15 Oct, 2024
15 Oct, 2024
Several types of cells are present in the lungs, such as alveolar cells, epithelial cells, and bronchial cells. The uncontrolled division of one or more of these cells results in a tumor, and this is referred to as lung cancer.
There are two types of lung cancer. These are:
Information present in the genes regulates the growth and multiplication of lung cells. However, when there is a change in this genetic information, the cell divides uncontrollably, resulting in lung cancer. It is not accurately known what causes changes in genetic information. However, certain factors, such as active and passive smoking, exposure to harsh chemicals, and a family history of lung cancer, increase the risk of these genetic changes.
Oncologists perform a comprehensive physical examination to determine the reasons for signs and symptoms. The common signs and symptoms of lung cancer are chest pain, persistent cough, hoarseness, shortness of breath, and coughing up blood.
The doctor may also ask about the medical and family health history of the patient. If the physical examination, medical history, and family history suggest the possibility of a lung cancer diagnosis, the doctor may advise the patient to undergo further lung cancer tests.
Patients suspected of lung cancer are apprehensive about how lung cancer can be detected. There are multiple tests available for lung cancer diagnosis. These tests can help doctors arrive at a conclusive diagnosis. Starting from a simple blood test to a biopsy, those suspected of having lung cancer are recommended to undergo a wide array of diagnostic tests for lung cancer.
We often come across this one question: how to detect lung cancer at home? It is important to note that there are no ways to detect lung cancer at home. When one is experiencing lung-related symptoms, it is important to promptly see a doctor.
Doctors recommend the following diagnostic tests for lung cancer. To begin with, doctors may recommend physical examinations and blood tests for lung cancer. If the results indicate cancer, further tests will be recommended. The following are the different types of diagnostic tests for lung cancer:
This imaging technique uses X-rays to obtain an initial view of the lungs to detect abnormalities.
This detailed imaging technique helps in obtaining cross-sectional images of the lungs to check and confirm the presence of tumors and evaluate their size and location.
An MRI scan uses powerful magnetic fields and radio waves to create detailed images of the lungs to evaluate the disease progression.
This scan employs a radioactive tracer to visualize metabolic activity in the lungs, aiding in cancer detection, staging, and treatment planning.
This test involves the microscopic evaluation of phlegm (sputum) for the presence of cancer cells. This test is particularly useful for larger tumors closer to the airways.
This is a definitive test for lung cancer, and it involves the removal of lung tissue samples and their examination under a microscope. Several biopsy methods are used, including fine needle aspiration biopsy, transtracheal FNA or transbronchial FNA, core biopsy, and transthoracic needle biopsy.
Molecular testing involves the genetic analysis of lung cancer cells to identify specific gene mutations or alterations that will be guiding the treatment of lung cancer.
Personalized lung cancer management involves extensive complexities during screening, diagnosing, staging, monitoring, treatment, and follow-up. These complexities develop due to the availability of a significant amount of data based on clinical presentation, tumor genomics, radiographic features, pathophysiology, tumor stage, biopsies, assessment of treatment response, treatment options, and overall outcomes.
Integrating artificial intelligence (AI) in cancer care plays an important role in diagnosing, monitoring, and managing lung cancer. Thus, there is a need to integrate AI technology in lung cancer detection and management that uses the available data and allows oncologists to offer individualized care to their patients suspected of or diagnosed with lung cancer. AI has the answer to the burning question: how to determine lung cancer early during the screening of high-risk patients?
Lung cancer screening, especially in high-risk patients, requires multiple CT scans from different angles. It is difficult to comprehensively review all the images and to detect any pulmonary nodule. AI may play an important role in lung cancer screening. AI in cancer increases the productivity and accuracy of screening methods. A recent study in Radiology (2023) reported that AI-based CAD software helps improve lung cancer detection without increasing exposure to hazardous imaging radiations. Artificial intelligence in oncology may also effectively differentiate between malignant and non-malignant tumors during the screening, which helps oncologists in detecting lung cancer in patients that require further evaluation.
AI also helps in detecting lung cancer risk by considering several factors, including the signs of abnormal growth and alterations in tissue patterns. A study published in Thorax (2020) found that an AI algorithm, the prediction convolutional neural network, improves risk prediction compared to a model advised in UK guidelines. Another study published in the Current Medical Imaging Reviews (2019) reported efficient identification of tumor location by using machine-learning representations. Another important study published in the Journal of Clinical Oncology (2023) found that a model developed using low-dose computed tomography images accurately predicts the future risk of lung cancer from a single scan.
Data indicates that people who smoke are almost 15 to 30 times at increased risk of developing or dying from lung cancer. One of the ways to predict the risk of lung cancer through AI is by evaluating the low-dose CT scan of smokers by an algorithm that uses the prespecified databases. A study published in EBioMedicine (2018) reported that cancer detection through machine learning and urine proteosome not only helps in differentiating the samples from lung cancer patients and healthy individuals but also differentiates the lung cancer samples from other cancers. Evaluating the composition of exhaled breath condensate by AI helps in molecular profiling, cancer diagnosis, screening high-risk individuals, and therapy monitoring. A study published in the Journal of Breath Research (2018) analyzed the breath samples through AI lung cancer detection and differentiated them from other lung diseases. The study found 96.2% for non-smokers and 95.8% for smokers, and the overall specificity was 92.3% for smokers and 90.6% for non-smokers.
Oncologists found it challenging to accurately diagnose the presence of lung cancer primarily due to the reason that several lung diseases mimic the signs and symptoms of lung cancer and vice versa. Reducing human intervention in screening and diagnosing lung cancer potentially reduces over-diagnosis and under-diagnosis. Artificial intelligence in oncology accurately distinguishes between malignant and benign lesions through multivariate logistic regression analysis of the radiological and molecular characteristics resulting in reduced incidences of over-diagnosis and under-diagnosis.
Cancer occurs when the cells divide uncontrollably due to mutation in specific genes. Some of the genes related to lung cancer include ALK, BRAF, EGFR, KRAS, LKB1, MET, PIK3CA, RET, and ROS1. AI can help oncologists to predict the risk of lung cancer by detecting lung cancer-related gene mutations. A study published in Nature Medicine (2018) found that neural network training might predict six mutated genes (EGFR, FAT1, KRAS, STK11, SETBP1, and TP53) through pathological images with accuracy between 73.3% to 85.6%.
There are several ways through which artificial intelligence in oncology assists in discovering new approaches for the management of lung cancer. First, it helps oncologists in developing customized strategies for lung cancer management by detecting the type of mutations and the drugs effective in those mutations. Second, AI detects the new anticancer targets in patients with lung cancer through genomic, proteomic, metabolomics, and epigenetic integration analysis. The analysis allows the development of new drugs by evaluating the molecular interaction between the drug and the novel target.
Pathological biopsies are the confirmatory tests for diagnosing lung cancer. However, as there are several types of lung cancer, it is difficult to determine the pathological type through manual evaluation. AI significantly reduces the incidences of false-negative rates by analyzing abnormal tissues.
Cytopathology assists in early diagnosis and treatment. However, limitations include a small sample size and a high misdiagnosis rate due to manual recognition. The classification accuracy was 85.3% with the antagonistic network (generative adversarial networks, GAN) method, which was 4.5% higher than the classification done without using the GAN-generated images. Thus, AI effectively enhances cytopathological efficiency in diagnosing lung cancer.
Lung tumor subtypes are usually evaluated by histopathological examination. AI improves the overall accuracy of the manual reading in the histopathological diagnosis system. A study published in IEEE Transaction of Image Processing (2020) reported that AI effectively differentiates lung adenocarcinoma from squamous cell carcinoma, and cancerous tumor from adjacent healthy tissues.
Several factors enhance the risk of lung cancer. These factors include smoking tobacco, second-hand tobacco smoking, exposure to asbestos and radon, exposure to inhaled chemicals, such as silica, cadmium, vinyl chloride, and chromium compounds, inhaling diesel exhaust, family history of lung cancer, and history of radiation therapy for other cancers. Smokers who take beta-carotene supplements are also at increased risk for lung cancer.
There has been extensive integration of machine learning in lung surgery. AI assists oncological surgeons in the diagnosis and management of lung cancer, surgical planning, assessing preoperative risk, and predicting the overall surgical outcomes. Researchers have predicted the pulmonary complications of various lung resection surgeries through machine learning. The prediction of the surgical risks assists the surgeons in effectively managing the perioperative risks. AI also helps in predicting lymph node metastasis during or after lung cancer surgery in patients with early-stage adenocarcinoma. AI also predicts post-surgical progression-free survival in patients suffering from lung adenocarcinoma.
The use of AI in cancer is evolving and there will be a widespread impact on early lung cancer detection with artificial intelligence in the future through more advanced algorithms. In the future, these models will be integrated with more comprehensive patient-related data, imaging modalities, electronic health record data, and molecular and histopathology biomarkers to accurately diagnose lung cancer. Despite the various benefits of AI in cancer, there are certain limitations. These include a lack of standardization in lung cancer health data and various barriers to the implementation of AI into clinical practice.
AI plays a crucial role in all the steps of lung cancer care, which include risk prediction, screening, diagnosis, monitoring, assessing surgical complications, and determining the overall outcomes in lung cancer patients. More advanced algorithms involving the use of lung cancer detection using machine learning have been developed to improve the accuracy of lung cancer diagnosis.
AI assists oncologists in lung cancer detection. AI helps detect lung cancer by evaluating computed tomography scans in detail. Several AI models have been developed that analyze the components of the exhaled breath for lung cancer detection.
High-risk patients usually ask doctors about how to detect lung cancer early. In most cases, patients do not present any signs and symptoms of lung cancer at an early stage. People with an increased risk of lung cancer, such as those who smoke heavily, are advised to undergo lung cancer screening at regular intervals. Individuals should not ignore any of the signs and symptoms of lung cancer.
Artificial intelligence is used in all steps of lung cancer management that includes risk prediction, screening, diagnosis, monitoring, treatment, and overall outcomes. In patients with a higher risk of lung cancer, such as smokers or family history of lung cancer, AI detects the gene mutation and predicts the lung cancer risk.
Several AI tools have been developed for lung cancer detection. One such tool is Sybil. The recent study on Sybil indicates that it predicts if the person is at risk for developing lung cancer. Further, several other AI tools, such as deep learning AI models, evaluate lung nodules and detect lung cancer at an early stage.
AI offers several benefits in diagnosing lung cancer. First, it predicts the risk of developing lung cancer. Second, it assists in diagnosing lung cancer at an early stage, which may not be possible with the manual reading of the CT scans. Third, it is accurate and precise, and thus, reduces the incidences of over-diagnosis and under-diagnosis. Fourth, it improves overall lung cancer outcomes as early diagnosis and treatment increase survival.
There are several lung cancer tests available. These include comprehensive physical examination, imaging tests, such as computerized tomography scans, magnetic resonance imaging, and positron emission tomography, sputum cytology, and tissue biopsy.
Artificial intelligence in lung cancer helps oncologists in developing optimal treatment strategies. It accurately differentiates between malignant and benign lung tumors and assists in early diagnosis and treatment. It also distinguishes between different subtypes of lung cancer and offers customized treatment options for lung cancer patients.
