The Complete Guide to Lung Cancer Detection and How AI could Improve Outcomes
More than 1 million* people worldwide die from lung cancer each year, making it the most prevalent cancer-related cause of death in both men and women. More people die of lung cancer than of the next three leading causes of cancer death combined: breast, colon, and prostate cancer.
Despite advances in surgical techniques and combined therapies, lung cancer remains a disease with a dismal prognosis. Although 1-year survival has improved over the past few decades, overall 5-year survival has remained relatively unchanged at 12% to 16% over the past 30 years. These findings highlight the necessity of creating novel treatment and diagnostic technologies to combat lung cancer.
Smoking is the primary risk factor in the majority of lung cancer patients, accounting for 90% of lung cancer incidences in men and 79% in women. The amount of tobacco use and the risk of lung cancer are correlated in a dose-dependent pattern. Therefore, lung cancer represents the most preventable cancer, and a strong emphasis should be focused on preventing the uptake of smoking as well as smoking cessation. In individuals who do stop smoking, the risk of developing lung cancer gradually falls for about 15 years to about twice that of someone who never smoked.
Lung cancer risk is also increased by exposure to asbestos and a number of other substances, including arsenic, beryllium, cadmium, nickel, and radon.
The majority of the exposure occurs in workplace environments. In the construction business, asbestos was frequently utilized as a fireproofing material. As a result, most lung cancer cases were noted among roofers, plumbers, and builders in addition to those working in the mining and processing of asbestos.
Types of lung cancers
Small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), with is most frequent forms (squamous and adenocarcinoma) are the two main subtypes of lung cancer (SCLC). Squamous cell carcinoma is believed to develop in stages, starting with hyperplasia and progressing through squamous metaplasia and dysplasia. Dysplasia does not always develop into cancer, and the process itself could be relatively reversible.
Squamous cell carcinomas are centrally located tumors, which macroscopically may appear white or gray. Adenocarcinomas are more common in the periphery of the lung. Adenocarcinomas can arise in a variety of ways, including macroscopically tiny peripheral lesions, larger regions of consolidation, and nodules that are more solid.
What are the symptoms of lung cancer?
Early-stage lung cancer is frequently asymptomatic. Clinical manifestations include a prolonged cough, production of sputum, dyspnea, and pain in the chest. Consequently, the majority of patients at clinical presentation have advanced disease. Hemoptysis can be seen as blood streaks within the sputum, fresh blood, or clots, and typically prompts the patients to seek medical care.
It’s possible for the tumor to gradually invade nearby tissues like the ribs and vertebral segments, which can result in intense, ongoing pain. Effusions and accompanying dyspnea may result from extension to the pleura or pericardium. Recurrent pneumonia and may be caused on by an obstructive bronchial lesion. Adrenal glands, bone, the brain, the liver, and the skin are commonly sites of lung cancer metastasis. Particularly in individuals with liver metastases, weight loss is frequently a sign of metastatic illness.
More than 20% of body weight lost in the previous month is typically an indicator of bad prognosis. Patients with brain metastases may appear with neurological symptoms such as hemiplegia or lack of coordination, disorientation, personality abnormalities, and epileptic seizures. Small cell carcinoma is one form of lung cancer that is more likely to develop brain metastases. Up to 25% of patients may have involvement in the supraclavicular and anterior cervical lymph nodes, which should be regularly checked when evaluating lung cancer patients. The initial examinations mostly consist of a chest X-ray and a thorax and upper abdomen CT scan. This typically establishes whether there is a pulmonary opacity that needs additional examination. Typically, biopsy from the pathological region is necessary for diagnosis. It is crucial to think about diagnosis and staging simultaneously.
Classification and stages of Lung Cancer
The current staging of lung cancer is based on the TNM classification.
TUMOR: T STAGING
The T staging is based on tumor size, location, and involvement of local structures. Infiltration of the heart and vertebral body confers a staging of T3.
NODAL: N STAGING
The N staging is based on lymph node involvement. The new unified lymph node match describes the lymph node as in both stations and zones. Contralateral or supraclavicular lymph nodes have the highest grading.
METASTASES: M STAGING The M staging has now been classified into three groups according to organ involvement and number of sites of involvement.
How to diagnose lung cancer?
The computed tomography of the chest and upper abdomen offers important details regarding the tumor’s features, closeness to critical organs, size of the lymph nodes, and potential metastatic condition. A PET-CT can reveal important details about the progression and activity of the suspected cancer. Even though PET-CT is more specific than CT scanning, pathological biopsy is still necessary to determine whether mediastinal lymph nodes are implicated. The scan can also be used to show extra thoracic disease involvement in other organs, such as the bones or the adrenal glands. When there is no central endobronchial component or mediastinal lymph nodes present, a CT-guided biopsy may be necessary to get a tissue diagnosis for nodules or masses.
The area of medicine will undergo significant change as a result of artificial intelligence. The availability of digital data, machine learning, and computer infrastructure have allowed AI applications to develop into fields that were previously thought to be inaccessible to machines. In the last several years, the domains of speech recognition, picture identification, and caption creation have all seen rapid advancements thanks to deep learning, an AI approach. Over the next ten years, it is anticipated that the use of technology will significantly increase the depth, quality, and value of the radiological contribution to patient treatment and community health.
AI and healthcare
The adoption of computed tomography for lung screening has increased the quantity of information that radiologists must evaluate, while automated image processing allows for the interpretation of more studies. Even Google AI researchers are attempting to scan CT pictures for lung cancer using a deep learning model*. Their outcomes were encouraging and on par with or better than radiologists’ performance. Future revolutionary improvements in the area of radiology are predicted by technologies used to identify the best possible combinations of biomarkers, such as molecular or image-based biomarkers.
ML is a part of AI and it is a great algorithm to find millions of microcirculatory genomics fragments that are more sensitive and earlier to be found compared with traditional methods. By focusing on chemicals on tumor cells and creating potent blockers to interfere with the carcinogenesis process, targeted treatment is a new strategy for treating cancer with the least amount of harm to normal tissues.
Determining which location changed and what happened in the mutation region is an important step in the development. A smart system might provide a wide range of information and allow for a thorough inquiry. Additionally, an intelligent computer could create a linked model to locate and highlight a targeted gene or associated biological pathway. The use of AI to understand and cure lung cancer is growing, but there are still many challenges to be solved. To realize the goal of individual treatment, more sensitive and precise machines or algorithms are needed.
Lung nodules detection with XVision
XVision’s software offers a recognition algorithm that enables radiologists to better detect lung nodules with a diameter between 3 and 30 mm. The results are displayed using an easy-to-read User-Interface allowing higher accuracy in a shorter amount of time. By highlighting the existence of nodules on each slice and individually measuring the diameter and volume of the ones that are found, XVision improves the total time needed to diagnose each patient.
Our CT solutions can be customized to be easily integrated into the hospital infrastructure, allowing a wide and efficient scaling. As a result, medical professionals are empowered to identify high-risk individuals and notify them, by detecting possible malignant nodules during regular check-ups or glancing over historical medical data. We’re enabling seamless process optimization, from preventive monitoring to early detection, emergency care and much more, all from a single CT.
You can also read about how XVision is helping lung cancer detection by clicking here.
*Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. https://doi.org/10.3322/caac.21492.
*Application of Artificial Intelligence in Lung Cancer Hwa-Yen Chiu 1 2 3 4, Heng-Sheng Chao 1 5, Yuh-Min Chen– DOI: 10.3390/cancers14061370