0.77 0.87 <0.001
45° 0.68 0.85 <0.001
90° 0.63 0.83 <0.001
135° 0.68 0.85 <0.001
5 0° 0.64 0.78 <0.001
45° 0.46 0.70 <0.001
90° 0.47 0.73 <0.001
135° 0.44 0.70 <0.001

Table A5. Gray level co-occurrence matrix dissimilarity feature measured across three pixel distances
{1, 3, 5} and four angles {0◦ , 45◦ , 90◦ , 135◦ }.
GLCM dissimilarity
distance angle µbenign µmalignant p-value
1 0° 4.84 3.41 <0.001
45° 8.67 5.75 <0.001
90° 7.09 4.58 <0.001
135° 8.69 5.74 <0.001
3 0° 12.10 8.84 <0.001
45° 15.33 10.07 <0.001
90° 17.04 10.97 <0.001
135° 15.40 10.06 <0.001
5 0° 16.76 12.76 <0.001
45° 23.10 15.95 <0.001
90° 22.06 14.98 <0.001
135° 23.18 15.90 <0.001
 12 of 13

Table A6. Gray level co-occurrence matrix energy feature measured across three pixel distances {1, 3, 5}
and four angles {0◦ , 45◦ , 90◦ , 135◦ }.
GLCM energy
distance angle µbenign µmalignant p-value
1 0° 0.30 0.38 <0.001
45° 0.28 0.37 <0.001
90° 0.29 0.38 <0.001
135° 0.28 0.37 <0.001
3 0° 0.26 0.36 <0.001
45° 0.24 0.35 <0.001
90° 0.25 0.35 <0.001
135° 0.24 0.35 <0.001
5 0° 0.24 0.34 <0.001
45° 0.19 0.32 <0.001
90° 0.22 0.33 <0.001
135° 0.19 0.32 <0.001

Table A7. Gray level co-occurrence matrix homogeneity feature measured across three pixel distances
{1, 3, 5} and four angles {0◦ , 45◦ , 90◦ , 135◦ }.
GLCM homogeneity
distance angle µbenign µmalignant p-value
1 0° 0.49 0.56 <0.001
45° 0.41 0.49 <0.001
90° 0.43 0.51 <0.001
135° 0.41 0.50 <0.001
3 0° 0.37 0.45 <0.001
45° 0.33 0.44 <0.001
90° 0.33 0.43 <0.001
135° 0.33 0.44 <0.001
5 0° 0.33 0.41 <0.001
45° 0.26 0.32 <0.001
90° 0.30 0.40 <0.001
135° 0.27 0.39 <0.001

References
1. Kuhl, C.K.; Schrading, S.; Leutner, C.C.; Morakkabati-Spitz, N.; Wardelmann, E.; Fimmers, R.; Kuhn, W.; Schild,
H.H. Mammography, breast ultrasound, and magnetic resonance imaging for surveillance of women at high
familial risk for breast cancer. Journal of clinical oncology 2005, 23, 8469–8476.
2. Al-Dhabyani, W.; Gomaa, M.; Khaled, H.; Fahmy, A. Dataset of breast ultrasound images. Data in brief 2020,
28, 104863.
3. Moon, W.K.; Lee, Y.W.; Ke, H.H.; Lee, S.H.; Huang, C.S.; Chang, R.F. Computer-aided diagnosis of breast
ultrasound images using ensemble learning from convolutional neural networks. Computer methods and
programs in biomedicine 2020, 190, 105361.
4. Samulski, M.; Hupse, R.; Boetes, C.; Mus, R.D.; den Heeten, G.J.; Karssemeijer, N. Using computer-aided
detection in mammography as a decision support. European radiology 2010, 20, 2323–2330.
5. Sahiner, B.; Chan, H.P.; Roubidoux, M.A.; Hadjiiski, L.M.; Helvie, M.A.; Paramagul, C.; Bailey, J.; Nees, A.V.;
Blane, C. Malignant and benign breast masses on 3D US volumetric images: effect of computer-aided diagnosis
on radiologist accuracy. Radiology 2007, 242, 716–724.
6. Jiménez-Gaona, Y.; Rodríguez-Álvarez, M.J.; Lakshminarayanan, V. Deep-Learning-Based Computer-Aided
Systems for Breast Cancer Imaging: A Critical Review. Applied Sciences 2020, 10, 8298.
7. Castelvecchi, D. Can we open the black box of AI? Nature News 2016, 538, 20.
 13 of 13

8. Arrieta, A.B.; Díaz-Rodríguez, N.; Del Ser, J.; Bennetot, A.; Tabik, S.; Barbado, A.; García, S.; Gil-López, S.;
Molina, D.; Benjamins, R.; others. Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities
and challenges toward responsible AI. Information Fusion 2020, 58, 82–115.
9. Zhu, J.; Liapis, A.; Risi, S.; Bidarra, R.; Youngblood, G.M. Explainable AI for designers: A human-centered
perspective on mixed-initiative co-creation. 2018 IEEE Conference on Computational Intelligence and Games
(CIG). IEEE, 2018, pp. 1–8.
10. Preece, A.; Harborne, D.; Braines, D.; Tomsett, R.; Chakraborty, S. Stakeholders in explainable AI. arXiv preprint
arXiv:1810.00184 2018.
11. Masud, M.; Rashed, A.E.E.; Hossain, M.S. Convolutional neural network-based models for diagnosis of breast
cancer. Neural Computing and Applications 2020, pp. 1–12.
12. Byra, M.; Jarosik, P.; Szubert, A.; Galperin, M.; Ojeda-Fournier, H.; Olson, L.; O’Boyle, M.; Comstock, C.;
Andre, M. Breast mass segmentation in ultrasound with selective kernel U-Net convolutional neural network.
Biomedical Signal Processing and Control 2020, 61, 102027.
13. Irfan, R.; Almazroi, A.A.; Rauf, H.T.; Damaševičius, R.; Nasr, E.A.; Abdelgawad, A.E. Dilated semantic
segmentation for breast ultrasonic lesion detection using parallel feature fusion. Diagnostics 2021, 11, 1212.
14. Tuceryan, M.; Jain, A.K. Texture analysis. Handbook of pattern recognition and computer vision 1993, pp. 235–276.
15. Materka, A.; Strzelecki, M.; others. Texture analysis methods–a review. Technical university of lodz, institute of
electronics, COST B11 report, Brussels 1998, 10, 4968.
16. Varghese, B.A.; Cen, S.Y.; Hwang, D.H.; Duddalwar, V.A. Texture analysis of imaging: what radiologists need
to know. American Journal of Roentgenology 2019, 212, 520–528.
17. Srinivasan, G.; Shobha, G. Statistical texture analysis. Proceedings of world academy of science, engineering
and technology, 2008, Vol. 36, pp. 1264–1269.
18. Kim, N.D.; Amin, V.; Wilson, D.; Rouse, G.; Udpa, S. Ultrasound image texture analysis for characterizing
intramuscular fat content of live beef cattle. Ultrasonic imaging 1998, 20, 191–205.
19. Sebastian V, B.; Unnikrishnan, A.; Balakrishnan, K. Gray level co-occurrence matrices: generalisation and some
new features. arXiv preprint arXiv:1205.4831 2012.
20. Iqbal, F.; Pallewatte, A.S.; Wansapura, J.P. Texture analysis of ultrasound images of chronic kidney disease.
2017 Seventeenth International Conference on Advances in ICT for Emerging Regions (ICTer). IEEE, 2017, pp.
1–5.
21. Xu, S.S.D.; Chang, C.C.; Su, C.T.; Phu, P.Q. Classification of liver diseases based on ultrasound image texture
features. Applied Sciences 2019, 9, 342.
22. Sharma, H.; Kumar, S. A survey on decision tree algorithms of classification in data mining. International
Journal of Science and Research (IJSR) 2016, 5, 2094–2097.
23. Myles, A.J.; Feudale, R.N.; Liu, Y.; Woody, N.A.; Brown, S.D. An introduction to decision tree modeling. Journal
of Chemometrics: A Journal of the Chemometrics Society 2004, 18, 275–285.
24. Safavian, S.R.; Landgrebe, D. A survey of decision tree classifier methodology. IEEE transactions on systems,
man, and cybernetics 1991, 21, 660–674.
25. Ke, G.; Meng, Q.; Finley, T.; Wang, T.; Chen, W.; Ma, W.; Ye, Q.; Liu, T.Y. Lightgbm: A highly efficient gradient
boosting decision tree. Advances in neural information processing systems 2017, 30, 3146–3154.
26. Rezazadeh, A. A Generalized Flow for B2B Sales Predictive Modeling: An Azure Machine-Learning Approach.
Forecasting 2020, 2, 267–283.
27. Chen, C.; Zhang, Q.; Ma, Q.; Yu, B. LightGBM-PPI: Predicting protein-protein interactions through LightGBM
with multi-information fusion. chemometrics and intelligent laboratory systems 2019, 191, 54–64.
28. Sun, X.; Liu, M.; Sima, Z. A novel cryptocurrency price trend forecasting model based on LightGBM. Finance
Research Letters 2020, 32, 101084.
29. Hossin, M.; Sulaiman, M.N. A review on evaluation metrics for data classification evaluations. International
journal of data mining & knowledge management process 2015, 5, 1.
30. Lundberg, S.M.; Erion, G.G.; Lee, S.I. Consistent individualized feature attribution for tree ensembles. arXiv
preprint arXiv:1802.03888 2018.