applied

sciences
Article
Lung Cancer Detection Model Using Deep Learning Technique
Abdul Rahaman Wahab Sait

Department of Archives and Communication, Center of Documents and Administrative Communication,

King Faisal University, P.O. Box 400, Hofuf 31982, Al-Ahsa, Saudi Arabia; asait@kfu.edu.sa

Abstract: Globally, lung cancer (LC) is the primary factor for the highest cancer-related mortality
rate. Deep learning (DL)-based medical image analysis plays a crucial role in LC detection and
diagnosis. It can identify early signs of LC using positron emission tomography (PET) and computed
tomography (CT) images. However, the existing DL-based LC detection models demand substantial
computational resources. Healthcare centers face challenges in handling the complexities in the model
implementation. Therefore, the author aimed to build a DL-based LC detection model using PET/CT
images. Effective image preprocessing and augmentation techniques were followed to overcome
the noises and artifacts. A convolutional neural network (CNN) model was constructed using the
DenseNet-121 model for feature extraction. The author applied deep autoencoders to minimize
the feature dimensionality. The MobileNet V3-Small model was used to identify the types of LC
using the features. The author applied quantization-aware training and early stopping strategies
to improve the proposed LC detection accuracy with less computational power. In addition, the
Adam optimization (AO) algorithm was used to fine-tune the hyper-parameters in order to reduce
the training time for detecting the LC type. The Lung-PET-CT-Dx dataset was used for performance
evaluation. The experimental outcome highlighted that the proposed model obtained an accuracy
of 98.6 and a Cohen’s Kappa value of 95.8 with fewer parameters. The proposed model can be
implemented in real-time to support radiologists and physicians in detecting LC in the earlier stages.
In the future, liquid neural networks and ensemble learning techniques will be used to enhance the
performance of the proposed LC detection model.

Keywords: lung cancer; machine learning; deep learning; positron emission tomography; computed
tomography; convolutional neural network

Citation: Wahab Sait, A.R. Lung

Cancer Detection Model Using Deep
Learning Technique. Appl. Sci. 2023, 1. Introduction
13, 12510. https://doi.org/10.3390/
LC is a malignancy originating from the lung cells, especially within the epithelial
app132212510
lining of the bronchi, bronchioles, or alveoli [1,2]. It is prevalent and associated with high
Academic Editor: Jorge Igual mortality rates on a global scale. In its earliest stages, LC exhibits no symptoms or presents
with moderate manifestations [2]. As a result, it is typically diagnosed in an advanced
Received: 30 October 2023
Revised: 16 November 2023
stage. The delayed identification of a medical condition affects the efficacy of treatment and
Accepted: 18 November 2023
diminishes the likelihood of achieving long-term survival [3]. The two common types of
Published: 20 November 2023
LC are non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) [4]. NSCLC
is commonly classified into two subtypes: lung squamous cell carcinoma (LUSC) and
lung adenocarcinoma (LUAD) [4]. The fine-grained categorization of LC, including LUSC,
LUAD, and SCLC, has a significant role in determining the prognosis of LC compared
Copyright: © 2023 by the author. to benign and malignant classifications. The precise classification of LC at the primary
Licensee MDPI, Basel, Switzerland. diagnostic stage significantly improves therapy efficacy and subsequently increases patients’
This article is an open access article survival rate [5]. PET and CT are extensively employed as non-invasive diagnostic imaging
distributed under the terms and
modalities in clinical practice, serving as valuable tools for assessing the specific diagnosis
conditions of the Creative Commons
of LC [6].
Attribution (CC BY) license (https://
The early detection and treatment of LC through effective screening methods are
creativecommons.org/licenses/by/
vital in enhancing patient outcomes. Based on the National Lung Screening Trial findings,
4.0/).

Appl. Sci. 2023, 13, 12510. https://doi.org/10.3390/app132212510 https://www.mdpi.com/journal/applsci

Appl. Sci. 2023, 13, 12510 2 of 17

low-dose helical CT screening is more effective in reducing mortality among high-risk

populations [7]. Nevertheless, the screening process for LC is susceptible to yielding false
positive (FP) results, leading to increased costs due to unwarranted medical interventions
and may induce psychological distress in individuals [8]. Computer-aided diagnosis has
notable advantages in LC detection, including enhanced scope in early cancer screening
and a diminished occurrence of FP results throughout the diagnostic process [9].
In the realm of LC detection, there have been notable developments in the form
of novel approaches and technologies for enhancing early diagnosis and treatment effi-
cacy [10]. Liquid biopsies are used to examine blood samples for detecting cancers [11].
These diagnostic tests can identify genetic abnormalities and modifications associated
with LC. Consequently, these tests offer a non-invasive approach to diagnosing the dis-
ease and monitoring the effectiveness of treatment. Low-dose computed tomography
(LDCT) screening has emerged as a widely applied method for the timely identification of
LC [12]. LDCT scans employ reduced radiation levels compared to conventional CT scans
while delivering high-resolution images of the pulmonary region. Novel bronchoscopic
methodologies, including electromagnetic navigation bronchoscopy and robotic-assisted
bronchoscopy, facilitate the performance of minimally invasive lung lesion biopsies. These
tools promote the prompt and precise identification of medical conditions. Integrating
genomes, proteomics, and metabolomics has enabled the development of diverse strategies
for identifying LC [13]. These methodologies employ several molecular markers to enhance
diagnostic precision and discover possible targets for therapeutic intervention.
DL-based LC screening techniques can reduce mortality by detecting the disease in
the primary stage [14]. It can aid in lowering false negatives (FN) by detecting subtle or
early indicators of LC that humans may leave unnoticed [15]. Imaging modalities including
CT, magnetic resonance imaging (MRI), and PET can be integrated using DL algorithms to
understand the disease and assist in diagnosis and therapy planning. Cancer staging is
directly related to the extent of disease metastasis [15]. A combination of imaging modalities
and biopsies of suspicious tissue identifies cancer types. Cancer staging assists caregivers
in selecting chemotherapy, immunotherapy, radiation, and surgical strategies [16]. In
particular, the higher cancer stage increases the mortality rate. The effectiveness of medical
therapies is based on the cancer stages. Providing more accurate and trustworthy diagnoses
can help reduce misdiagnosis and unwarranted care. The process of LC screening produces
substantial quantities of medical imaging data. DL models effectively analyze CT scans,
chest radiography, and other imaging modalities due to their ability to analyze massive
data [16–18]. Using DL models, physicians may gain insights into a patient’s state by
combining data from several imaging modalities and other clinical data sources.
Due to privacy concerns and the cost of data acquisition, medical image datasets are
typically small. Pre-trained models can apply broader image dataset expertise to medical
images, enabling model training with minimal medical data [19]. These models can extract
hierarchical information from images, including fine-grained details and crucial patterns.
Feature extraction supports the medical image classification models to detect significant
patterns, abnormalities, and disease indicators [20]. The process of training CNN models
from scratch can pose significant computing costs and require a substantial amount of time.
The deep transfer learning technique reduces training time and time spent fine-tuning the
model for medical image classification [21].
LC diagnosis and classification rely heavily on PET/CT imaging. PET and CT images
are combined in PET/CT-based LC detection models to understand lung lesions’ location,
size, and metabolic activity. PET/CT scans are employed to monitor the efficacy of cancer
treatment. The assessment of therapy success can be facilitated by tracking changes in
metabolic activity and tumor growth over a period of time. Differentiating LC types and fea-
tures with PET/CT imaging allows customized treatment. Bhandary et al. [20] developed a
DL framework to detect LC using the CT images. Kanavati et al. [21] proposed a weekly-
supervised learning-based model to identify lung carcinoma. Hallitschke et al. [22] built a
multimodal LC lesion segmentation technique for annotating the images. They utilized
Appl. Sci. 2023, 13, 12510 3 of 17

the geodesic distance transformation technique and implemented an ellipsoid-based user

simulation scheme. Ardila et al. [23] introduced an end-to-end LC screening model using
LDCT images. Marentakis et al. [24] proposed an LC histology classification technique.
Similarly, the studies [25–33] employed DL models for classifying the medical images to
detect LC. Barbouchi et al. [34] introduced a deep neural network for LC classification.
They obtained an accuracy of 97% in classifying the PET/CT images. El Hamdi et al. [35]
proposed a VGG-16 network-based LC detection model. They generalized their model
using the Lung-PET-CT-Dx dataset [36,37], which contains 251,135 PET/CT images. The
model obtained an accuracy of 94% and an area under the receiver operating characteristic
curve (AU-ROC) of 97%. Goswami and Singh [38] proposed a pulmonary LC classification
model using a deep neural network. They applied accuracy, precision, recall, F1-score, and
Cohen’s Kappa metrics for the performance evaluation. Liu et al. [39] discussed the role of
deep reinforcement learning in finding LC using complex images.
In some healthcare settings, computing resources are inaccessible for training huge and
complicated deep learning models for lung cancer diagnosis. To adapt DL-based solutions,
healthcare centers demand high computational resources. DL algorithms may face chal-
lenges in generating an optimal outcome with limited datasets. The authors’ motivation
for employing DL techniques for LC detection emerges from the potential to significantly
transform early diagnosis, enhance patient care, contribute to scientific advancements, and
generate substantial benefits for public health. Radiologist and physician effort may be
reduced by automating LC detection using DL techniques. This automation can improve
speed and consistency. Thus, the authors intend to build a cost-effective LC detection
model using the DL technique. The study includes the following contributions:
1. A DL model for extracting critical features from the PET/CT images;
2. An LC detection model for identifying the type of LC with less computational resources;
3. Performance evaluation of the proposed LC model using state-of-the-art techniques.
The proposed study is structured as follows: Section 2 offers the proposed method-
ology for classifying types of LC using PET/CT images. The experimental results are
presented in Section 3. Section 4 describes the significance of the proposed LC detection
model. The study’s contribution and limitations are presented in Section 5.

2. Materials and Methods

The quality of medical images can affect feature segmentation approaches. The pres-
ence of noise, artifacts, or low-resolution images can diminish the quality of feature ex-
traction. Feature segmentation techniques typically depend on pre-established feature
sets, which may not include all pertinent data for the identification of lung cancer [33].
Specific feature segmentation algorithms may demand high computational resources when
working with large medical image datasets. These algorithms may restrict applications to
operate with limited resources [34]. Different patient positions and imaging settings may
negatively impact the accuracy of feature segmentation methods.
DenseNet-121 possesses a high degree of parameter efficiency. It operates with fewer
parameters, leading to an LC detection model with fewer computational resources [40].
It can extract rich and diverse information from PET/CT images. A dense connection
improves backpropagation gradient flow. Therefore, the vanishing gradient challenge can
be resolved by allowing gradients to autonomously and swiftly relocate throughout the
network. This approach facilitates the proper training of deeper networks and ensures
the acquisition of characteristics at multiple levels of abstraction. However, multiple
connections within the network may result in elevated memory consumption. This could
constrain the maximum batch size during the training process. The MobileNet V3-Small
design offers exceptional efficiency and suitability for mobile and embedded devices [41].
It can be used in real-time object identification and image classification applications. The
model is comparatively compact in relation to numerous deep learning models. It can
be deployed on devices that possess restricted storage capacity. However, the MobileNet
V3-Small model may not learn more complex characteristics than larger DL models. The
 Appl. Sci. 2023, 13, x FOR PEER REVIEW 4 of 18

Appl. Sci. 2023, 13, 12510 4 of 17

model is comparatively compact in relation to numerous deep learning models. It can be
deployed on devices that possess restricted storage capacity. However, the MobileNet V3-
Small model may not learn more complex characteristics than larger DL models. The pres-
presence of complicated patterns in datasets can potentially restrict its performance. The
ence of complicated patterns in datasets can potentially restrict its performance. The
depthwise separable convolution layers of the MobileNet V3-Small model make it difficult
depthwise separable convolution layers of the MobileNet V3-Small model make it difficult
to train them from scratch on complex datasets. This motivates the user to implement the
to train them from scratch on complex datasets. This motivates the user to implement the
proposed LC detection model using the DenseNet-121 model for feature extraction and
proposed LC detection model using the DenseNet-121 model for feature extraction and
the
the MobileNetV3-Small
MobileNet V3-Small model
model for the LC
for the LCimage
imageclassification.
classification.The
The author
author presents
presents an an
LCLC
detection model using the CNN models in this study. Initially, an image
detection model using the CNN models in this study. Initially, an image preprocessing preprocessing
technique
techniqueisisused
usedtotoimprove
improve the
the visibility of PET/CT
visibility of PET/CTimages.
images.Subsequently,
Subsequently,thethe author
author
followed an image augmentation technique to increase the dataset size. The
followed an image augmentation technique to increase the dataset size. The DenseNet- DenseNet-121
model is used
121 model for feature
is used extraction.
for feature extraction.The
Theauthor
author employed
employed aadimensionality
dimensionality reduction
reduction
technique to reduce the number of features. Finally, enhanced MobileNet V3-Small
technique to reduce the number of features. Finally, enhanced MobileNet V3-Small model model
weights are used for detecting LC. Figure 1 outlines the proposed methodology
weights are used for detecting LC. Figure 1 outlines the proposed methodology for de- for detecting
LC usingLC
tecting PET/CT images.images.
using PET/CT

Figure 1. The proposed LC detection framework.

Figure 1. The proposed LC detection framework.

2.1.Image
2.1. ImageAcquisition
Acquisition
The author utilizesthe
The author utilizes the PET/CT(Lung-PET-CT-Dx)
PET/CT (Lung-PET-CT-Dx)dataset datasettotogeneralize
generalizethethe pro-
proposed
posed LC detection model. Lung-PET-CT-Dx is a recently developed
LC detection model. Lung-PET-CT-Dx is a recently developed dataset that comprises dataset that com-
prisesannotated
31,562 31,562 annotated
images.images. It is publicly
It is publicly available
available in the repository
in the repository [36,37].[36,37]. The im-
The images were
ages were retrospectively
retrospectively collected fromcollected fromhealthcare
multiple multiple healthcare centers
centers across acrossThe
China. China. The
computer
computer
center center and
and cancer cancerHarbin
institute, institute, HarbinUniversity
Medical Medical University
in Harbin,in Harbin, China, sup-the
China, supported
ported the dataset owners in obtaining permission to collect the PET/CT
dataset owners in obtaining permission to collect the PET/CT images. In addition, images. In addi-the
tion, the patient’s personal information was not included in the dataset. The dataset con-
patient’s personal information was not included in the dataset. The dataset contains CT
tains CT and PET-CT DICOM images of LC patients. It provides XML annotation files to
and PET-CT DICOM images of LC patients. It provides XML annotation files to indicate
indicate tumor location and bounding boxes. The tissue histopathological diagnosis was
tumor location and bounding boxes. The tissue histopathological diagnosis was used for
used for the patients’ classification. The CT image resolution was 512 × 512 pixels at 1
the patients’ classification. The CT image resolution was 512 × 512 pixels at 1 mm × 1 mm,
mm × 1 mm, and the PET image resolution was 200 × 200 pixels at 4.07 mm × 4.07
and the PET image resolution was 200 × 200 pixels at 4.07 mm× 4.07 mm. Five academic
mm. Five academic thoracic radiologists with expertise in LC detection were employed to
thoracic radiologists with expertise in LC detection were employed to annotate the tumor
annotate the tumor location. The Labellmg 1.4.0 package was used to capture the annota-
location. The Labellmg 1.4.0 package was used to capture the annotation. The images were
tion. The images were broadly classified into adenocarcinoma (A), small cell carcinoma
broadly classified into adenocarcinoma
(B), large cell carcinoma (E), and squamous (A), small cell carcinoma
cell carcinoma (B), largethe
(G). However, celldataset
carcinoma
is
(E), and squamous cell carcinoma (G). However, the dataset is highly imbalanced. Table 1
offers the dataset characteristics. Figure 2 highlights the sample LC images.
 Appl. Sci. 2023, 13, x FOR PEER REVIEW 5 of 18

Appl. Sci. 2023, 13, 12510 5 of 17

highly imbalanced. Table 1 offers the dataset characteristics. Figure 2 highlights the sam-
ple LC images.
Table 1. Dataset characteristics.
Table 1. Dataset characteristics.
Typesof
Types ofLC
LC Patients
Patients Annotated
Annotated Images
Images
AA 266266 20,894
20,894
BB 44 44 3116
3116
E 5 201
E 5 201
G 62 7351
G 62 7351

Figure 2. Sample images.

Figure 2. Sample images.

2.2.Image
2.2. ImagePreprocessing
Preprocessing
PET and
PET and CT CT images
images are usually
usually obtained
obtainedas asseparate
separatescans
scansandandmaymayexhibit
exhibitmisalign-
misalign-
mentdue
ment duetotopatient
patient mobility.
mobility. Various
Various artifacts,
artifacts, including
including motion
motion artifacts
artifacts in PETin PET and
and metal
metal artifacts
artifacts in CT may
in CT scans, scans, may influence
influence these images.
these images. The image
The image registration
registration is usedistoused to
guaran-
guarantee
tee spatial alignment
spatial alignment betweenbetween two images.
two images. The symmetric
The symmetric normalization
normalization (SyN)
(SyN) function
isfunction is widely
widely used used for non-linear
for non-linear image registration.
image registration. The author The authorthe
applied applied the SyNto
SyN function
function
handle thetodeformations
handle the deformations
and improve and improve
the image the image alignment.
alignment. The SyN integrates
The SyN function function
integrates
PET PETactivity
metabolic metabolic activity
with with CT anatomical
CT anatomical data. It employs
data. It employs a forward a forward transfor-for
transformation
generating the images. A similarity metric measures the similarity between the sourcethe
mation for generating the images. A similarity metric measures the similarity between and
sourceimage.
target and target
Theimage. The SyNemploys
SyN function function employs cross-correlation
cross-correlation computation
computation to gen-
to generate the
erate the normalized
normalized image. Inimage. In addition,
addition, it computes
it computes velocityvelocity
to applytosmoothness
apply smoothness to the
to the images.
images.
The The and
forward forward andtransformations
inverse inverse transformations
are used are used to the
to improve improve the alignment
alignment accuracy.ac-
The
images were resized to 512 × 512 pixels to improve the classification accuracy. Equation (1)
shows the computation of the SyN function. The function captures the complex defor-
Appl. Sci. 2023, 13, 12510 6 of 17

mation in the PET/CT images by building a Gaussian pyramid. It identifies the spatial
transformation for generating the spatially coincident features.

Ii = SyN ( Ii , ϕ1 , ϕ2 ), i = 1, . . . , n (1)

where ϕ1 and ϕ2 are diffeomorphism variables used to compute cross-correlation and

velocities to normalize the images.
PET and CT images may have unique spatial resolutions and voxel sizes. The author
employs the resample image filter function to normalize the spatial resolution of the
PET/CT images. The function uses the interpolating pixel values to improve the spatial
resolution of the PET/CT images. It minimizes high-frequency objects and restricts the
irrelevant patterns during the resampling process. The mathematical form of the image
resampling process is presented in Equation (2). The function modifies the dimensions
of the image grid to maintain an effective resolution of the images. An interpolation
technique is selected to compute the voxel values at non-grid positions depending on the
image characteristics. The resample image filter function standardizes the PET/CT image
dimensions and supports the proposed model to achieve a meaningful outcome.

Ii = Resample_Image_Filter ( Ii , s, scale, N I ), i = 1, . . . , n (2)

where s is the image size, scale is the scaling factor, and NI is the nearest neighbor interpolation.
The attenuation correction function enhances the anatomical and functional informa-
tion of the PET/CT images. It reconstructs the images by reducing the artifact impressions.
The author applies attenuation correction to address the artifacts of the PET images. In
addition, the metal artifact correction technique eliminates the artifacts of the CT images.
Equation (3) outlines the attenuation correction and metal artifact correction process. The
attenuation correction function identifies variations in tissue density by measuring the
attenuation of the gamma ray photons. It reduces the impact of scatter artifacts by con-
structing the primary photon attenuation. The metal attenuation correction function applies
an iterative reconstruction technique to reduce the metal artifacts. In addition, it replaces
the metal artifact regions with interpolated values of the surrounding tissue regions of
the images.

Ii = attenuation( Ii , A) + Metal_attenuation( Ii , M), i = 1, . . . , n (3)

where A and M are the attenuation coefficients for reducing the impression of the artifacts.
To improve the image contrast level, the author employs the Retinex filtering technique
to address non-uniform illumination and shading of PET/CT images. Retinex filtering
enhances the color and brightness of the PET/CT images. It preserves the overall structure
and details of the images. It decomposes the images using the illumination and reflectance
components. The logarithmic transformation technique modifies the image by multiplying
the illumination and reflectance components into an additive relationship. As a result, the
images are reconstructed with improved contrast and visibility. Equation (4) highlights
the computational form of the Retinex filtering function. A logarithmic transformation
decomposes the images into reflectance and illumination components. The function applies
convolution to separate low- and high-frequency components. An exponential function
transforms the images into linear space.

Ii = Retinex_Filter ( Ii , I M, RF ), i = 1, . . . , n (4)

where IM is the illumination component and RF is the reflectance component.

2.3. Image Augmentation

In order to increase the dataset size, enhance model generalization, and reduce overfit-
ting, the author applied image augmentation techniques. The geometric transformations
are used to generate variations in the image sizes. The images are rotated to 30, 60, 90, and
 Appl. Sci. 2023, 13, x FOR PEER REVIEW 7 of 18

Appl. Sci. 2023, 13, 12510 7 of 17

2.3. Image Augmentation
In order to increase the dataset size, enhance model generalization, and reduce over-
fitting, the author applied image augmentation techniques. The geometric transfor-
120 degrees
mations aretoused
simulate deviations
to generate in the in
variations orientation
the imageof the patients.
sizes. The images Theare
author
rotatedflipped
to 30,the
images
60, 90, to
and generate
120 degreesanatomical variations.
to simulate deviations Theinelastic deformation
the orientation of theispatients.
employed Thetoauthor
produce
changes
flipped inthetissue
images shape and position.
to generate anatomical Thevariations.
superimposition
The elastictechnique
deformation is used to include
is employed
synthetic
to produce lesions
changesfor pathological
in tissue shape condition simulations.
and position. It supports the
The superimposition evaluation
technique of the
is used
model’s
to includecapability
syntheticinlesions
classifying the LC types.
for pathological Generative
condition adversarial
simulations. networks
It supports (GANs)
the evalu-
facilitate
ation ofthe thegeneration of high-quality
model’s capability synthetic
in classifying theimages. TheGenerative
LC types. system comprises
adversarialtwo net-
neural
works (GANs) facilitate the generation of high-quality synthetic
networks, namely a generator and a discriminator, which operate in a mutually compet- images. The system com-
prises
itive two neural
manner. networks,
In order namely
to address theadata
generator and a discriminator,
imbalance, the author employs which operate
multi-channelin a
mutually
GANs [42]competitive
to generatemanner. In order
the synthetic to address
images. The the data imbalance,
multi-channel GAN themodel
authorwas employstrained
multi-channel
and tested using GANs [42] to
the lung generate
PET/CT the synthetic
images. images. The multi-channel
The preprocessed images were used GANfor modelgener-
was the
ating trained
images.and The
tested using the GANs
pre-trained lung PET/CT
allow the images.
author The preprocessed
to produce images were
the PET/CT images
used for
related to generating
the real PET/CTthe images.
images Thewithout
pre-trained GANs allow
additional the author
training. to produce
In addition, the
the images
PET/CT
are images
generated at related
multiple to the real in
points PET/CT images
the latent without
space additionalsemantic
for achieving training. In addition,
interpolation.
the images
Equation (3) are generated
shows at multiple points
the mathematical form of in generating
the latent space
imagesfor using
achieving
GANs. semantic in-
terpolation. Equation (3) shows the mathematical form of generating images using GANs.
I = GANs( I𝑖i )=
𝐼 i= 𝐺𝐴𝑁𝑠(𝐼 ),
, i1,=…1,
,𝑛
..., n (5) (5)
2.4. Feature Extraction
2.4.The
Feature Extraction
author builds a feature extraction model using the weights of the pre-trained
The author
DenseNet-121 model. builds
A asequence
feature extraction model using
of convolutional layers,the weights
batch of the pre-trained
normalization, and ReLu
DenseNet-121 model. A sequence of convolutional layers, batch normalization,
function is employed for extracting the features. Figure 3 reveals the architecture and ReLu of
function
the proposedis employed for extracting
feature extraction the features.
model. The finalFigure
set of3 reveals
layers in thethe
architecture of the
feature extraction
proposed
model featureusing
is trained extraction model. The final
the DenseNet-121 set ofweight
model’s layers in the afeature
with extraction
learning rate of 1model
× 10−3 .
is trained
The frozen using the DenseNet-121
pre-trained DenseNet-121model’s weight
model with afewer
requires learning × 10
rate of 1 for
parameters the .feature
The
frozen pre-trained DenseNet-121 model requires fewer parameters for the feature
extraction. Finally, the global average pooling layer obtains a fixed-size feature vector for extrac-
tion.image.
each Finally,Thethe global
featureaverage pooling
extraction layerminimizes
process obtains a fixed-size feature vector
the computational for eachfor
resources
image. The feature extraction process minimizes the computational resources
classifying the PET/CT images. Equations (6) and (7) show the mathematical forms of for classify-
ing the PET/CT images. Equations (6) and (7) show the mathematical forms of generating
generating fixed-size features.
fixed-size features.
Features==𝐶𝑜𝑛𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛
𝐹𝑒𝑎𝑡𝑢𝑟𝑒𝑠 Convolution++ ReLu
𝑅𝑒𝐿𝑢 + Batch_Normalization),( I𝑖i )=, 1,
+ 𝐵𝑎𝑡𝑐ℎ_𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛(𝐼 i= , 𝑛. . . , n
… 1, (6) (6)

Fixed_size_ f eatures
𝐹𝑖𝑥𝑒𝑑_𝑠𝑖𝑧𝑒_𝑓𝑒𝑎𝑡𝑢𝑟𝑒𝑠 = Global_Average_Pooling_2D ( f eatures)
= 𝐺𝑙𝑜𝑏𝑎𝑙_𝐴𝑣𝑒𝑟𝑎𝑔𝑒_𝑃𝑜𝑜𝑙𝑖𝑛𝑔_2𝐷(𝑓𝑒𝑎𝑡𝑢𝑟𝑒𝑠) (7) (7)

Figure The
3. 3.
Figure proposed
The feature
proposed extraction
feature process.
extraction process.
2.5. Dimensionality Reduction
Deep autoencoders are used to minimize PET/CT image dimensionality by developing
a compact latent space representation [38]. This method employs an encoder–decoder
architecture to attempt a reconstruction of the input image from its compressed lower-
dimensional representation. The author uses the multilayer encoder network to reduce the
feature dimensionality. The encoder’s layers facilitate the conversion of input data into an
expression of reduced dimensionality by utilizing learned weights and biases. Lowering
dimensions is accomplished by employing several processes like convolutions, pooling,
Appl. Sci. 2023, 13, 12510 8 of 17

and non-linear activations. The encoder’s final stage compresses the input picture into
latent space. Compared to the original image, the dimensionality of this latent space is
substantially less. In the autoencoder design, the latent space is the limiting factor. The
condensed version of the input image captures its fundamental elements and patterns.
Using the low-dimensional representation from the latent space, the decoder network
attempts to recreate the original feature. The number of dimensions of the representation
expands with decoder layers. The decrease in dimensionality can be attributed to the
constriction in the latent space. The encoder network preserves essential characteristics of
the features while removing irrelevant or identical data.

2.6. LC Detection Model

The primary objective of MobileNet V3 is to optimize the performance of mobile
devices, embedded systems, and Internet of Things devices [41]. MobileNet models are
ideal for developing lightweight image classifications. However, its small size may restrict
its capacity to capture complex shapes and patterns compared to larger, parameter-heavy
models. The efficient MobileNet V3 may not perform complicated feature extraction tasks,
including high-resolution images or detailed object recognition [41]. MobileNet V3 models
are developed for real-time applications and mobile devices. Certain trade-offs should be
considered regarding the model size and its accuracy. A MobileNet V3-Small model may
be unsuitable for applications requiring precision and minimal computational resources. It
may exhibit limited generalization capabilities in classifying PET/CT models. In order to
overcome the challenges, the author develops a classification model with the weights of the
MobileNet V3-Small model for classifying the PET/CT images. A total of four convolutional
layers are used for the classification. The final layers are trained with MobileNet weights.
During the training, the author unfreezes three layers of the MobileNet model. The fully
connected layers are integrated into the proposed model with the Softmax function to
predict the type of LC. Equation (8) shows the computational form of the Softmax function.

→ e Ii
 
σ I = Ij
(8)
∑4j=1 e
→
where σ() is the Softmax function, I is the image, e Ii is the standard exponential factor of
the input vector, and e Ij is the standard exponential factor of the output vector.
Furthermore, the author employs a learning rate and decay schedule for improving
training stability and convergence. The data loading and preprocessing pipeline is opti-
mized to reduce the training time. The author introduces quantization-aware training to
maintain the trade-off between model size and inference speed [43]. Quantization-aware
training optimizes the proposed model using weight and activation redundancy [43]. It
allows the model to be lowered to a quarter of their original size and memory footprint.
In addition, the inference rates are increased by a factor of two to four with minimal or
no loss of accuracy [43]. The training process enables the proposed model accessible for
edge devices and real-time applications. Finally, early stopping is implemented to prevent
overfitting and reduce the training time.

2.7. Hyper-Parameter Optimization

In order to fine-tune the parameters, the author employs the Adam optimization (AO)
algorithm. The AO algorithm integrates the features of RMSProp and Adagrad algorithms.
It adapts the mean of the second moments of the uncentered variance. The exponential
moving average of the squared gradients is used to control the decay rates of the proposed
LC detection model.

2.8. Evaluation Metrics

In the context of LC detection, true positives (TPs), true negatives (TNs), FPs, and
FNs have unique meanings. A TP represents the correctly identified specific LC type,
Appl. Sci. 2023, 13, 12510 9 of 17

TN indicates the correctly identified image that belongs to a different LC type, FP is an

incorrectly identified LC type, and FN represents the incorrectly identified image that
belongs to a different LC type.
The author follows the benchmark evaluation metrics for evaluating the proposed
model in this study. The accuracy metric assesses the prediction quality of the pro-
posed model. It determines the proportion of accurately predicted instances, which in-
cludes TP and TN, in relation to the overall number of instances contained within the
dataset. Precision is the percentage of cases in which the model made an optimistic pre-
diction. Recall is the percentage of TP predictions among all positive cases. The F1-score
presents the harmonic mean of precision and recall. A higher F1-score signifies a model
consistently exhibiting elevated recall and precision levels. In addition, AU-ROC, the
precision–recall curve (AU-PRC) and the loss function are computed for the LC detec-
tion models. Equations (9)–(12) show the mathematical form for computing the accuracy,
precision, recall, and F1 measure.

TP + TN
Accuracy = (9)
TP + TN + FP + FN

TP
Precision = (10)
TP + FP
TP
Recall = (11)
TP + FN

2 × Precision × Recall
F1 − Score = (12)
( Precision + Recall )
Cohen’s Kappa measures the model’s consistency in medical image categorization
or classification settings [44]. It is a standardized method for measuring the multi-class
classification ability of the DL models. It is beneficial for imbalanced datasets where preci-
sion may not be reliable for evaluating the model’s effectiveness. In addition, it quantifies
inter-rater or inter-classifier agreement in multi-class classification problems based on
chance agreement. Equation (13) represents the mathematical form of Cohen’s Kappa.

2 × ( TP × TN − FN × FP)
Kappa = (13)
( TP + FP) + ( FP + TN ) + ( TP + FN ) + ( FN + TN )

Standard deviation (SD) and confidence intervals (CI) are used to analyze the data
distribution and population parameter estimation. The author used these metrics to identify
the variability and uncertainty in the outcome. The mean absolute deviation (MAD) and
the root mean square error (RMSE) are employed to evaluate the precision and reliability of
predictions. The MAD is used to compare the predicted and actual outcomes. It determines
the average of the absolute differences between each anticipated value and its associated
real value. The RMSE measures the average squared deviation from the projected value. It
is sensitive to outliers compared to the MAD and emphasizes the significant errors.

3. Results
The author conducted the experiment in Python 3.8.3 using the Keras library. The
proposed model is implemented in Windows 10, Intel i7, 16 GB RAM, and GeForce GTX
950 GPU. The weights of the DenseNet-121 and MobileNet V3-Small models were ex-
tracted from the repository [40,41]. Table 2 presents the hyper-parameters of the proposed
LC model.
Appl. Sci. 2023, 13, 12510 10 of 17

Table 2. Model development settings.

Parameters Value
Image dimension 512 × 512
Convolutional layers 4
Size of strides 2
Activation function Softmax
Loss function Cross-Entropy
Batch Size 256
Fully connected layer 1

Adagrad, RMSProp, and AO algorithms are used to fine-tune the parameters of

the proposed LC detection model. Table 3 highlights the outcome of the proposed LC
detection model. It indicates that the AO-based LC detection model achieved an excep-
tional performance in classifying the type of LC using the PET/CT images. The compu-
tation of bias correction estimates supported AO to outperform RMSProp and Adagrad
optimization algorithms.

Table 3. Findings of hyper-parameter optimization.

Hyperparameter Optimization Accuracy Precision Recall F1-Score Kappa

AO-based Proposed LC detection model 98.6 97.9 98.1 98.0 95.8
Adagrad-based Proposed LC detection model 97.1 96.1 96.3 96.2 94.2
RMSProp-based Proposed LC detection model 96.8 96.9 96.5 96.7 95.3

Table 4 illustrates the computational practices utilized by the LC detection models.

The suggested model exhibits a decline in the number of parameters, FLOPs, learning rate,
and training duration compared to the existing models.

Table 4. Computational settings.

Parameters FLOPs Training Time

Models Learning Rate
(in Millions (m)) (in Millions (m)) (in Seconds (s))
Proposed LC detection model 1 × 10−3 2.1 176 128
DenseNet-121 model 1 × 10−3 8.3 679 187
Barbouchi et al. model [34] 1 × 10−3 9.2 892 372
El-Hamdi et al. model [35] 1 × 10−3 18.3 912 486
Goswami and Singh model [38] 1 × 10−3 16.2 872 332
EfficientNet model 1 × 10−3 4.8 496 386
ShuffleNet model 1 × 10−3 3.2 276 512

Table 5 shows the performance of the proposed LC detection model with different
batch sizes. Figure 4 presents the proposed model’s performance in the multiple epochs.
The proposed model showed a significant improvement in the 180th batch. However, the
author extended the batch size to 256. The proposed feature extraction supported the
proposed LC detection model to produce an exceptional outcome.

Table 5. Findings of the batch-wise performance evaluation.

Batch Size Accuracy Precision Recall F1-Score Kappa

60 95.3 94.8 94.3 94.5 91.2
130 97.1 96.2 96.4 96.3 93.4
180 98.4 96.8 96.7 96.7 94.6
256 98.6 97.9 98.1 98.0 95.8
 Table 5. Findings of the batch-wise performance evaluation.

Batch Size Accuracy Precision Recall F1-Score Kappa

60 95.3 94.8 94.3 94.5 91.2
130 97.1 96.2 96.4 96.3 93.4
Appl. Sci. 2023, 13, 12510 180 98.4 96.8 96.7 96.7 94.611 of 17
256 98.6 97.9 98.1 98.0 95.8

Findings of
Figure 4. Findings
Figure of the
the performance
performance analysis
analysis in the multiple epochs.

Table 66 reveals
Table revealsthe
theperformance
performance of of
thethe
proposed model.
proposed TheThe
model. proposed model
proposed obtained
model ob-
an exceptional outcome for each class. The proposed image augmentation
tained an exceptional outcome for each class. The proposed image augmentation tech- technique
effectively
nique supported
effectively the proposed
supported modelmodel
the proposed in identifying each each
in identifying class’s key patterns.
class’s The
key patterns.
findings of the performance analysis are presented in Table 6. Figure 5 highlights
The findings of the performance analysis are presented in Table 6. Figure 5 highlights the the
outcome of the performance evaluation.
outcome of the performance evaluation.
Table 6. Findings of the performance evaluation.
Table 6. Findings of the performance evaluation.
Types of LC Accuracy Precision Recall F1-Score Kappa
Types of LC Accuracy Precision Recall F1-Score Kappa
A 98.6 97.8 97.5 97.6 96.1
BA 98.6
98.6 97.8
97.6 97.5
97.7 97.6
97.6 96.1
95.8
EB 98.6
98.8 97.6
98.1 97.7
98.5 97.6
98.3 95.8
96.1
Appl. Sci. 2023, 13, x FOR PEER REVIEW 12 of 18
GE 98.5
98.8 98.3
98.1 98.7
98.5 98.5
98.3 95.4
96.1
Average
G 98.6
98.5 97.9
98.3 98.1
98.7 98.0
98.5 95.8
95.4
Average 98.6 97.9 98.1 98.0 95.8

Findings of
Figure 5. Findings
Figure of the
the multi-class
multi-class classification.
classification.

Table 77outlines
Table outlinesthe
thegeneralization
generalizationofof
thethe
LCLC detection
detection models
models on Lung-PET-CT-
on the the Lung-PET-
CT-Dx dataset. The proposed model outperformed the existing LC detection
Dx dataset. The proposed model outperformed the existing LC detection models models by
by
achieving an
achieving an exceptional
exceptional outcome.
outcome. The
The MobileNet
MobileNet V3-Small
V3-Small architecture
architecture supported
supported the
the
proposed LC detection model to achieve a better outcome. In addition, the suggested
proposed LC detection model to achieve a better outcome. In addition, the suggested im-image
age preprocessing technique assisted the proposed model in detecting the crucial LC pat-
terns in the PET/CT images. Figure 6 shows the outcome of the comparison analysis.

Table 7. Comparative analysis outcome.

Appl. Sci. 2023, 13, 12510 12 of 17

preprocessing technique assisted the proposed model in detecting the crucial LC patterns
in the PET/CT images. Figure 6 shows the outcome of the comparison analysis.

Table 7. Comparative analysis outcome.

Models Accuracy Precision Recall F1-Score Kappa

Proposed LC detection model 98.6 97.9 98.1 98.0 95.8
DenseNet-121 model 96.4 96.7 96.8 96.7 93.8
Barbouchi et al. model [34] 97.0 96.7 96.5 96.6 94.7
El-Hamdi et al. model [35] 97.5 97.2 97.4 97.3 95.7
Goswami and Singh model [38] 96.8 97.6 97.3 97.4 96.1
EfficientNet
Appl. Sci. 2023, 13, x FOR model
PEER REVIEW 97.5 96.9 96.7 96.8 95.1
13 of 18
ShuffleNet model 96.7 97.2 97.5 97.3 95.4

Figure 6. Findings of the comparative analysis [34,35,38].

Table 8 highlights the loss, AU-ROC, AU-PRC, AU-PRC, SD,SD, and

and CI
CI of the models. TheThe sug-
sug-
gested LC
LCdetection
detectionmodel
modelidentified
identifiedthethe
multiple types
multiple of LC
types of with minimum
LC with loss. loss.
minimum AU-ROCAU-
and AU-PRC
ROC and AU-PRCvaluesvalues
indicated the significance
indicated of the proposed
the significance model in
of the proposed detecting
model FP and
in detecting
FN images. The SD value of 0.0002 indicates that the proposed model’s
FP and FN images. The SD value of 0.0002 indicates that the proposed model’s perfor- performance is
stable and
mance consistent.
is stable Lastly, the Lastly,
and consistent. CI range ofCI
the 98.1–98.5 shows
range of that the
98.1–98.5 proposed
shows LCproposed
that the detection
model
LC is reliable
detection andiscan
model be implemented
reliable in healthcare centers.
and can be implemented in healthcare centers.

Table 8. Findings of the AU-ROC, AU-PRC, and loss analysis with SD and CI.

Models Loss AU-ROC AU-PRC SD CI

Proposed LC detection model 0.89 0.98 0.98 0.0002 [98.1–98.5]
DenseNet-121 model 1.13 0.97 0.96 0.0003 [96.7–97.4]
Barbouchi et al. model [34] 1.02 0.97 0.97 0.0004 [96.7–97.4]
El-Hamdi et al. model [35] 0.95 0.97 0.96 0.0003 [97.5–98.1]
Goswami and Singh model [38] 1.12 0.96 0.95 0.0004 [96.4–96.8]
EfficientNet model 0.99 0.97 0.97 0.0005 [97.1–97.6]
Appl. Sci. 2023, 13, 12510 13 of 17

Table 8. Findings of the AU-ROC, AU-PRC, and loss analysis with SD and CI.

Models Loss AU-ROC AU-PRC SD CI

Proposed LC detection model 0.89 0.98 0.98 0.0002 [98.1–98.5]
DenseNet-121 model 1.13 0.97 0.96 0.0003 [96.7–97.4]
Barbouchi et al. model [34] 1.02 0.97 0.97 0.0004 [96.7–97.4]
El-Hamdi et al. model [35] 0.95 0.97 0.96 0.0003 [97.5–98.1]
Goswami and Singh model [38] 1.12 0.96 0.95 0.0004 [96.4–96.8]
EfficientNet model 0.99 0.97 0.97 0.0005 [97.1–97.6]
ShuffleNet model 0.89 0.97 0.96 0.0004 [96.7–97.1]

4. Discussions
The author proposed an LC detection model using this study’s PET/CT images. An
extensive image preprocessing technique was developed to overcome the challenges of
enhancing the image quality. The SyN function was used to normalize the CT and PET
images. Additionally, the author applied resampling, Retinex filtering, and attenuation
techniques to improve the image quality. The image augmentation was performed to
overcome the data imbalance. A CNN model was developed using the DenseNet-121
model’s weights for the feature extraction. Deep autoencoders reduced the dimensionality
of the images. Finally, a CNN model with a MobileNet V3 model weights was employed
for classifying the type of LC. The author improved the performance of the proposed
model by applying a quantization-aware training approach. In addition, the AO-based
hyperparameter optimization supported the proposed LC model to detect the type of LC
with a minimum hardware and software requirement. The experimental results revealed
the significance of the proposed LC detection model in identifying the LC types.
Table 3 outlined the hyper-parameter optimization’s significance in optimizing the
proposed LC detection parameters. AO-based hyper-parameter optimization supported the
model to yield an exceptional outcome compared to Adagrad and RMSProp optimizations.
Table 4 revealed that the proposed model demands minimum computational power to
generate an effective outcome. It required parameters of 2.1 m and FLOPs of 176 m with a
training time of 128 s for identifying the types of LC. Table 5 and Figure 4 highlighted the
number of epochs of the proposed model to deliver a superior outcome. The proposed LC
detection model achieved an accuracy of 98.6 with a batch size of 256. However, it obtained
an accuracy of 98.4 in the 180th batch. Table 6 outlined the performance evaluation findings
of the suggested LC detection model. The proposed model produced an outstanding
result for each class using the PET/CT images. It is evident that the suggested image
preprocessing and augmentation techniques assisted the proposed LC detection model in
identifying the LC types. Table 7 emphasized the key findings of the comparison analysis.
The suggested model surpassed the existing LC detection approaches by reaching an
optimal accuracy and F1-score using minimal resource utilization. Finally, Table 8 showed
that the proposed model obtained a higher AU-ROC and AU-PRC with a minimum loss
of 0.89.
Barbouchi et al. [34] built a transformer-based LC detection model. They developed
a histologic subtype classification to detect the tumor location from the PET/CT images.
However, the vision transformers demand enormous computational resources compared
to the proposed LC detection model. The proposed LC detection model outperformed
Barbouchi et al.’s model with minimum hardware resources.
El-Hamdi et al. [35] proposed an LC detection model and generalized their model
using the Lung-PET-CT-Dx dataset. They built a multi-class CNN model for classifying the
PET/CT images. The VGG-16 network was used to extract the features. Subsequently, a
classifier was employed to classify the extracted features. Similarly, the proposed model
used the DenseNet-121-based feature extractor. In addition, the author constructed an im-
age classifier using the weights of the MobileNet V3-Small model. The findings highlighted
that the proposed LC detection model outperformed the El-Hamdi et al. model.
Appl. Sci. 2023, 13, 12510 14 of 17

Goswami and Singh [37] developed an image classification model for detecting LC
from PET/CT images. They obtained an optimal outcome using the Lung-PET-CT-Dx
dataset. Likewise, the proposed model achieved superior results by employing effective
image processing, feature extraction, and image classification models.
The DenseNet-121-based LC detection model achieved an accuracy of 96.4 with param-
eters of 8.3 and FLOPs of 679. However, the proposed model obtained exceptional accuracy
with limited computational resources. The EfficientNet-based LC detection model obtained
a higher accuracy and F1-score. However, the model may not perform optimally on the
real-time images. It required a considerable computational resource for classifying the
PET/CT images. In addition, it demands additional training time to capture the fine details
of the images. The ShuffleNet-based LC detection model offered an exceptional outcome.
However, the shallower architecture of the ShuffleNet model may limit the LC detection
performance. In contrast, the proposed LC detection model presented an outstanding
result with limited computational resources. The combination of the DenseNet-121 and
MobileNet V3 models yielded better results.
PET/CT imaging offers a complete assessment of lung tissue, effectively highlighting
regions exhibiting elevated metabolic activity while offering valuable anatomical context.
Healthcare practitioners can choose the surgery, chemotherapy, or radiation therapy for
LC by precisely identifying the types of LC. The proposed model was trained to analyze
and extract features from complex images. It can be implemented in edge devices to render
effective healthcare services to individuals. It can facilitate the detection and classification
of various forms of LC. In addition, it can locate tiny lung nodules that human radiologists
may ignore due to the early stage of development. The timely identification of medical
conditions can significantly enhance patients’ overall prognosis and treatment outcomes.
The proposed LC detection model can assist physicians in customizing LC therapies for
each patient. It can minimize the number of FPs and FNs associated with cancer detection.
This intervention enhances the precision of diagnostic assessments.
Furthermore, the proposed model can substantially accelerate the diagnostic proce-
dure. Radiologists can use the proposed model to select cases and focus on complicated
tasks. It can contribute to advancing research and facilitating clinical trials. The proposed
LC detection model generates outcomes in a limited time. It enables healthcare centers to
identify LC in the earlier stages. Healthcare centers can employ the model to detect LC
without professional radiologists. The reduced number of FPs and FNs guaranteed the
optimal accuracy of the proposed model. Additionally, the proposed model is cost-effective
and can be implemented with low computational resources.
However, the author encountered a few limitations while implementing the proposed
model. The PET/CT images contained motion artifacts and metallic artifacts. The low
spatial resolution caused challenges in identifying the crucial patterns. Tissue misalignment
during the scan reduced the image quality. In addition, PET and CT images were obtained
separately. Thus, there was a misalignment due to the functional information of PET
and the anatomical information of CT images. In addition, the dataset samples were
highly imbalanced. Nonetheless, the author employed compelling image preprocessing
techniques, including the SyN function, artifact correction technique, and attenuation
approaches, to improve image quality. The proposed image augmentation technique
supported the proposed LC detection model to overcome the data imbalance limitation.
While implementing the proposed model, the author faced challenges with the graphical
processing unit configuration. The quantization-aware training approach has supported
the author deploying the model while using low computational resources. The lack of
accessible, high-quality labeled data to train DL models is a significant barrier. Generating
massive datasets with precise annotations is essential for LC diagnosis. The research
requires more diversified, publicly available, and well-annotated datasets. The strong
generalization capabilities of DL models can support a physician in treating a wide range
of populations. Maintaining model robustness and performance in clinical contexts is
Appl. Sci. 2023, 13, 12510 15 of 17

challenging. A balance between FP and FN is crucial. PET/CT images contain noises and
artifacts. Extensive image preprocessing was required to improve image clarity.
To enhance the proposed LC detection model, it is essential to integrate technology de-
velopments with clinical expertise, information availability, and compliance standards. By
considering these factors, the proposed model can serve as a significant resource in assisting
healthcare professionals in making efficient and prompt decisions concerning the diagnosis
and treatment of LC. The liquid neural network (LNN) extends the recurrent neural net-
work’s capabilities by including dynamic hidden states for analyzing video and images [45].
It computes the expected outcome and its shifting hidden state at each prediction step. In
the future, LNN-based applications can significantly outperform currently used machine
vision approaches. Using the LNN technique, the feature engineering process can be im-
proved. A set of optimal features can optimize the multi-class classification accuracy. The
proposed model can be extended using LNN and ensemble learning techniques to reduce
the computation time and resources in identifying LC from the multi-modality images.

5. Conclusions
The author built an LC detection model using the DL technique in this study. The pro-
posed LC detection model employed image preprocessing and augmentation techniques to
improve the performance of the proposed model. The SyN function and Retinex filters were
used to remove noise and artifacts. The GANs technique was used to increase the dataset
size. The author developed the DenseNet-121-based feature extraction model for extracting
the critical features. A deep autoencoder model was employed to reduce the number of
features without losing the essential details. A MobileNet V3-Small model-based image
classifier was constructed using quantization-aware training and early stopping strategies.
The author generalized the proposed model using the Lung-PET-CT-Dx dataset. The find-
ings highlighted the significant contribution of the proposed model in identifying the LC
types. The proposed LC detection model outperformed the existing models. It required
limited computational resources for detecting LC from the complex images. Healthcare
centers can deploy the proposed model without any implementation complexities. In
addition, the lightweight architecture of the proposed model offers an opportunity to
integrate it into the edge devices. However, a few limitations were encountered during the
development. There is a lack of highly balanced datasets for providing extensive training
for the LC detection model. The low-quality images caused challenges in classifying the
LC types. As a result, extended training is required to implement the proposed model in
the real-time application. The integration of the LNN technique can improve the process of
feature extraction. In addition, the ensemble learning approach can accurately classify the
medical images. The author will employ the LNN and ensemble learning approaches to
extract crucial features and classify low-quality images.

Funding: This work was supported by the Deanship of Scientific Research, Vice Presidency for
Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia (Grant No. 4702).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Lung-PET-CT-Dx Dataset. The Cancer Imaging Archive. https://doi.
org/10.7937/TCIA.2020.NNC2-0461, accessed on 1 January 2023.
Conflicts of Interest: The author declares no conflict of interest.

References
1. Gu, Y.; Chi, J.; Liu, J.; Yang, L.; Zhang, B.; Yu, D.; Zhao, Y.; Lu, X. A survey of computer-aided diagnosis of lung nodules from CT
scans using deep learning. Comput. Biol. Med. 2021, 137, 104806. [CrossRef]
2. Sharif, M.I.; Li, J.P.; Naz, J.; Rashid, I. A comprehensive review on multi-organs tumor detection based on machine learning.
Pattern Recognit. Lett. 2020, 131, 30–37. [CrossRef]
Appl. Sci. 2023, 13, 12510 16 of 17

3. Adhikari, T.M.; Liska, H.; Sun, Z.; Wu, Y. A Review of Deep Learning Techniques Applied in Lung Cancer Diagnosis. In Signal and
Information Processing, Networking and Computers: Proceedings of the 6th International Conference on Signal and Information Processing,
Networking and Computers (ICSINC), Guiyang, China, 13–16 August 2019; Springer: Singapore, 2020; pp. 800–807.
4. Saba, T. Recent advancement in cancer detection using machine learning: Systematic survey of decades, comparisons and
challenges. J. Infect. Public Health 2020, 13, 1274–1289. [CrossRef]
5. Ozdemir, O.; Russell, R.L.; Berlin, A.A. A 3D probabilistic deep learning system for detection and diagnosis of lung cancer using
low-dose CT scans. IEEE Trans. Med. Imaging 2019, 39, 1419–1429. [CrossRef]
6. Yu, L.; Tao, G.; Zhu, L.; Wang, G.; Li, Z.; Ye, J.; Chen, Q. Prediction of pathologic stage in non-small cell lung cancer using machine
learning algorithm based on CT image feature analysis. BMC Cancer 2019, 19, 464. [CrossRef]
7. Asuntha, A.; Srinivasan, A. Deep learning for lung Cancer detection and classification. Multimed. Tools Appl. 2020, 79, 7731–7762.
[CrossRef]
8. Shakeel, P.M.; Burhanuddin, M.A.; Desa, M.I. Lung cancer detection from CT image using improved profuse clustering and deep
learning instantaneously trained neural networks. Measurement 2019, 145, 702–712. [CrossRef]
9. Xie, Y.; Meng, W.Y.; Li, R.Z.; Wang, Y.W.; Qian, X.; Chan, C.; Yu, Z.F.; Fan, X.X.; Pan, H.D.; Xie, C.; et al. Early lung cancer
diagnostic biomarker discovery by machine learning methods. Transl. Oncol. 2021, 14, 100907. [CrossRef]
10. Sori, W.J.; Feng, J.; Godana, A.W.; Liu, S.; Gelmecha, D.J. DFD-Net: Lung cancer detection from denoised CT scan image using
deep learning. Front. Comput. Sci. 2021, 15, 152701. [CrossRef]
11. Ak, M.F. A comparative analysis of breast cancer detection and diagnosis using data visualization and machine learning
applications. Healthcare 2020, 8, 111. [CrossRef]
12. Pradhan, K.; Chawla, P. Medical Internet of things using machine learning algorithms for lung cancer detection. J. Manag. Anal.
2020, 7, 591–623. [CrossRef]
13. Heuvelmans, M.A.; van Ooijen, P.M.; Ather, S.; Silva, C.F.; Han, D.; Heussel, C.P.; Hickes, W.; Kauczor, H.U.; Novotny, P.; Peschl,
H.; et al. Lung cancer prediction by Deep Learning to identify benign lung nodules. Lung Cancer 2021, 154, 1–4. [CrossRef]
14. Singh, G.A.P.; Gupta, P.K. Performance analysis of various machine learning-based approaches for detection and classification of
lung cancer in humans. Neural Comput. Appl. 2019, 31, 6863–6877. [CrossRef]
15. Sajja, T.; Devarapalli, R.; Kalluri, H. Lung Cancer Detection Based on CT Scan Images by Using Deep Transfer Learning. Trait. Du
Signal 2019, 36, 339–344. [CrossRef]
16. Coccia, M. Deep learning technology for improving cancer care in society: New directions in cancer imaging driven by artificial
intelligence. Technol. Soc. 2020, 60, 101198. [CrossRef]
17. Bhatia, S.; Sinha, Y.; Goel, L. Lung cancer detection: A deep learning approach. In Soft Computing for Problem Solving: SocProS
2017; Springer: Singapore, 2019; Volume 2, pp. 699–705.
18. Nasrullah, N.; Sang, J.; Alam, M.S.; Mateen, M.; Cai, B.; Hu, H. Automated lung nodule detection and classification using deep
learning combined with multiple strategies. Sensors 2019, 19, 3722. [CrossRef]
19. Chaunzwa, T.L.; Hosny, A.; Xu, Y.; Shafer, A.; Diao, N.; Lanuti, M.; Christiani, D.C.; Mak, R.H.; Aerts, H.J. Deep learning
classification of lung cancer histology using CT images. Sci. Rep. 2021, 11, 5471. [CrossRef]
20. Bhandary, A.; Prabhu, G.A.; Rajinikanth, V.; Thanaraj, K.P.; Satapathy, S.C.; Robbins, D.E.; Shasky, C.; Zhang, Y.D.; Tavares, J.M.R.;
Raja, N.S.M. Deep-learning framework to detect lung abnormality—A study with chest X-Ray and lung CT scan images. Pattern
Recognit. Lett. 2020, 129, 271–278. [CrossRef]
21. Kanavati, F.; Toyokawa, G.; Momosaki, S.; Rambeau, M.; Kozuma, Y.; Shoji, F.; Yamazaki, K.; Takeo, S.; Iizuka, O.; Tsuneki, M.
Weakly-supervised learning for lung carcinoma classification using deep learning. Sci. Rep. 2020, 10, 9297. [CrossRef]
22. Hallitschke, V.J.; Schlumberger, T.; Kataliakos, P.; Marinov, Z.; Kim, M.; Heiliger, L.; Seibold, C.; Kleesiek, J.; Stiefelhagen, R.
Multimodal Interactive Lung Lesion Segmentation: A Framework for Annotating PET/CT Images based on Physiological and
Anatomical Cues. arXiv 2023, arXiv:2301.09914.
23. Ardila, D.; Kiraly, A.P.; Bharadwaj, S.; Choi, B.; Reicher, J.J.; Peng, L.; Tse, D.; Etemadi, M.; Ye, W.; Corrado, G.; et al. End-to-
end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat. Med. 2019,
25, 954–961. [CrossRef] [PubMed]
24. Marentakis, P.; Karaiskos, P.; Kouloulias, V.; Kelekis, N.; Argentos, S.; Oikonomopoulos, N.; Loukas, C. Lung cancer histology
classification from CT images based on radiomics and deep learning models. Med. Biol. Eng. Comput. 2021, 59, 215–226. [CrossRef]
25. Chabon, J.J.; Hamilton, E.G.; Kurtz, D.M.; Esfahani, M.S.; Moding, E.J.; Stehr, H.; Schroers-Martin, J.; Nabet, B.Y.; Chen, B.;
Chaudhuri, A.A.; et al. Integrating genomic features for non-invasive early lung cancer detection. Nature 2020, 580, 245–251.
[CrossRef]
26. Zhang, R.; Cheng, C.; Zhao, X.; Li, X. Multiscale mask R-CNN–based lung tumor detection using PET imaging. Mol. Imaging
2019, 18, 1536012119863531. [CrossRef] [PubMed]
27. Kriegsmann, M.; Haag, C.; Weis, C.A.; Steinbuss, G.; Warth, A.; Zgorzelski, C.; Muley, T.; Winter, H.; Eichhorn, M.E.; Eichhorn, F.;
et al. Deep learning for the classification of small-cell and non-small-cell lung cancer. Cancers 2020, 12, 1604. [CrossRef] [PubMed]
28. Polat, H.; Danaei Mehr, H. Classification of pulmonary CT images by using hybrid 3D-deep convolutional neural network
architecture. Appl. Sci. 2019, 9, 940. [CrossRef]
29. Patel, D.; Shah, Y.; Thakkar, N.; Shah, K.; Shah, M. Implementation of artificial intelligence techniques for cancer detection.
Augment. Hum. Res. 2020, 5, 6. [CrossRef]
Appl. Sci. 2023, 13, 12510 17 of 17

30. Baranwal, N.; Doravari, P.; Kachhoria, R. Classification of histopathology images of lung cancer using convolutional neural
network (CNN). In Disruptive Developments in Biomedical Applications; CRC Press: Boca Raton, FL, USA, 2022; pp. 75–89.
31. Levy-Jurgenson, A.; Tekpli, X.; Kristensen, V.N.; Yakhini, Z. Spatial transcriptomics inferred from pathology whole-slide images
links tumor heterogeneity to survival in breast and lung cancer. Sci. Rep. 2020, 10, 18802. [CrossRef]
32. Mishra, S.; Thakkar, H.K.; Mallick, P.K.; Tiwari, P.; Alamri, A. A sustainable IoHT based computationally intelligent healthcare
monitoring system for lung cancer risk detection. Sustain. Cities Soc. 2021, 72, 103079. [CrossRef]
33. Luna, J.M.; Chao, H.H.; Diffenderfer, E.S.; Valdes, G.; Chinniah, C.; Ma, G.; Cengel, K.A.; Solberg, T.D.; Berman, A.T.; Simone,
C.B., II. Predicting radiation pneumonitis in locally advanced stage II–III non-small cell lung cancer using machine learning.
Radiother. Oncol. 2019, 133, 106–112. [CrossRef]
34. Barbouchi, K.; El Hamdi, D.; Elouedi, I.; Aïcha, T.B.; Echi, A.K.; Slim, I. A transformer-based deep neural network for detection
and classification of lung cancer via PET/CT images. Int. J. Imaging Syst. Technol. 2023, 33, 1383–1395. [CrossRef]
35. El Hamdi, D.; Elouedi, I.; Slim, I. Computer-Aided Classification of Cell Lung Cancer Via PET/CT Images Using Convolutional
Neural Network. Int. J. Image Graph. 2023, 2450040. [CrossRef]
36. Li, P.; Wang, S.; Li, T.; Lu, J.; HuangFu, Y.; Wang, D. A Large-Scale CT and PET/CT Dataset for Lung Cancer Diagnosis
(Lung-PET-CT-Dx) [Dataset]. Cancer Imaging Arch. 2020. [CrossRef]
37. Clark, K.; Vendt, B.; Smith, K.; Freymann, J.; Kirby, J.; Koppel, P.; Moore, S.; Phillips, S.; Maffitt, D.; Pringle, M.; et al. The Cancer
Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository. J. Digit. Imaging 2013, 26, 1045–1057.
[CrossRef]
38. Goswami, J.; Singh, K.K. Pulmonary Lung Cancer Classification Using Deep Neural Networks. In Machine Vision and Augmented
Intelligence: Select Proceedings of MAI 2022; Springer Nature Singapore: Singapore, 2023; pp. 395–407.
39. Liu, Z.; Yao, C.; Yu, H.; Wu, T. Deep reinforcement learning with its application for lung cancer detection in medical Internet of
Things. Future Gener. Comput. Syst. 2019, 97, 1–9. [CrossRef]
40. DenseNet-121 Model. Available online: https://github.com/flyyufelix/DenseNet-Keras/blob/master/densenet121.py (accessed
on 1 January 2023).
41. MobileNet-V3 Small Model. Available online: https://github.com/topics/mobilenetv3-small (accessed on 3 January 2023).
42. Bi, L.; Kim, J.; Kumar, A.; Feng, D.; Fulham, M. Synthesis of positron emission tomography (PET) images via multi-channel
generative adversarial networks (GANs). In Molecular Imaging, Reconstruction and Analysis of Moving Body Organs, and Stroke
Imaging and Treatment, Proceedings of the Fifth International Workshop, CMMI 2017, Second International Workshop, RAMBO 2017, and
First International Workshop, SWITCH 2017, Held in Conjunction with MICCAI 2017, Québec City, QC, Canada, 14 September 2017;
Springer International Publishing: Cham, Switzerland, 2017; pp. 43–51.
43. Abid, A.; Sinha, P.; Harpale, A.; Gichoya, J.; Purkayastha, S. Optimizing medical image classification models for edge devices. In
Distributed Computing and Artificial Intelligence, Volume 1: 18th International Conference; Springer International Publishing: Cham,
Switzerland, 2022; pp. 77–87.
44. Dominik, M.; Soto-Rey, I.; Kramer, F. Towards a guideline for evaluation metrics in medical image segmentation. BMC Res. Notes
2022, 15, 210.
45. Ramin, H.; Lechner, M.; Amini, A.; Rus, D.; Grosu, R. Liquid time-constant networks. In Proceedings of the 2021 AAAI Conference
on Artificial Intelligence, Virtual, 2–9 February 2021; Volume 35, pp. 7657–7666.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.