"Yunyan" AI System Enables Precise Benign-Malignant Differentiation of Pulmonary Nodules, Empowering Early Lung Cancer Diagnosis for Lifelong Patient Benefit

From the detection of multiple ground-glass nodules in her lungs last October to undergoing minimally invasive surgery in December, Ms. Yang, a resident of Guangzhou, endured more than a month of intense anxiety. Although she has finally breathed a sigh of relief, she remains apprehensive. Postoperative pathological examination revealed adenocarcinoma in situ in the right lung, with a tumor size of 7 mm and no evidence of lymph node metastasis. Had timely surgical intervention not been performed, the consequences would have been dire.

“I consulted many experts, all of whom recommended a three-month observation period. However, I was experiencing significant psychological distress. I wanted to undergo surgery for a prompt resolution, yet I also wished to adhere to medical advice and avoid overtreatment. Thanks to the WeDoctor Pulmonary Nodule Diagnosis and Treatment Center, which assessed an 80% probability of malignancy, I made the decisive choice to proceed with surgery. I am deeply frightened when reflecting on what might have happened had I not undergone surgery and the condition had progressed.” Although Ms. Yang has recovered well physically after the operation, her emotional state has not yet fully stabilized.

Ms. Yang was undoubtedly fortunate to receive surgical treatment at an early stage.

The results of the “Analysis of the Epidemiology of Malignant Tumors in China, 2015,” released in China, showed that as of 2015, lung cancer ranked among the highest in both incidence and mortality rates; even with targeted therapy and immunotherapy, the five-year survival rate reached only 19.7%.

If lung cancer is diagnosed and surgically treated at an early stage (carcinoma in situ and stage IA lung cancer), the 10-year survival rate can be as high as 90%. However, early-stage lung cancers are mostly less than 1 cm in diameter, usually asymptomatic and occult, and approximately two-thirds of patients are already at an advanced stage when the disease is detected. Under current medical conditions, physicians rely on manual interpretation of imaging to assess whether pulmonary nodules are benign or malignant; accurate evaluation of small nodules smaller than 10 mm remains a global challenge, and clinicians generally recommend follow-up observation.

As Ms. Yang’s CT scan revealed, multiple ground-glass nodules were present in the apicoposterior segment of the left upper lobe, the lingular segment of the left upper lobe, and the apical segment of the right upper lobe, with the largest nodule located in the apical segment of the right upper lobe, measuring approximately 3–9 mm in diameter. After the detection of these pulmonary nodules, Ms. Yang consulted several specialists, all of whom recommended follow-up observation. Anxious about the potential malignancy of what she perceived as a “ticking time bomb” within her body, she eagerly sought more accurate diagnostic methods. Through a relative’s recommendation, she contacted the WeDoctor Pulmonary Nodule Diagnosis and Treatment Center and underwent the PNapp 5A diagnostic and therapeutic protocol. This fortuitous encounter ultimately changed Ms. Yang’s fate.

The WeDoctor Pulmonary Nodule Diagnosis and Treatment Center is directed by Professor Bai Chunxue, Chair of the Chinese Alliance for Lung Cancer Prevention and Control and a leading figure in respiratory medicine, who also serves as its chief expert. Professor Bai spearheaded the development of the “Chinese Consensus on the Diagnosis and Treatment of Pulmonary Nodules” and the “Asia-Pacific Guidelines for the Diagnosis and Treatment of Pulmonary Nodules.” Leveraging the Internet of Things (IoT), he developed PNapp, an intelligent AI-assisted diagnostic and therapeutic tool for pulmonary nodules that integrates domestic and international consensus guidelines. This innovation has established the distinctive PNapp 5A diagnostic and treatment pathway, enabling the annual diagnosis of over 200,000 early-stage lung cancer cases and delivering lifelong benefits to patients.

It is understood that the PNapp 5A diagnostic and treatment protocol comprises five steps: medical history inquiry, examination and assessment, recommendations, intervention planning, and intelligent assistance. During the recommendation phase, which is critical for differentiating between benign and malignant nodules, the WeDoctor Pulmonary Nodule Diagnosis and Treatment Center employs multiple methods, including artificial intelligence-assisted diagnosis and Circulating Abnormal Cells (CAC) testing.

Leveraging artificial intelligence and big data to help doctors analyze hundreds of thin-slice chestCTPrecise and rapid localization of pulmonary nodules in imaging, with risk probability assessment based on international clinical imaging standards, the Chinese consensus on the diagnosis and treatment of pulmonary nodules, and the Asia-Pacific guidelines for the diagnosis and management of pulmonary nodules., providing objective imaging data support for the diagnosis of pulmonary nodules, enablingEffectively Reduce Missed and Misdiagnoses, significantly enhancing the ability to differentiate between benign and malignant pulmonary nodules, andEstablishing a Comprehensive Personalized Management System for Patients with Pulmonary Nodules,achieving less thanBenign and Malignant Nature of 10 mm Pulmonary NodulesIdentification AccuracyGundam90% Professional Competency。

Previously, Ms. Yang consulted with specialists and made numerous trips to various hospitals. It was not until she visited the WeDoctor Pulmonary Nodule Diagnosis and Treatment Center that she experienced a different approach to medical care. As a digital “online + offline” center for whole-course disease management, all of Ms. Yang’s services were conducted online, with a health manager providing personalized one-on-one support throughout the entire process.

Ms. Yang explained that she located the “Pulmonary Nodule Diagnosis and Treatment Center” through the “Specialized Diseases and Departments” service on the WeDoctor app and purchased a package deal. Under the guidance of a health manager, Ms. Yang uploaded her CT imaging data online, and a nurse visited her home to collect peripheral blood samples for CAC testing. From purchasing the service to receiving the test report and expert recommendations, Ms. Yang completed the entire consultation process from home.

Based on a comprehensive analysis of the AI report and CAC test results, Ms. Yang was assessed to have a high risk of malignancy in her pulmonary nodules, and experts recommended surgical intervention. On December 8, Ms. Yang underwent minimally invasive surgery at the First Affiliated Hospital of Guangzhou Medical University. Thanks to early diagnosis and timely treatment, her prognosis is favorable, and no subsequent radiotherapy, chemotherapy, or other anti-tumor therapies are required.

Applying artificial intelligence to the early diagnosis of lung cancer, starting with pulmonary nodules—the precursors of lung cancer—to improve patient survival rates embodies the concept of “preventive treatment” advocated by Professor Bai Chunxue and the WeDoctor Pulmonary Nodule Diagnosis and Treatment Center.

A 73-year-old female patient was found to have nodules in the left upper lobe and right middle lobe of the lung eight years ago, and she adhered to annual follow-up visits as medically advised for eight years. Two months ago, during a consultation at Professor Bai Chunxue’s AI Clinic for Pulmonary Nodules, AI analysis indicated that the lesions were highly likely malignant and had progressed to an advanced stage. She was ultimately diagnosed with advanced-stage lung adenocarcinoma, a regrettable outcome.

Bai Chunxue analyzed this case and concluded that the primary reason for the adverse outcome was that the pace of medical technological advancement failed to keep up with clinical demand. AI-based analysis of the patient’s 2017 CT scans indicated a 77% malignancy risk for the nodule in the right middle lobe and a 90% malignancy risk for the nodule in the left upper lobe, both classified as high-risk. Had the patient undergone surgical resection or precise radiotherapy at that time, the prognosis would have been significantly improved. To prevent similar regrets, physicians need not only clinical experience but also new technologies such as AI-assisted analysis to promote standardized diagnosis and treatment. This approach embodies the principle of “preventing disease before it manifests,” thereby facilitating early detection, diagnosis, and treatment of lung cancer and benefiting a broader population.