Robert CHANG – Promises & Pitfalls of AI and Digital Health
From telemedicine and artificial intelligence, miniature sensors and consumer wearables, big data and precision health–there is a lot of excitement surrounding digital health innovation. This talk introduces exciting new technologies and how they may shape the future of medicine.
Divya NAG – Mobile Technology’s Role in Advancing Ophthalmology
Apple’s journey into the world of healthcare has uncovered unprecedented opportunities for mobile technology to play a significant role in medical research and care for consumers all around the world.
Theodore LENG – Artificial Intelligence’s Potential to Fight Blindness
There are 455 million people with diabetes worldwide and 4.2 million are visually impaired from diabetic retinopathy (DR). Diabetics need an eye exam every year to screen for DR, but there are not enough human resources to make this happen. 98% of vision loss from DR is preventable with early detection and treatment. Dr. Leng will speak about the power of automated artificial intelligence (AI) algorithms and their role in aiding in this screening process. Using an AI algorithm trained on over 75 thousand images, internally validated 5 times and externally validated 2 times, Dr. Leng was able to obtain a level of DR detection on the level of a human ophthalmologist. Implementing such an algorithm on a global scale could drastically reduce the rate of vision loss due to DR.
Andre ESTEVA – Artificial Intelligence in Detecting & Tracking Skin Conditions
Skin cancer, the most common human malignancy, is primarily diagnosed visually, beginning with an initial clinical screening and followed potentially by dermoscopic analysis, a biopsy and histopathological examination. Automated classification of skin lesions using images is a challenging task owing to the fine-grained variability in the appearance of skin lesions.
Deep convolutional neural networks (CNNs) show potential for general and highly variable tasks across many fine-grained object categories. Here we demonstrate classification of skin lesions using a single CNN, trained end-to-end from images directly, using only pixels and disease labels as inputs. We train a CNN using a dataset of 129,450 clinical images—two orders of magnitude larger than previous datasets — consisting of 2,032 different diseases. We test its performance against 21 board-certified dermatologists on biopsy-proven clinical images with two critical binary classification use cases: malignant carcinomas versus benign seborrheic keratoses; and malignant melanomas versus benign nevi. The first case represents the identification of the most common cancers, the second represents the identification of the deadliest skin cancer.
The CNN achieves performance on par with all tested experts across both tasks, demonstrating an artificial intelligence capable of classifying skin cancer with a level of competence comparable to dermatologists. Outfitted with deep neural networks, mobile devices can potentially extend the reach of dermatologists outside of the clinic. It is projected that 6.3 billion smartphone subscriptions will exist by the year 2021 and can therefore potentially provide low-cost universal access to vital diagnostic care.
Suman THAPA – Tele-Ophthalmology Potential in Nepal
Tele-ophthalmology in Nepal has a definite role for screening eye diseases. The difficult geographic terrain and limited human resource challenges the establishment of a sustainable eye care infrastructure in certain parts of the country. This poses a serious barrier for patients residing in remote areas of the country. Opportunistic screening, use of affordable portable equipment and the utilization of available human resource is the way to start addressing the problem. This presentation will cover tele-ophthalmology research work that has been conducted at Tilganga Institute of Ophthalmology in Kathmandu for screening eye diseases with the hope of providing sustainable eye care.
Tien-Yin WONG – Deep Learning Technology & Its Applications in Diabetic Screening
Diabetic retinopathy (DR) is a major complication of diabetes, and the leading cause of blindness among working adult people worldwide. About 600 million will have diabetes by 2040, with a third having DR and 10% with severe vision-threatening DR (VTDR). DR screening along with timely referral and treatment, is a universally accepted strategy for the prevention of visual impairment. Currently, DR screening by fundus photography, usually within a tele-ophthalmology framework, with assessment of the fundus photographs by human assessors (e.g., ophthalmologists, general physicians, technicians) is the most commonly used method for DR screening. However, this type of DR screening program is limited by availability and training of human assessors, and long-term financial sustainability. The need for low cost, sustainable DR screening programs is substantial
Deep learning technology is a relatively new branch of artificial intelligence (AI) that has substantial potential for DR screening. Previous technology for automated DR screening using traditional “pattern recognition” techniques to detect specific DR lesions (e.g., microaneurysms) have been promising but has not broken the “translational gap” from research to clinical adoption. Deep learning uses much larger datasets and uses a “black box” approach to mine, extract and learn patterns and/or features to determine a disease state or condition. Recently, researchers from Google using deep learning technology have reported high sensitivity and specificity (>90%) in detecting referable DR from retinal photographs. However, for translational impact, deep learning technology should be trained and validated in “real-world” screening programs where fundus images have varying qualities (e.g. cataract, poor pupil dilation, poor contrast/focus), and with patient samples of different ethnicity (i.e. different fundi pigmentation) and systemic control (poor and good control). Furthermore, in any screening programs for DR, the detection of incidental but common vision-threatening conditions such as glaucoma and age-related macular degeneration should be incorporated, as missing such cases may not be acceptable to clinicians. Only then will deep learning technology be applicable in large scale screening programs for DR
Nam Han CHO – Diabetes & Telemedicine: IDF Challenges on Diabetic Retinopathy
2015 IDF Diabetes Atlas, 7th edition uses age-stratified data and a consistent methodology to estimate the diabetes prevalence in adults aged 20-79 years, across 170 countries and territories. Diabetes is one of the largest global health emergencies of the 21st century. Each year more and more people live with this condition, which can result in life-changing complications. Approximately 415 million people worldwide, or 8.8% of adults aged 20-79, are estimated to have diabetes, and there are 318 million adults with impaired glucose tolerance which put these adults at high risk of developing diabetes in the future. About 75% live in low- and middle-income countries. If these trends continue, by 2040 some 642 million people, or one adult in ten, will have diabetes. The largest increases will take place in regions where economies are moving from low- to middle-income levels. Currently there are more people with diabetes in urban (269.7 million) than in rural (145.1 million) areas. In low- and middle-income countries, the number of people with diabetes in urban areas is 186.2 million while 126.7 million live in rural areas.
Approximately 5.0 million people globally, between 20 and 79 years of age, died from diabetes in 2015, equivalent to one death every six seconds. Diabetes accounted for 14.5% of global all-cause mortality among people in this age group. Close to half (46.6%) of deaths due to diabetes are in people under the age of 60. The highest number of deaths due to diabetes occurred in countries with the largest numbers of people with diabetes: China, India, USA, and the Russian Federation. Diabetes related morbidities such as macrovascular and microvascular diseases are yet to be tackles effectively. On of area the IDF put more emphasis is in diabetes retinopathy. Diabetes retinopathy effects over on third of all people with diabetes and is the leading cause of vision loss in working-age adult. The management of diabetes and its complications begins in primary health care and this should include screening for diabetic retinopathy. Moreover, more conservative estimates suggest health spending on diabetes accounted for 11.6% of the total health expenditures worldwide in 2015. Over 80% of the countries covered in this report dedicated between 5% and 20% of their total health expenditure to diabetes. Furthermore, global health spending to treat diabetes and prevent complications was estimated to be 673 in 2015, and projected to exceed 802 billion US$ by 2040. Thus, placement of economical, efficient, and effective diabetes care management system, such as telemedicine are critical to prevent diabetes related morbidity and mortality in the future.
