Safer Global Surgery with SDSC

Surgical Data Science Collective (SDSC)

Safer Global Surgery with SDSC

Supercharging neurosurgical training in low- and middle-income countries via a novel AI-powered video training platform: a pilot in Tanzania

Washington, DC, USA

USA

Nonprofit

We aim to address the severe shortage of trained surgeons in low- and middle-income countries (“LMIC”), starting with a pilot effort in Neurosurgery in Tanzania. Neurological diseases including stroke, trauma, congenital disorders, cancer, and neurodegenerative conditions are the leading cause of disability adjusted life years among all global disease categories (Steinmetz et al., The Lancet, 2021). In addition, more than 5 million people need neurosurgical care each year but cannot access it (Dewan et al., Journal of Neurosurgery, 2018). As expected, unmet neurosurgical needs and gaps in access to care are greatest in LMIC. Globally, there is one neurosurgeon for every 107,000 people. Yet, in Tanzania, where our work will take place, there are just 27 neurosurgeons serving 67 million individuals, or one per 2,500,000. Surgical capacity in LMICs is limited by funding, workforce, technology, and overwhelming clinical volumes. However, new technologies can provide scalable, efficient, and effective neurosurgical training in LMICs, helping to expand local surgical capacity and improve patient outcomes by closing the enormous gaps in access to specialty care.

SDSC is a non-profit founded by pediatric neurosurgeon Dr. Daniel Donoho. SDSC applies novel AI algorithms to intraoperative surgical video and improve the efficiency, quality, and accessibility of global surgical training. Each year, millions of hours of surgical video go unwatched and unutilized. The SDSC leverages data to disseminate the best surgical techniques worldwide and address the global neurosurgical workforce shortage. SDSC has created a Surgical Video Platform (“SVP”) which uses computer vision algorithms to analyze surgical video and give automated feedback on performance and technique. Currently, hundreds of surgeons upload and store their surgical videos, receive technical feedback, and engage with a global community of physicians on SVP. We have designed a pilot collaboration with the Muhimbili Orthopaedic Institute (“MOI”), the largest neurosurgical center in Tanzania. MOI neurosurgical trainees will record their procedures and upload the footage to SVP. SVP’s automated feedback will be combined with remote mentorship, where expert neurosurgeons provide targeted coaching sessions utilizing SVP metrics. Improving the effectiveness and efficiency of neurosurgical training through technology will yield greater surgical capacity and access to care for patients globally.

The SVP aims to serve the millions of patients in LMICs who do not have access to quality surgical care. To help patients in Tanzania, we will use the SVP to enhance the training of neurosurgical residents (also known as “Registrars”) at MOI. These future neurosurgeons may lack the expert mentorship and advanced training tools that have drastically improved outcomes in higher resource settings. This training gap leads to uneven skill development, exacerbates surgical complication rates, and constrains the growth of health systems. By equipping the MOI trainees with an AI-driven platform for in-depth analysis and feedback on their neurosurgical techniques, our solution provides a structured, scalable framework for improving surgical skill and ultimately patient outcomes. To our knowledge, no other training platforms currently exist which apply data science and artificial intelligence to surgical video as a global health tool. This collaboration with MOI in Tanzania is the first step in our mission to transform global surgical education by leveraging a highly available data source - surgical video - to change the course of care for millions.