Lung Cancer Detection using Machine Learning

Lung cancer is a malicious tumor that originates in the respiratory tissues when the cells lining the airways of the lungs proliferate and expand uncontrollably. While computed tomography (CT) scans are capable of illuminating problematic areas, they are not proficient at independently diagnosing lung tumors. The task of accurately determining the nodule distribution within the lung CT scans through automatic identification is challenging. As a consequence, the delineation of lung CT images facilitates the identification and classification of lung tumors. This study involved the acquisition of 1097 lung CT image datasets from the Iraq-Oncology Teaching Hospital/National Centre for Cancer Diseases (IQ-OTH/NCCD). In the initial stage of the work, lung CT images are segmented to find malignancies utilizing the U-Net and Attention Gate Residual U-Net (AGRes U-Net) algorithms. The Intersection over Union (IoU) and binary focal loss metrics are applied to evaluate the segmented lung images. AGRes U-Net achieves 97% greater accuracy than the standard U-Net architecture, as determined by performance evaluation. Subsequently, a YOLOv5 network is implemented to annotate the segmented lung CT images with lesions. Thus, the segmented lung CT outputs are labelled for tumors by the YOLOv5 network are fed into the VGG-19 architecture. With an accuracy of 94.8%, this VGG-19 framework distinguishes the lung tumors as normal, benign, and malignant. To compare the performance of the segmented lung output to that of the classified output by means of a convolutional neural network (CNN), the labelled lung CT images are fed as input to a CNN model in the second phase of the work. In conjunction with the Adaptive Moment Estimation (Adam) optimizer, the CNN model is implemented. The Adam optimizer demonstrates a 98% accuracy rate for the classified tumor outputs. The study illustrates that the accuracy of AGResU-Net segmentation for lung cancer detection and classification, at 97%, is nearly equivalent to that of the CNN model, which achieves 98% accuracy.

Chapter 26 - Optimizing CNN architecture using whale optimization algorithm for lung cancer detection

Lung cancer, with the highest morbidity and mortality in 21th century, is difficult to cure because people often fail to detect it in time. With the blossom of artificial intelligence and big data research, people seem to see the dawn of solving this problem. Computer aided diagnosis (CAD) system of lung cancer based on deep learning techniques have been a popular research topic, which aims at utilize the computer system to analyze the possibility of cancer which is beneficial to the treatment in early stages and increase the possibility of recovery. Here we propose an integrated network system based on 3D Convolutional Neural Network (CNN), consisting mainly of two different 3D CNNs. And finally we reached 82.89% accuracy by combining 3D CNNs' feature outputs and using the softmax activation function to get the final disease probability. Also, we modified the decision making method to make the final result tending to get the negative output to decrease the false positive rate and the true positive rate reached 90.38% eventually.

Lung cancer is a serious type of cancer, leading to increased mortality to death in both men and women as the symptoms are noticed only at later stages. Life span of individuals' may be extended if lung cancer is detected in its early stages. One of the imaging modalities used to diagnose lung cancer is computed tomography (CT). A nodule or mass is a small abnormal growth observed in a lung CT scan that, in most cases, may turn out to be benign. A computer‐aided system is essential to help physicians in precisely diagnosing the disease. The main objective of this work is to detect and classify the nodules in the lung CT scan images as benign or malignant. A context‐aware attention UNET architecture is proposed to segment the nodule from the lung CT scan image. Further, the segmented nodule is classified as benign or malignant using a Convolutional Neural Network architecture. The experiments are performed using the LUNA 16 and LIDC‐IDRI lung CT scan image datasets. From the results obtained, it is observed that the context‐aware attention UNET shows a noteworthy improvement in the following metrics: Dice Score, Sensitivity, Specificity, and F‐Measure. A significant improvement is obtained compared to the existing systems in detecting the lesion as a benign nodule or malignant nodule. Further, an ablation study is performed to validate the significance of each component in the architecture. The experimental results have reported 98.81% and 99.15% for specificity and sensitivity, respectively, and therefore the proposed system has potential clinical value in the detection of lung cancer.

The deadliest red and prevalent disease responsible for huge death cases every year is cancer. In various cancers, the highly predominant cancer-associated towards most incredible mortality ratio is lung cancer. Thereby, to identify or detect lung cancer, the utilization of computed tomography (CT) scans is essential as that presents a comprehensive image of tumor located in the body and its evolution tracking. However, even though CT prefers different imaging modalities, the visual interpretation of such CT scan images may result in erroneous function, causing lag in detecting lung cancer. Consequently, image processing methods have been extensively applied in the medical domain intended for the detection of lung tumor in formative years. Over the past years, extensive studies have been effectuated towards the CAD system for lung cancer detection employing CT images. Generally, the entire process involves four stages: pre-processing or lung segmentation, nodule detection, nodule segmentation, and classification. The study presents performance execution and investigation of the image processing technique for the detection of lung cancer. In this research, a detection technique of lung cancer attributed to image segmentation is proposed since it is halfway in image processing. Further, the approach is extended to Marker control watershed and region growing methodology for segmenting of CT scan image. With the image enhancement followed towards the phases of detection. In this light, the detection segments outcome using Gabor filter, image segmentation, and features extraction. With the assistance of experimental results, it is observed that the efficacy of the proposed approach has reached the highest mark for detecting foremost topographies of lung cancer through the watershed by masking technique attaining extreme precision and vigorous.

Computer Aided Diagnosis (CAD) systems for automatic detection of pulmonary diseases and lung cancer mainly depend on the segmentation of different pulmonary components like right and left lung lobes, airways, vessels, and nodules from the medical imaging modalities like CTs, MRIs, etc. Lung segmentation and nodule segmentation are the important steps to detect any lung related abnormalities. It requires many image processing operations to be performed on the medical images. Computed Tomography (CT) imaging is the most preferred modal because of its popularity, ease of use, and capability of showing different anatomical structures of thorax region. This review paper includes a study of various state of the art techniques explaining the methods applied on CT scans to find the ROIs along with their segmentation accuracies parameters in terms of similarity coefficient, mean error, and overlap ratio.

Lung cancer is considered one of the leading causes of death all across the world. Various radiology-related fields increasingly have used Computer-aided diagnosis (CAD) systems. It just has already become a part of clinical work for lung cancer detection. In this article, we proposed an Adaptive Boost-based Grid Search Optimized Random Forest (Ada-GridRF) classifier that best optimized the hyperparameters of the base random forest model to identify the malignant and non-malignant nodules from the trained CT images. Improved performance speed and reduced computational complexity were the advantages of the proposed method. The proposed methodology was compared with other hyperparameter optimization techniques and also with different conventional approaches. It even outperformed the popular state-of-the-art deep learning techniques such as transfer learning and convolutional neural network. The experimental results proved that the proposed method yielded the best performance metrics of 97.97% accuracy, 100% sensitivity, 96% specificity, 96.08% precision, 98% F1-score, 4% False positives rate, and 99.8% Area under the ROC curve (AUC). It took only 8 msec to train the model. Thus, the proposed Ada-GridRF model can aid radiologists in fast lung cancer detection.

The detection of lung and colon cancer is a critical challenge in medical diagnosis, and machine learning (ML) and deep learning (DL) techniques are increasingly being used to enhance accuracy and efficiency. This review focuses on the integration of ML and DL methods for the combined detection of lung and colon cancer, emphasizing their strengths, limitations, and future potential. The motivation behind this study is to address the growing demand for accurate and early detection of these cancers, which significantly impacts treatment outcomes. Current methods often struggle with feature complexity, image variability, and computational intensity, which limit their real-world applicability. The aim is to consolidate various ML and DL techniques that have been employed for this purpose, highlighting how hybrid models can improve detection rates. The objective of this review is to provide a comprehensive analysis of different methodologies, their datasets, pre-processing techniques, feature extraction methods, and evaluation parameters. This review also explores recent advancements, such as transfer learning combined with fine-tuning techniques, which can further optimize performance in cancer detection. The findings suggest that while current methods show promise, further improvements in model generalization, interpretability, and computational efficiency are required to overcome existing limitations and expand clinical use.

Lung cancer is one of the dangerous diseases that cause huge cancer death worldwide. Early detection of lung cancer is the only possible way to improve a patient’s chance for survival. This study presents an innovative automated diagnosis classification method for Computed Tomography (CT) images of lungs. In this paper, the CT scan of lung images was analyzed with the multiscale convolution. The entire lung is segmented from the CT images and the parameters are calculated from the segmented image. The use of image processing techniques and identifying patterns in the detection of lung cancer from CT images reduces human errors in detecting tumors, and speeds up diagnosis time. Artificial Neural Network (ANN) has been widely used to detect lung cancer, and has significantly reduced the percentage of errors. Therefore, in this paper, Convolution Neural Network (CNN), which is the most effective method, is used for the detection of various types of cancers. This study presents a Multiscale Convolutional Neural Network (MCNN) approach for the classification of tumors. Based on the structure of MCNN, which presents CT picture to several deep convolutional neural networks with different size and resolutions, the classical handcrafted features extraction step is avoided. The proposed approach gives better classification rates than the classical state of the art methods allowing a safer Computer-Aided Diagnosis of pleural cancer. This study reaches a diagnosis accuracy of [Formula: see text] using multiscale convolution technique, which reveals the efficiency of the proposed method.

Cancer is the second leading cause of death worldwide, and within this type of disease, lung cancer is the second most diagnosed, but the leading cause of death. Early detection is crucial to increase patient survival rates. One of the primary methods for detecting this disease is through medical imaging, which, due to its features, is well-suited for analysis by deep learning techniques. These techniques have demonstrated exceptional results in similar tasks. Therefore, this paper focusses on analyzing the latest work related to lung cancer detection using deep learning, providing a clear overview of the state of the art and the most common research directions pursued by researchers. We have reviewed DL techniques for lung cancer detection between 2018 and 2023, analyzing the different datasets that have been used in this domain and providing an analysis between the different investigations. In this state-of-the-art review, we describe the main datasets used in this field and the primary deep learning techniques used to detect radiological signs, predominantly convolutional neural networks (CNNs). As the impact of these systems in medicine can pose risks to patients, we also examine the extent to which explainable AI techniques have been applied to enhance the understanding of these systems, a crucial aspect for their real-world application. Finally, we will discuss the trends that the domain is expected to follow in the coming years and the challenges that researchers will need to address.

BackgroundCurrently, no satisfactory biomarkers are available to screen for lung cancer. Surface-Enhanced Laser Desorption/ionization Time-of- Flight Mass Spectrometry ProteinChip system (SELDI-TOF-MS) is one of the currently used techniques to identify biomarkers for cancers. The aim of this study is to explore the application of serum SELDI proteomic patterns to distinguish lung cancer patients from healthy individuals.MethodsA total of 208 serum samples, including 158 lung cancer patients and 50 healthy individuals, were randomly divided into a training set (including 11 sera from patients with stages I/II lung cancer, 63 from patients with stages III/IV lung cancer and 20 from healthy controls) and a blinded test set (including 43 sera from patients with stages I/II lung cancer, 41 from patients with stages III/IV lung cancer and 30 from healthy controls). All samples were analyzed by SELDI technology. The spectra were generated on weak cation exchange (WCX2) chips, and protein peaks clustering and classification analyses were made using Ciphergen Biomarker Wizard and Biomarker Pattern software, respectively. We additionally determined Cyfra21-1 and NSE in the 208 serum samples included in this study using an electrochemiluminescent immunoassay.ResultsFive protein peaks at 11493, 6429, 8245, 5335 and 2538 Da were automatically chosen as a biomarker pattern in the training set. When the SELDI marker pattern was tested with the blinded test set, it yielded a sensitivity of 86.9%, a specificity of 80.0% and a positive predictive value of 92.4%. The sensitivities provided by Cyfra21-1 and NSE used individually or in combination were significantly lower than that of the SELDI marker pattern (P < 0.005 or 0.05, respectively). Based on the results of the test set, we found that the SELDI marker pattern showed a sensitivity of 91.4% in the detection of non-small cell lung cancers (NSCLC), which was significantly higher than that in the detection of small cell lung cancers (P < 0.05); The pattern also had a sensitivity of 79.1% in the detection of lung cancers in stages I/II.ConclusionThese results suggest that serum SELDI protein profiling can distinguish lung cancer patients, especially NSCLC patients, from normal subjects with relatively high sensitivity and specificity, and the SELDI-TOF-MS is a potential tool for the screening of lung cancer.

The purpose of this study was to compare the diagnostic value of 11C-choline positron emission tomography (PET) and [18F]fluorodeoxyglucose (FDG) PET imaging in the detection of primary lung cancer and mediastinal lymph node metastases. Seventeen patients with histologically proven primary lung cancer were examined with both 11C-choline and FDG PET within a week of each study. Lung cancers were analysed visually and semiquantitatively using the ratio of tumour-to-normal radioactivity (T/N ratio) and standardized uptake value (SUV). Mediastinal lymph node metastases were analysed visually. Although both techniques delineated focal lesions with an increase in tracer accumulation in 13 patients, FDG PET identified three additional patients in whom 11C-choline PET did not visualize any lesion. In the detection of lung cancer <2 cm in size, FDG PET provided higher sensitivity (six of seven, 85.7%) than 11C-choline PET (four of seven, 57.1%). The T/N ratio and SUV were significantly higher with FDG PET (T/N ratio, 7.43+/-6.22; SUV, 4.05+/-3.05) than these were with 11C-choline PET (T/N ratio, 2.93+/-1.19; SUV, 2.93+/-0.79) (P<0.001). There was a significant positive correlation between the T/N ratios and SUVs of FDG and 11C-choline. In the assessment of mediastinal lymph node involvement, FDG PET detected lymph node metastases in two patients who were negative on 11C-choline PET, whereas both techniques could not detect tumour involvement in one patient. Both techniques have clinical value for the non-invasive detection of primary lung cancer that is 2 cm or greater in size. However, FDG PET is superior to 11C-choline PET in the detection of lung cancer that is less than 2 cm in diameter and in mediastinal lymph node metastases.

Lung Cancer in today’s world is one of the major widespread dangerous diseases which is the subject of maximum deaths every year. Accurate detection of lung cancer could boost the endurance rate. Medical image processing has a significant impact on the recognition of lung tumors using Computer Tomography (CT) scan images. Images from a CT scan are widely used because they provide comprehensive imaging of tumor progression inside the lungs. Although different types of noise might be experienced while doing CT scans, producing it a monotonous task for recognizing tumors in the lung. Elimination of noise in CT images is a challenging task for medical diagnoses. The presence of noise in an image is inevitable. Hence reducing noise from the CT scan image is critical for further analysis. Hence various filtering techniques have been used that denoise and enhance the image and help in further evaluation of CT images for accurate lung cancer detection. This paper analyses the noises of different kinds in the CT images and different noise removal techniques which help in improving the accuracy of segmentation and feature extraction as they remove unwanted noise and contribute to the accurate detection of lung cancer. The various filtering methods are analyzed with salt along with pepper noise and speckle noise. The performances of different filters are computed in terms of metrics for evaluation like PSNR, SSIM, MSE, and SNR. The experimental results show that the median filter is more efficient in comparison to other filtering methods in eliminating noises that exist in lung CT images by owning fewer MSE values of 214.8522, high SNR value of 19.36304, SSIM value 0.595997, and high PSNR value of 24.80941.

The incidence and mortality rates of lung cancer have been increasing in many countries. However, the true effectiveness of screening for lung cancer is still controversial. This study aimed to examine the growth pattern of lung cancer and to evaluate the efficacy of screening. The authors linked the records of annual radiologic screening to cancer registry data and conducted a retrospective follow-up study of radiographs in all patients with lung cancer arising in a population undergoing screening. Among a total of 305,934 participants, screening detected 206 lung cancers, 103 of which were Stage I disease. Seventy-one of the 131 adenocarcinomas were Stage I, and 58% of them showed evidence of cancer for 2 years on a retrospective review of radiographs. Presentation as small faint lesions overlapping the normal chest structures delayed the early detection of adenocarcinoma. The overall sensitivity of screening was 70%, 52% for squamous cell carcinoma and 50% for small cell carcinoma. Rapidly growing Stage II-IV tumors without retrospective evidence of cancer on previous radiographs accounted for most of the cancers detected during the intervals between screening. Both the low detectability of Stage I adenocarcinoma and the late recognition of rapidly growing small cell and squamous cell carcinomas reduced the effectiveness of screening. More effective imaging methods and an antismoking campaign are required to reduce lung cancer mortality.

Cancer is one of the most deadly diseases that results in death. Lung cancer is the most common type of cancer and the main cause of death from cancer. Lung cancer survival rates are highly dependent on early identification and staging of the tumor. Image processing technique has a major role in the area of medical, in the detection of diseases. Computer Aided Diagnosis (CAD) systems were created to identify lung cancer in its early stages. This research aims to improve lung cancer identification performance in terms of accuracy, range and responsiveness. This work demonstrates a computerized method in detecting lung cancer using computed tomography images. The algorithm for the detection of lung cancer identification has steps such as image pre-processing, which is carried out by median filter, next is to obtain the region of interest using watershed segmentation and feature extraction. Then the SVM classifier and the KNN classifier are used in the detection of lung cancer.