Opportunities for the Medical AI Industry Post-Pandemic
-
2020 was supposed to be a year of 'reshuffling' for the medical AI industry, with companies facing fierce competition. The outbreak of the pandemic highlighted AI's ability to gradually replace certain mechanical tasks, assist doctors in rapid disease diagnosis, and improve treatment efficiency. What opportunities will the medical AI industry encounter after the pandemic?
2018 was dubbed the 'year of medical AI implementation' by the healthcare industry. However, the medical AI sector lacked a mature business model and was often criticized for 'burning money without profitability.' By 2019, the industry entered a capital winter, with fewer companies able to survive on funding alone. Then, 2020 began with the black swan event of the COVID-19 pandemic.
Major medical AI companies have stepped up their efforts, playing significant roles in pandemic prevention. Key contributions include:
1. AI Doctor Q&A
AI chatbots or online AI doctors can provide answers to pandemic-related questions and common health concerns for groups like the elderly, children, and pregnant women. This reduces hospital visits, lowers cross-infection risks, and supplements medical resources.2. Triage Robots
During the pandemic, patients couldn't confirm their infection status before seeking medical care. Robot-assisted triage reduced the risk of infection for healthcare workers conducting screenings and minimized their contact with other patients, significantly protecting medical staff.3. AI-Assisted Diagnosis
AI medical imaging serves as a clinical diagnostic tool. Technologies like AI+CT and AI algorithms + genetic analysis help doctors review scans, quickly identify suspected cases, and improve diagnostic efficiency, enabling faster isolation and treatment to curb virus spread.4. Robotic Treatment
Smart treatment robots assist doctors, reducing or eliminating direct contact with patients and lowering infection risks for healthcare workers. They also provide treatment recommendations for patients with pre-existing conditions, aiding in personalized care.5. AI Drug Development
AI algorithms and computing power accelerate tasks like viral gene sequencing, vaccine/drug development, and protein screening.During the pandemic, companies in these five areas demonstrated their capabilities. For example:
- Alibaba DAMO Academy's AI-based COVID-19 CT imaging diagnosis achieved 96% accuracy in under 20 seconds.
- Shanghai Children's Medical Center deployed robot 'Xiaobai' to reduce face-to-face interactions amid protective gear shortages.
- Ping An Good Doctor saw 1.1 billion platform visits during the pandemic, with daily new user visits nine times higher than usual.
The pandemic exposed the severe shortage of medical resources during emergencies. Technologies like the internet, big data, and AI proved effective in supplementing resources and enabling remote consultations, earning recognition for internet-based healthcare. Future prospects look promising, with long-term investment potential.
In 2019, over 140 medical AI companies focused on medical imaging, with about 100 specializing in lung nodule analysis. The pandemic showcased the strengths of companies like Yitu Healthcare's intelligent imaging system and Infervision's COVID-19 AI system, though some firms had to pivot to new areas.
The pandemic also revealed new potential applications for medical AI, such as:
1. Decision Support for Healthcare Workers: Providing diagnostic and treatment recommendations to improve accuracy and speed.
2. Medical Imaging Analysis for Other Diseases: Alibaba DAMO Academy developed an AI algorithm trained on 5,000+ CT scans, showcasing the potential of machine learning in analyzing MRI, CT, and other imaging data.
3. Medical Robots in Treatment: Robots for diagnostics, Q&A, disinfection, and delivery can reduce cross-infection risks and protect medical staff, with future roles in inspections, surgeries, and more.
4. Health Wearables: Smart devices for continuous health monitoring are crucial for high-risk patients, diabetics, chronic disease sufferers, and pregnant women unable to attend regular check-ups.
5. AI in Drug Development: AI can shorten the traditional 10-15 year drug development cycle and reduce costs by screening potential treatments and accelerating research.
A major hurdle for medical AI is the lack of high-quality, labeled data for training algorithms, leading to limited disease coverage and product homogeneity. The pandemic prompted some hospitals to share data, easing this challenge temporarily, though nationwide data sharing remains difficult due to incomplete digitization and reluctance post-pandemic.
Advances in blockchain technology and national policies may enable authorized data sharing in the future, breaking down long-standing data silos in healthcare.
AI-assisted diagnostics have excelled during the pandemic. For example, Alibaba DAMO Academy's AI+CT system analyzes 300+ CT slices in under 20 seconds with 96% accuracy, compared to the 10-15 minutes required by human experts.
This poses a challenge to traditional clinicians. In the long run, AI replacing some mechanical and repetitive tasks is an inevitable trend. This requires future medical practitioners to further enhance their professional capabilities, not just as pure clinicians but also with some IT knowledge. For AI researchers, an accuracy rate of 96% or even 99% does not mean the product is flawless. Doctors are still needed to address the potential misdiagnoses in the remaining 4% or 1%, as well as any errors within the 96% or 99%.
Healthcare + AI requires medical AI teams to translate medical problems into engineering language, while healthcare professionals need to understand AI-related knowledge. Only then can the two communicate effectively and collaborate to develop truly high-availability AI products. This inevitably demands higher professional competence from practitioners.
During the SARS outbreak in 2003, we stood united to overcome the crisis. Over a decade later, China's technology has advanced rapidly, and the battle against the novel coronavirus is a fight of the technological era.
During SARS, I was still a student, and the world protected us. Now, I want to say: healthcare workers protect the world, and it’s our turn as tech workers to protect you. I believe that after this pandemic, the medical AI industry will see broader and deeper research, leading to more high-availability intelligent products that contribute to human health.
Technology makes our fight against the pandemic stronger, and I hope it also fosters a more harmonious coexistence among all species: humans, nature, and intelligent machines.