Akari

Opportunities for Surgeon-Scientists

Nakamura invites qualified healthcare professionals to explore collaborative opportunities regarding the development of Akari:

• Consulting, Advisory and KOL Roles
• Sponsored Research Agreements (SRA)
• Joint Research Agreements (JRA)

Such collaborations may potentially include joint publications, patent applications, and development of imaging study protocols for IRB review at collaborator sites.

Key Akari Capabilities

Akari includes several innovative features not present in other computer surgical planning modules:

• automated surgical planning of orthognathic surgery, facial bone contouring surgery, and placement of dental implants.
  
• automated generation of alloplastic facial implants and bone graft meshes suitable for the optimized surgical plan.
  
• proprietary planning and optimization algorithms that overcome the problems of deep-learning based planning approaches being developed by incumbents, which include accuracy, availability of data, and explainability.
• computational modeling and surgical feasibility verification systems for microsurgical reconstruction.
• a novel workflow for automated prosthodontic restorative planning in orthognathic surgery cases.
• robotic trajectory optimization algorithms for craniofacial surgery indications.

Generating optimal aesthetic movements for orthognathic surgery

Correcting malocclusion is just one piece of delivering a successful orthognathic case. Akari helps surgeons determine the optimal aesthetic movement for each patient by automatically generating a surgical plan, with or without concomittant alloplastic implant placement or facial bone contouring.

Akari automates highly manual processes in computer surgical and dental restorative planning. It reduces the cost and time burden on biomedical engineering and clinical teams in planning and promises to boost fee-for-service margins for biomedical engineering and device vendors providing computer surgical planning services. 

Surgeon-tunable automation and Explainable AI

Akari augments surgeon's expertise, not replacing it. Akari is the first computer surgical planning system that allows surgeons to create and change AI-generated plans on the fly. Surgeons are able to change desired 2D- or 3D-cephalometric values or measures within Akari, which then reoptimized the entire plan accordingly.

Akari was designed to overcome the limitations of deep learning based planning systems currently in development by incumbents. The main limitation of such systems is nondeterminism and variance in results. That is, due to stochastic variances inherent to many deep learning methods, different outputs are returned by the program despite identical inputs for different executions of the same program. Furthermore, another limitation of such systems is that they are not explainable; that is, first principles explanations of how such an algorithm came to decide that the returned plan was optimal cannot be generated. Finally, such systems are generally not tunable, that is they learn optimal patterns from a previously curated set of medical imaging scans and associated metadata, and if a surgeon or other clinical user wishes to modify the extent to which certain landmarks are optimized for, or disregarded somewhat as a “soft constraint”, as would be necessary to compare multiple potential plans and address unique and varying patient morphology, they generally cannot do so. Finally, such deep learning methods still require significant manual intervention, including the manual selection of osteotomy planes, which can cause the plan to use a suboptimal osteotomy path; selecting correct osteotomy planes can prove to be the one of the most technique-sensitive and potentially difficult aspects of a computer surgical planning workflow. 

Automated feasibility verification for double-barrel fibula free flap reconstruction

Double-barrel fibula free flap reconstruction cases are prone to a complication known as "planned failiure," where kinking of the peroneal artery impacts perfusion of the distal segement, leading to necrosis. This is because while such cases are planned using computer surgical planning tools, the angulation and spacing of ostetomies are determined by manual heuristics. Akari includes a feasibility verification module that can indicate whether a proposed fibula free flap reconstruction is feasible with respect to anatomical constraints, thereby allowing for a greater cohort of surgeons to carry out double-barrel fibula free flap reconstruction surgery with confidence.

A new frontier in facial bone contouring surgery

The roadmap of the Akari research program includes a craniofacial robotic surgery system which has been in active development since 2021. Such robotic surgery system is planned to help surgeon, increasing the predictability and reliability of completing complex facial bone counturing surgical cases in high-volume settings. Indications for the Akari surgical robot include:

• Zygomatic repositioning osteotomies (e.g. ZSO)
• Mandible Angle Osteotomies (MAO, colloquially "V-line surgery")
• LeFort 1 Osteotomies and Sagittal Split Ramus Osteotomies

Akari is a preclinical research prototype. For research use only. Not for sale. Not for clinical use. Not cleared or approved by FDA.