Media coverage / 17 Aug, 2026

AI-Assisted Lung Cancer Screening: Early Detection Improves Survival Chances

AI-Assisted Lung Cancer Screening: Early Detection Improves Survival Chances

Lung cancer is the leading cause of cancer death in Hong Kong, and many patients are already at an advanced stage when diagnosed. Clinical oncologist Dr Daryn Chow Yi Ler points out that this is not simply due to insufficient public awareness of cancer prevention, but is related to the pathology of lung cancer. Early-stage lung cancer usually presents no specific symptoms because lung tissue lacks sensory cells. By the time symptoms such as shortness of breath, coughing up blood or weight loss appear, the disease is often already advanced. Currently, Hong Kong does not have a universal lung cancer screening programme. Screening is mainly recommended for high-risk individuals, such as long-term smokers or those with a family history, and should be discussed with a doctor as a personalised decision.

The Importance of Lung Cancer Screening

Implementing effective lung cancer screening has a significant impact on patient outcomes and overall healthcare resources. Dr Chow notes that a well-designed screening programme can help detect asymptomatic tumours early and enable timely treatment. According to Hong Kong data, the five-year survival rate for early-stage tumours exceeds 70%, but drops to around 7% for stage IV advanced disease—a substantial difference. Early-stage patients may require only local treatment such as surgery or radiotherapy, whereas advanced cases need comprehensive treatment and are managed as a chronic condition, resulting in higher costs and greater pressure on families.

Current Screening Methods and Limitations

The primary method for lung cancer screening at present is low-dose computed tomography (CT), which identifies lung shadows or nodules and determines whether further investigations are needed. However, low-dose CT is highly sensitive and may be able to screen out even benign shadows, leading to risks of overdiagnosis or overtreatment. Therefore, the decision to undergo screening remains a personal one. The doctor’s interpretation of the scan results is also very important.

How Artificial Intelligence Assists in Lung Cancer Imaging Screening

Artificial intelligence plays a supportive role in lung cancer imaging screening. Dr Chow explains that AI can analyse X-ray or CT images, automatically detect abnormal shadows or nodules, and measure their size for doctors’ reference. The Hospital Authority in Hong Kong already uses AI related applications to assist doctors in accident and emergency departments or clinics with triage, as well as for academic research. Compared with traditional methods, AI’s greatest advantages are speed and consistency. It can process large volumes of images simultaneously, assisting to accelerate diagnosis; reliably identify different shadows and measure their size; and, if it judges a scan as normal, significantly reduce the workload of radiologists, streamlining classification processes. Screening with low-dose CT combined with AI is not yet widespread, but once systematic screening steps are in place, AI will be a valuable tool for improving efficiency.

Further Investigations and Treatment After Detecting Nodules

If screening detects a lung nodule or abnormality, further investigations depend on characteristics such as size and location. Larger nodules may require needle biopsy or tissue sampling via endoscopy or external navigation. Once cancer is confirmed, additional tests such as positron emission tomography (PET) or magnetic resonance imaging (MRI) may be needed to determine the stage and suitable treatment options.

After diagnosis, these tests help formulate personalised treatment plans. Dr Chow states that most lung cancer patients first need classification of the pathological type, such as small-cell or non-small-cell lung cancer, as well as staging. Early-stage patients typically receive local treatment such as surgery or radiotherapy, which may be followed by adjuvant chemotherapy or targeted therapy. Advanced cases may require immunotherapy or other targeted drugs. Genetic testing is also important to identify suitability for drugs targeting EGFR, ALK and other mutations. Targeted therapies have advanced rapidly in recent years, expanding from early options such as EGFR and ALK to include NTRK, BRAF and others, with development from first-generation to third-generation drugs, significantly increasing treatment choices.

Future Outlook for AI in Lung Cancer Screening and Treatment

Looking ahead, the application of artificial intelligence in cancer screening and diagnosis will extend beyond image analysis. Dr Chow believes AI can integrate patients’ electronic medical records, imaging, pathology reports and emerging ctDNA (circulating tumour DNA in blood) data to help assess risk levels and design more personalised treatment plans, such as adjusting the duration of adjuvant therapy.

In summary, artificial intelligence is providing faster and more consistent support for lung cancer screening. Combined with low-dose CT, it has the potential to improve early detection rates. High-risk individuals who need screening advice may consult their family doctor or a specialist in family medicine. Patients who have been diagnosed or require treatment can discuss suitable options in detail with their attending clinical oncologist.

 

More information about lung nodules: What Are Lung Nodules? | A Comprehensive Guide to Causes, Symptoms, and Treatment Options | OncoCare

About lung health screening: Lung Cancer Screening: Don't Delay | About more infomration and booking

About genetic testing of multi-cancer early screening: Multi-cancer Early Screening (Genetic Testing)

About lung cancer treatment: Lung Cancer Treatment | OncoCare


Frequently Asked Questions Q&A

Q1: Why is lung cancer often diagnosed at an advanced stage?

A: Early-stage lung cancer usually has no specific symptoms because lung tissue lacks sensory cells. Symptoms such as shortness of breath, coughing up blood or weight loss often appear only when the disease is already advanced.

Q2: Does Hong Kong currently have a universal lung cancer screening programme?

A: At present, Hong Kong does not have a universal lung cancer screening programme. Screening is mainly targeted at high-risk individuals, such as long-term smokers or those with a family history, and should be discussed with a doctor.

Q3: What are the limitations of low-dose CT scanning?

A: It is highly sensitive and may be able to screen out even benign shadows, carrying risks of overdiagnosis or overtreatment.

Q4: How does artificial intelligence assist doctors in lung cancer screening?

A: AI can automatically detect abnormal shadows or nodules and measure their size, speeding up processing. If it judges a scan as normal, it can also reduce the workload of radiologists.

Q5: What should be done after a lung nodule is found?

A: Depending on characteristics such as size and location, needle biopsy or tissue sampling may be required. Further tests such as PET scanning may follow if cancer is confirmed.

Q6: What recent advances have there been in targeted therapies?

A: Options have expanded from early targets such as EGFR and ALK to include NTRK, BRAF and others, with development from first-generation to third-generation drugs, significantly increasing choices. Patients who require treatment should discuss suitable treatment options in detail with their attending clinical oncologist.

Q7: What future developments are expected for AI in lung cancer treatment?

A: AI may integrate medical records, imaging, pathology reports and ctDNA data to help assess risk and design more personalised treatment plans.

 

Related Media Coverage (Traditional Chinese Version Only):

Search