Multimodality Integrated Imaging for Foetal Intervention

Fetal surgeries are among the most complex surgical interventions because entail operating in location with difficult access, through an endoscopic instrument that reduces vision and manoeuvrability, there is usually a lack of information on what to expect inside the patient, time sensitivity and extreme fragility of the patient. Recent advances in medical imaging techniques, artificial intelligence and mixed reality promise to have a profound impact in how those surgeries are performed. It is increasingly common to use multi-modal multi-resolution medical images to perform a surgical planning well ahead of the real operation in order to detect the possible problems and plan in advance. Those images are very complex to correctly acquire, process and integrate.

Our proposal is centred in Twin-to-twin Transfusion Syndrome (TTTS), a condition that occurs in monochorionic twins where both fetuses are connected to one single placenta. In around one third of cases, a blood flow imbalance will develop that result in the death of both the twins in more than 90% of cases. When TTTS is established, the only option is to separate blood circulation of both fetuses by coagulation the blood vessels connections with a laser. The complexity of operating through the womb, restrained by the anatomy of the mother, the high variability of placental position and vasculature distribution, the low manoeuvrability of the endoscope, the extremely reduced field of view and time sensitivity of the whole intervention are all key factors that make this intervention complex and risky. In this context, the surgeon is under a great cognitive stress, trying to integrate all the information in a mental picture that enables him to take the correct decision at every step. Our aim is to give him a “superpower” sight sense that overcomes most of the limitations of the current clinical setting and enables him to deliver better care.

We propose to: (1) integrate multimodal medical imaging techniques such as MRI, US and endoscopic video feed to present the surgeon with an integrated view of the operating field inside the womb, (2) build a 3D map of the target vessels from endoscopic view and their real-time localisation with respect to the anatomy of the mother and the current position of the surgical instruments (3) present this information in a mixed reality solution and (4) test that the proposed solution has the capability to reduce the cognitive load of the surgeon and speed up the intervention.

The MIIFI project aims to assist fetal surgeries to treat Twin-Twin Transfusion Syndrome (TTTS). To this end, Vicomtech is developing algorithms to build 3D maps of the placental vessels on fetoscopic videos recorded during surgery. This could reduce the surgeon's cognitive load and speed up the intervention.

Developed jointly by Pompeu Fabra University (UPF), Sant Joan de Deu Foundation and Vicomtech, this will be a great breakthrough that will change forever the way fetal surgeries are performed and will positively impact life conditions for these patients.