Low-contrast Images Research Articles

A mathematical and neural network-based hybrid technique for detecting the prostate contour from medical image data

Extraction of prostate contour on transrectal ultrasound (TRUS) is significant in prostate cancer diagnosis. With the low-contrast images of TRUS and the existence of imaging artifacts, detecting the prostate accurately from TRUS images becomes more and more difficult. In this paper, we present four strategies to achieve improved segmentation precision on difficult images. First, to achieve the data sequence, we adopt a variant of the principal curve-based algorithm, where a small part of prior point information is utilized as the approximate initialization. Secondly, we devise an evolution neural network to assist in finding the optimal neural network. Thirdly, we train the fractional-order-based neural network using data sequence as input, thereby decreasing the model error and increasing the precision of results during training. Finally, a smooth and explainable mathematical function of organ boundary is represented by the parameters of a fractional-order-based neural network. The performance of our method was evaluated on two clinical datasets with 1,338 TRUS prostate images using Accuracy (ACC), Dice similarity coefficient (DSC), and Jaccard similarity coefficient (OMG). Results show that our model 1) achieves better performance than the state-of-the-art segmentation models; the corresponding ACC, DSC, and OMG of our method are 95.3 ± 2.2 %, 95.9 ± 2.3 %, and 94.9 ± 2.4 %, respectively, and 2) detects unseen or vague regions satisfactorily.

Fish Monitoring from Low-Contrast Underwater Images

A toolset supporting fish detection, orientation, tracking and especially morphological feature estimation with high speed and accuracy, is presented in this paper. It can be exploited in fish farms to automate everyday procedures including size measurement and optimal harvest time estimation, fish health assessment, quantification of feeding needs, etc. It can also be used in an open sea environment to monitor fish size, behavior and the population of various species. An efficient deep learning technique for fish detection is employed and adapted, while methods for fish tracking are also proposed. The fish orientation is classified in order to apply a shape alignment technique that is based on the Ensemble of Regression Trees machine learning method. Shape alignment allows the estimation of fish dimensions (length, height) and the localization of fish body parts of particular interest such as the eyes and gills. The proposed method can estimate the position of 18 landmarks with an accuracy of about 95% from low-contrast underwater images where the fish can be hardly distinguished from its background. Hardware and software acceleration techniques have been applied at the shape alignment process reducing the frame processing latency to less than 0.5 us on a general purpose computer and less than 16 ms on an embedded platform. As a case study, the developed system has been trained and tested with several Mediterranean fish species in the category of seabream. A large public dataset with low-resolution underwater videos and images has also been developed to test the proposed system under worst case conditions.

Multi-branch Context Awareness Network for Prostate MRI Segmentation

Prostate image segmentation is a precondition for the diagnosis of prostate diseases and subsequent treatment. However, the blurry organ edges and low image contrast make accurate segmentation difficult. In order to overcome the difficulties, this paper proposes an innovative Multi-branch Context Awareness Network (MBCA-Net) for prostate MRI segmentation. MBCA-Net uses the 3D UNet as the backbone network with an encoder-decoder framework. To improve the feature extraction capacity of the network, a multi-branch residual module with different convolution kernels is proposed. To better make use of extracted features and reduce semantic ambiguity, a context awareness module is used to fuse low-level features in the encoder and high-level features in the decoder in a full process manner. To improve the stability of the model during training, a mixed loss function is adopted. Experiments on the public dataset PROMISE12 show that our model’s Dice coefficient is 90.19%. Compared with other advanced techniques, the proposed model has better segmentation results.

Longitudinal mapping of the development of cortical thickness and surface area in rhesus macaques during the first three years

Studying dynamic spatiotemporal patterns of early brain development in macaque monkeys is critical for understanding the cortical organization and evolution in humans, given the phylogenetic closeness between humans and macaques. However, due to huge challenges in the analysis of early brain Magnetic Resonance Imaging (MRI) data typically with extremely low contrast and dynamic imaging appearances, our knowledge of the early macaque cortical development remains scarce. To fill this critical gap, this paper characterizes the early developmental patterns of cortical thickness and surface area in rhesus macaques by leveraging advanced computing tools tailored for early developing brains based on a densely sampled longitudinal dataset with 140 rhesus macaque MRI scans seamlessly covering from birth to 36 mo of age. The average cortical thickness exhibits an inverted U-shaped trajectory with peak thickness at around 4.3 mo of age, which is remarkably in line with the age of peak thickness at 14 mo in humans, considering the around 3:1 age ratio of human to macaque. The total cortical surface area in macaques increases monotonically but with relatively lower expansions than in humans. The spatial distributions of thicker and thinner regions are quite consistent during development, with gyri having a thicker cortex than sulci. By 4 mo of age, over 81% of cortical vertices have reached their peaks in thickness, except for the insula and medial temporal cortices, while most cortical vertices keep expanding in surface area, except for the occipital cortex. These findings provide important insights into early brain development and evolution in primates.

A Novel Light U-Net Model for Left Ventricle Segmentation Using MRI

MRI segmentation and analysis are significant tasks in clinical cardiac computations. A cardiovascular MR scan with left ventricular segmentation seems necessary to diagnose and further treat the disease. The proposed method for left ventricle segmentation works as a combination of the intelligent histogram-based image enhancement technique with a Light U-Net model. This technique serves as the basis for choosing the low-contrast image subjected to the stretching technique and produces sharp object contours with good contrast settings for the segmentation process. After enhancement, the images are subjected to the encoder–decoder configuration of U-Net using a novel lightweight processing model. Encoder sampling is supported by a block of three parallel convolutional layers with supporting functions that improve the semantics for segmentation at various levels of resolutions and features. The proposed method finally increased segmentation efficiency, extracting the most relevant image resources from depth-to-depth convolutions, filtering them through each network block, and producing more precise resource maps. The dataset of MICCAI 2009 served as an assessment tool of the proposed methodology and provides a dice coefficient value of 97.7%, accuracy of 92%, and precision of 98.17%.

Xray-Net: Self-supervised pixel stretching approach to improve low-contrast medical imaging

Medical imaging is crucial for monitoring the progression of a diagnosis. However, poor-quality images, particularly radiographs which are commonly used for their affordability, can lack sufficient visual information and lead to lengthy and inaccurate diagnoses, reducing clinicians’ ability to examine and decreasing the accuracy of their assumptions. To address this issue, this paper proposes an Xray-Net approach that utilizes self-supervised pixel-contrast stretching to improve low-contrast images. The approach consists of three main phases: (1) Pre-processing, which involves obtaining X-ray contrast-based classification through the measurement of average brightness, (2) Development of a conventional neural network model, and (3) Contrast-adaptive stretching, where the stretching control coefficients are predicted adaptively based on input image information. The proposed approach was evaluated using quantitative and qualitative experiments on three X-ray databases, and six different quality measures were considered. The results obtained by Xray-Net demonstrate superior low-contrast enhancement performance compared to several state-of-the-art methods.

Creating patient-specific digital phantoms with a longitudinal atlas for evaluating deformable CT-CBCT registration in adaptive lung radiotherapy.

Quality assurance of deformable image registration (DIR) is challenging because the ground truth is often unavailable. In addition, current approaches that rely on artificial transformations do not adequately resemble clinical scenarios encountered in adaptive radiotherapy. We developed an atlas-based method to create a variety of patient-specific serial digital phantoms with CBCT-like image quality to assess the DIR performance for longitudinal CBCT imaging data in adaptive lung radiotherapy. A library of deformations was created by extracting the longitudinal changes observed between a planning CT and weekly CBCT from an atlas of lung radiotherapy patients. The planning CT of an inquiry patient was first deformed by mapping the deformation pattern from a matched atlas patient, and subsequently appended with CBCT artifacts to imitate a weekly CBCT. Finally, a group of digital phantoms around an inquiry patient was produced to simulate a series of possible evolutions of tumor and adjacent normal structures. We validated the generated deformation vector fields (DVFs) to ensure numerically and physiologically realistic transformations. The proposed framework was applied to evaluate the performance of the DIR algorithm implemented in the commercial Eclipse treatment planning system in a retrospective study of eight inquiry patients. The generated DVFs were inverse consistent within less than 3mm and did not exhibit unrealistic folding. The deformation patterns adequately mimicked the observed longitudinal anatomical changes of the matched atlas patients. Worse Eclipse DVF accuracy was observed in regions of low image contrast or artifacts. The structure volumes exhibiting a DVF error magnitude of equal or more than 2mm ranged from 24.5% (spinal cord) to 69.2% (heart) and the maximum DVF error exceeded 5mm for all structures except the spinal cord. Contour-based evaluations showed a high degree of alignment with dice similarity coefficients above 0.8 in all cases, which underestimated the overall DVF accuracy within the structures. It is feasible to create and augment digital phantoms based on a particular patient of interest using multiple series of deformation patterns from matched patients in an atlas. This can provide a semi-automated procedure to complement the quality assurance of CT-CBCT DIR and facilitate the clinical implementation of image-guided and adaptive radiotherapy that involve longitudinal CBCT imaging studies.

Improved YOLOv4-tiny Target Detection Method Based on Adaptive Self-Order Piecewise Enhancement and Multiscale Feature Optimization

To improve the accuracy of material identification under low contrast conditions, this paper proposes an improved YOLOv4-tiny target detection method based on an adaptive self-order piecewise enhancement and multiscale feature optimization. The model first constructs an adaptive self-rank piecewise enhancement algorithm to enhance low-contrast images and then considers the fast detection ability of the YOLOv4-tiny network. To make the detection network have a higher accuracy, this paper adds an SE channel attention mechanism and an SPP module to this lightweight backbone network to increase the receptive field of the model and enrich the expression ability of the feature map. The network can pay more attention to salient information, suppress edge information, and effectively improve the training accuracy of the model. At the same time, to better fuse the features of different scales, the FPN multiscale feature fusion structure is redesigned to strengthen the fusion of semantic information at all levels of the network, enhance the ability of network feature extraction, and improve the overall detection accuracy of the model. The experimental results show that compared with the mainstream network framework, the improved YOLOv4-tiny network in this paper effectively improves the running speed and target detection accuracy of the model, and its mAP index reaches 98.85%, achieving better detection results.

Self-Guided Optimization Semi-Supervised Method for Joint Segmentation of Macular Hole and Cystoid Macular Edema in Retinal OCT Images.

Macular hole(MH) and cystoid macular edema(CME) are two common retinal pathologies that cause vision loss. Accurate segmentation of MH and CME in retinal OCT images can greatly aid ophthalmologists to evaluate the relevant diseases. However, it is still challenging as the complicated pathological features of MH and CME in retinal OCT images, such as the diversity of morphologies, low imaging contrast, and blurred boundaries. In addition, the lack of pixel-level annotation data is one of the important factors that hinders the further improvement of segmentation accuracy. Focusing on these challenges, we propose a novel self-guided optimization semi-supervised method termed Semi-SGO for joint segmentation of MH and CME in retinal OCT images. Aiming to improve the model's ability to learn the complicated pathological features of MH and CME, while alleviating the feature learning tendency problem that may be caused by the introduction of skip-connection in U-shaped segmentation architecture, we develop a novel dual decoder dual-task fully convolutional neural network(D3T-FCN). Meanwhile, based on our proposed D3T-FCN, we introduce a knowledge distillation technique to further design a novel semi-supervised segmentation method called Semi-SGO, which can leverage unlabeled data to further improve the segmentation accuracy. Comprehensive experimental results show that our proposed Semi-SGO outperforms other state-of-the-art segmentation networks. Furthermore, we also develop an automatic method for measuring the clinical indicators of MH and CME to validate the clinical significance of our proposed Semi-SGO. The code will be released on Github.

A Hybrid Method for Enhancement of Both Contrast Distorted and Low-Light Images

Many different histogram equalization (HE)-based image enhancement methods have been developed to overcome the problems of low or high image brightness, contrast sensitivity, and difficulty in revealing details of dark areas under low-light environments. In this paper, a novel image enhancement method based on HE and adaptive gamma correction with weight distribution (AGCWD) is proposed for natural and effective image enhancement. In this method, histogram stretching is performed on Red–Green–Blue (RGB) color components of image, and then the color space of RGB image is converted to Hue–Saturation–Intensity (HSI) color space. The histograms of S and I components are divided into sub-histograms according to the exposure threshold. The underexposure regions are enhanced with a new AGCWD. Then, the color space of HSI image is converted to RGB color space. Finally, the HE is applied to the input image with the obtained image histogram map. Thus, the method has not only effectively prevented the over-enhancement of the contrast but also obtained the quality and natural enhanced image. The proposed method is compared with the most known contrast enhancement methods and low-light enhancement methods. Experimental results have supported that the proposed method outperforms other methods in terms of both visual perception and objective evaluation.

DDeep3M+: adaptive enhancement powered weakly supervised learning for neuron segmentation.

Robust segmentations of neurons greatly improve neuronal population reconstruction, which could support further study of neuron morphology for brain research. Precise segmentation of 3D neuron structures from optical microscopy (OM) images is crucial to probe neural circuits and brain functions. However, the high noise and low contrast of images make neuron segmentation challenging. Convolutional neural networks (CNNs) can provide feasible solutions for the task but they require large manual labels for training. Labor-intensive labeling is highly expensive and heavily limits the algorithm generalization. We devise a weakly supervised learning framework Docker-based deep network plus (DDeep3M+) for neuron segmentation without any manual labeling. A Hessian analysis based adaptive enhancement filter is employed to generate pseudo-labels for segmenting neuron images. The automated segmentation labels are input for training a DDeep3M to extract neuronal features. We mine more undetected weak neurites from the probability map based on neuronal structures, thereby modifying the pseudo-labels. We iteratively refine the pseudo-labels and retrain the DDeep3M model with the pseudo-labels to obtain a final segmentation result. The proposed method achieves promising results with the score of 0.973, which is close to that of the CNN model with manual labels and superior to several segmentation algorithms. We propose an accurate weakly supervised neuron segmentation method. The high precision results achieved on 3D OM datasets demonstrate the superior generalization of our DDeep3M+.

A new X-ray images enhancement method using a class of fractional differential equation

Many image-processing applications heavily depend on the quality of medical images. Due to the unpredictable variation in the captured images, medical images frequently have problems with noise or low contrast; therefore, improving medical imaging is a challenging task. For better treatment, physicians need images with good contrast to provide the most detailed picture of the disease. The generalized k-differential equation based on the k-Caputo fractional differential operator (K-CFDO) is used in this study to determine the energy of the image pixels to improve the visual quality and provide a clearly defined problem. The logic behind using the K-CFDO approach in image enhancement is the ability of K-CFDO to efficiently capture high-frequency details using the probability of pixels as well as preserve the fine image details. Moreover, the visual quality of X-ray images is improved by performing a low-contrast X-ray image enhancement.•Determine the energy of the image pixels for better pixel intensity enhancement.•Capture high frequency image details using the image probability of pixels.The findings of this study indicate that the average Brisque, Niqe, and Piqe values for the provided chest X-ray were found to be (Brisque=23.25, Niqe=2.8, Piqe21.58), and for the dental X-ray, they were (Brisque=21.12, Niqe=3.77, Piqe=23.49). The results of this study show potential improvements with the proposed enhancement methods that may contribute to increasing efficiency in healthcare processes at rural clinics. Generally, this model improves the details of medical images, which may aid medical staff throughout the diagnostic process by increasing the efficiency and accuracy of clinical decisions. Due to the improper setting of the suggested enhancing parameters, the current study included a limitation on image over-enhancement.

Neural Network-Assisted Data Processing Improved Tomography Characterizations of Reverse Osmosis Polyamide Layers

Electron tomography (ET) can provide comprehensive and detailed structural information on the polyamide selective layer of reverse osmosis membranes. Albeit powerful, the segmentation of low-contrast grayscale images from transmission electron microscope tomography remains a challenge. In this study, a neural network-assisted image segmentation method was developed to solve this problem. A well-trained neural network based on the U-Net model was able to automatically identify the boundary between polymer and vacuum, which reached a 93.0% compatibility with the baseline within a 0.5 h effort of data segmentation, significantly improved upon other segmentation methods. The results were compared with those of a previous study and conventional characterization results for validation. In addition, we performed topography analyses to quantitatively characterize the surface geometry and proposed an entire “sample preparation to 3D structural model” workflow that calibrates the data sets and minimizes tedious manual inputs of ET. For the first time, ET is qualified to be a powerful quantitative characterization technique for separation membranes and other polymer materials widely used in energy and environmental applications.

Defect detection method for high-resolution weld based on wandering Gaussian and multi-feature enhancement fusion

As the weakest point of the pipeline, the weld seam is prone to various internal defects and has great safety risks, so it is necessary to conduct a weld seam inspection. In addition, accurately detecting small-sized defects from high-resolution and low-contrast X-ray images is still a challenging task. Therefore, this paper proposes a defect detection method for high-resolution weld images, including three steps: welded extraction, weld restructuring, and defect detection. Because the resolution of the original image is high and the defect features are mainly concentrated in the weld area, it is not easy to directly defect detection. Firstly, a new irregular long weld extraction algorithm based on wandering Gaussian was designed. Secondly, because the existing welding seams aspect ratio is large and the information loss of conventional compression methods is serious, a new feature reorganization method was provided to maximize the effective information of the weld. Finally, a new method of welding defects detection method based on cross-layer feature fusion has been redesigned to take into efficiency and accuracy. The experiments show that the proposed method achieves the best detection performance on the public GDXray dataset compared with other advanced defect detection methods, from which it can locate the weld defects automatically and fast with high accuracy. It provides an effective solution for high-resolution weld defect detection.

LAPLACIAN-BASED BLUR DETECTION ALGORITHM FOR DIGITAL BREAST TOMOSYNTHESIS IMAGES IN IMPROVING BREAST CANCER DETECTION

The most challenging aspect of working with digital images captured in an uncontrolled environment is to determine whether the image is of sufficient quality to be studied further. One of the most frequent reasons for a decrease in the quality of digital images is the presence of blur artefacts, especially when the images are taken from various angles of the x-ray source with limited angular range such as in digital breast tomosynthesis (DBT). The unwanted artefacts might substantially obscure the breast cancer location, especially in extremely dense fibroglandular breast tissue. It is almost impossible to differentiate breast cancer lesions in blurry and lowcontrast DBT images, thereby reducing the accuracy of lesion diagnosis. Due to this blurry artefact issue, this study aims to assess the performance of Laplacian-based Blur Detection (LbBD) algorithm for the blurry detection of DBT images. The LbBD algorithm is designed and developed using MATLAB R2021a software. The algorithm identifies the amount of blurriness by calculating the image variance; the farther the variance is from the threshold value, the less blurry is the image. On the contrary, the lower the variance level is from the threshold value, the greater the blur level. An online survey was conducted with an expert to assess the quality of 20 DBT images using two subjective measurements of blur or non-blur to ascertain the algorithm's performance. Three threshold values were used to compare the outcome with the output of the algorithm's status. With a low error rate (0.05), an accuracy of 95% at the ideal threshold value of 200, and an image size reduction of 10%, the system successfully predicts blurry images.

Directed searching optimized texture based adaptive gamma correction (DSOTAGC) technique for medical image enhancement.

Because of complexity and low contrast in medical images, few enhancement techniques result unwanted artifacts and information loss by affecting the structure similarity and peak signal to noise ratio. To meet these challenges, a Directed searching optimized texture-based adaptive gamma correction technique is proposed in this article. This proposed technique utilizes the textured regions of the image and suppresses the effect of non-textured regions for eliminating the artifacts. An adaptive clipping threshold is used in the textured image to control the enhancement rate. For improving the contrast, the transfer function of the enhanced image is evaluated using the modified weighted probability density function and adaptive gamma parameter. To make the algorithm more adaptive, parameters like clipped threshold, gamma parameter, and textural threshold are to be optimized using directed searching optimization algorithm. For improving the information contents and noise suppression capability, the proposed technique incorporated a fitness function which is a combination of entropy and peak signal to noise ratio. Equal weightage has been given to each parameter in the fitness function for obtaining a balanced optimal result. Then, the performance of the proposed technique is evaluated in terms of visual quality, information contents, average mean brightness error, noise suppression, and structural similarity. Experimental results show the proposed technique results in better visual effects without information loss. It effectively suppresses the effect of artifacts and significantly improves the contrast by making edges clearer and textures richer over other algorithms.

Noise reduction performance of a deep learning-based reconstruction in brain computed tomography images acquired with organ-based tube current modulation.

We aimed to evaluate the image quality of brain computed tomography (CT) images reconstructed using deep learning-based reconstruction (DLR) in organ-based tube current modulation (OB-TCM) acquisition. An anthropomorphic head phantom and a cylindrical low-contrast phantom were scanned at the standard dose level for adult brain CT in axial volume acquisition without OB-TCM. Moreover, image acquisition with OB-TCM was performed. The radiation dose on the eye lens was measured using a scintillation fibre-optic dosimeter placed on the anthropomorphic phantom's eye surface. The task transfer function (TTF), contrast-to-noise ratio (CNR), and low-contrast object specific CNR obtained from low-contrast phantom images reconstructed with filtered back projection (FBP), hybrid iterative reconstruction (HIR), and two types of DLR (DLRCTA and DLRLCD) were compared. In result, OB-TCM achieved a 32.5% dose reduction in the eye lens. Although HIR, DLRCTA, and DLRLCD showed lower TTF than FBP, the difference in TTF at the highest contributing spatial frequency corresponding to the contrast rod diameter was < 10%. Despite the OB-TCM acquisition, DLRCTA and DLRLCD achieved significantly lower noise and a higher CNR than FBP without OB-TCM (p < 0.05). However, low-contrast object specific CNR was equivalent among all reconstruction methods for the objective diameter of 5mm and slightly improved in DLRLCD for the objective diameter of 7mm. DLR with OB-TCM acquisition enabled dose reduction for the eye lens and high CNR image appearance, whereas the low contrast detectability evaluated by low-contrast object specific CNR did not always improve.

22‐4: Invited Paper: Deep Learning‐based Image Deblurring for Display Vision Inspection

Image quality acts as a major factor in determining its performance in a vision inspection task. The moire pattern caused by frequency aliasing severely degrades the visual quality in display devices, where such high‐quality images are required. To remove these quality undermining patterns, the images are acquired by intentional defocusing. Then, to restore the details lost during the image acquisition, image deblurring is used. The existing deblurring methods fail to output satisfactory results for low contrast Mura images. To solve this problem, we present a novel approach using a generalized Gaussian kernel for real‐world vision inspection tasks. We evaluated the performance and experimented under different settings to validate the robustness of the proposed method. The performance for the proposed method has improved in no‐reference image quality assessment metrics.

SLAAHE: Selective Apex Adaptive Histogram Equalization

Medical imaging enables specialists to make diagnoses our internal body organs for detecting anomalies without being intrusive. Unfortunately, the images produced by various modalities are not only plagued by noises, but the images contrasts are also subpar, making it impossible to use them for a delicate task like diagnosis, which is directly concerned with our health. Pre-processing the image becomes obligatory to ameliorate its medical worth and contrast enhancement is a sacrosanct part of diagnostic image pre-processing. Even though some popular contrast enhancement algorithms, such as Adaptive Histogram Equalization (AHE), Contrast Limited Adaptive Histogram Equalization (CLAHE), and Minimum Mean Brightness Error Bi-Histogram Equalization (MMBEBHE), etc., are widely used, they are insufficient for some medical imaging modalities, such as ultrasound imaging (USG). This paper, named as Selective Apex Adaptive Histogram Equalization (SLAAHE), proposes an effective and efficient contrast enhancement algorithm that dynamically selects the best window size for dividing the low contrast image into image grids and clip limit for clipping the histogram of the low contrast sub-image only. Unique feature of the proposed algorithm is that it clips the image histogram only when it is not well-distributed and ennobles the output image contrast. This makes the algorithm best concerning the contrast enhancement quality, and reduced time and space complexity. Both theoretical analysis and assessment of experimental results with statistical metrics like Universal Image Quality Index (UIQI), Absolute Mean Brightness Error (AMBE), Multi-Scale Structural Similarity Index Measure (MSSSIM), Entropy, and Root Mean Square Error (RMSE) exhibit how the proposed method has gained momentous improvement over existing state-of-the-art methods. For the evaluation metrics UIQI, AMBE, MSSSIM, Entropy, and RMSE, the suggested technique received average scores of 0.79, 0.12, 0.88, 0.92, and 11.52 respectively.

English

Breast cancer is among the top causes of fatalities related to cancer in females. Radiologists commonly use mammogram images to detect breast tumors in their early stages. However, mammography can produce low-contrast images, making it difficult and time-consuming to segment abnormal regions. Deep convolutional neural networks (CNNs) are commonly used for image evaluations. This study used deep CNN models to develop a computer-aided diagnostic (CAD) system for feature extraction and classification. The proposed approach consists of three phases. In the first phase, a shallow, deep CNN model comprising five convolutional layers, five max-pooling layers, one batch normalization layer, and one dropout layer was developed and used to extract recombined images and novel features. In the second phase, the Inception-v3 model was used for label smoothing and classification due to its multiple filters with different sizes. In the third phase, features were extracted using shallow, deep CNN and Inception-v3 models. The Infallible Euclidean distance-based nonlinear dimensionality reduction approach was used to minimize dimensionality. Finally, the Gini-index-based C4.5 decision tree was used for the binary classification of mammogram images from the Digital Database for Screening Mammography (DDSM) + Curated Breast Imaging Subset of DDSM (CBIS-DDSM) and Mammographic Image Analysis Society (MIAS) datasets. The proposed hybrid shallow, deep CNN and Inception-v3 model achieved 99.52% accuracy, a 96% AUC on the DDSM + CBISDDSM dataset, and an accuracy of 97.53% and an AUC value of 97% on the MIAS dataset. Compared with other cutting-edge CAD systems, the proposed hybrid approach achieved higher accuracy by combining in-depth features across both datasets.