How Do Artificial Intelligent Systems Improve Efficiency in Diagnosing Lung Cancer and Breast Cancer through Automated Medical Image Analysis?

Professor. HENG Pheng Ann, Professor, Department of Computer Science and Engineering, The Chinese University of Hong KongAccording to the statistics revealed by the World Health Organization (WHO), cancer is the second leading cause of deaths in the world and took away the lives of 8.8 million people in 2015, i.e. nearly one-sixth of deaths were caused by cancer. Among the various types of cancers, the most common one is lung cancer causing the death of 1.69 million people while breast cancer ranked fifth with 571,000 deaths. If we study the cancer statistics in Asia, we find 51 percent of the world’s lung cancer cases come from Asia and 21 percent of deaths from it come from the region. Breast cancer, with nearly 1.7 million new cases diagnosed in 2012, is the most common cancer found in women in the world. It represents about 12 percent of all new cancer cases and 25 percent of all cancers in women. It is the fifth most common cause of death from cancer in women3.

Is Application of Artificial Intelligence in Medical Sciences a Way Out?
With the big number of new lung and breast cancer cases registered as well as the public’s awareness of their health condition, it is not a surprise to find a rapidly growing demand for medical services in the global medical sector. As it takes years of training to nurture a medical professional, I always wonder how we, as scientists, can contribute to the healthcare sector with our professional knowledge. If we can apply the latest technologies in the healthcare sector in a proper way, we can help in improving the quality and effectiveness of medical services and benefit the community. Artificial intelligence (AI) has been applied in recent years in various sectors from analyzing big data to automation with robotics, which greatly improves our quality of life. Is there any way we can apply AI in medical treatment to improve efficiency in diagnosis?

Detection of pulmonary nodules through Deep Learning
At an early stage, lung cancer mostly exists in the form of small pulmonary nodules, which appear on medical images as shades of small lumps. Currently, doctors depend
on chest CT scans to reveal those nodules. However, each scan often results in hundreds of images. Assuming that going through each image requires three seconds, an analysis of these images by the naked eye will take five minutes to complete. Such examinations are time consuming, and must rely on the doctors’ experience and sharpness of focus. Recently a research team from a Hong Kong tertiary institution applied deep learning technology to CT scans, which were able to locate the pulmonary nodules in 30 seconds, with an accuracy of 90 percent. With positive feedback from the medical sector, the team expects the new technology will be widely adopted by medical practitioners in the next couple of years. Deep learning makes use of advanced training to improve the sensitivity of the technology, so that it is able to tackle a major challenge that a naked-eye examination faces; that is, to remove noise and reduce false positives. In order to further improve the technology, the team will be working with top hospitals in Beijing, to provide solid evidence in support of early diagnosis and treatment of lung cancer.

A key advantage of artificial intelligent deep learning is that it is able to analyse large quantities of parameters

Automated Detection of Metastatic Breast Cancer in Histology Images
To determine whether a patient has breast cancer, doctors often must extract and examine tissue samples. Using mammograms or MR scans to locate the lump; samples are extracted and examined under the microscope to see if there are signs of tumor and whether the tumor is benign or malignant. A digital histology is of high resolution, often up to one gigabyte in file size - equivalent to a 90-minute high resolution movie. Examining such an image requires a lot of time and energy. To solve the problem, a scientific research team at a Hong Kong tertiary institution has developed a novel deep cascaded convolutional neural network to process the histo pathological images. Making use of a fully convolutional network, the model can efficiently and accurately detect the metastatic cancer with a high-resolution score-map. The whole automated analysis process takes about five - ten minutes, as compared to the 15 - 30 minutes that are required if examined by the naked eye. In terms of accuracy, the system has achieved a rate of 98.75 percent, two percent higher than analysis conducted by experienced doctors. This indicates that it is an invaluable reference for clinical diagnosis on breast cancer.

A key advantage of artificial intelligent deep learning is that it is able to analyse large quantities of parameters. The more the data, the higher is its accuracy. When this automated screening and analysis system is applied to the medical sector, it acts as a tireless assistant to the doctor that can quickly identify the source of an illness, enabling a timely and appropriate treatment.

Benefits of Applying Technology in the Medical Sector
The above two cases are only some of the examples of how applications of AI contribute to the advancement of medical services. With more frequent collaborations between the IT experts and the medical professionals, I expect it to definitely help improve both efficiency and accuracy in diagnosing the diseases as well as the treatment provided to patients in the healthcare system. Moreover, the technical achievements will not only benefit citizens in the APAC region but will eventually reach out to the world and benefit mankind globally.