One-stop-shop Early Diagnosis of Joint Instability using AI-assisted 6D Computed Tomography

PROJECT SUMMARY/ABSTRACT Around 80 million US adults will have arthritis by 2040, which causes a significant societal problem resulting in an annual healthcare cost of > $136.8 billion. These progressive degenerative diseases result in long-term functional disabilities and morbidities leading to activity limitations and productivity loss. Joint instability typically precedes before the symptom onset, but it is typically not detectable in standard imaging exams. Patients can therefore suffer for years, as symptoms worsen before an accurate diagnosis is made. There is a critical need for an accurate imaging technique for early diagnosis, which is crucial to providing the most effective medications, therapies, and surgical interventions and restoring joint function before the onset of disabling symptoms. Various limitations have hindered the accurate detection of joint instability in standard imaging exams: radiography and fluoroscopy cannot visualize 3-dimensional (3D) joint anatomy; ultrasound lacks visualization of bone structure; static MRI/CT cannot assess the effect of dynamic structural and biomechanical pathologies. Accurate diagnosis is therefore oftentimes acquired using invasive arthroscopy, which exposes patients to risks of the associated complications. Although there have been some recent advancements in 4D (3D + time) CT to provide dynamic imaging capability, its clinical use is limited due to (1) joint motion artifacts that affect accurate anatomical and functional analyses, and (2) poor soft-tissue visualization that prohibits robust assessment of connective tissue involvement. The specific goal of this application is to develop and validate a new diagnostic imaging technique – 6DCT (3D + time + spectral information + kinematics) in combination with custom deep-learning methods. The 6DCT will provide an accurate, informative, one-stop-shop diagnostic tool to assess early dynamic joint pathologies before the onset of arthritis, which will obviate the need for invasive tests and procedures, with improved patient care and associated cost savings. We will accomplish this goal through three specific aims: 1. Develop a robust motion correction technique for dynamic joint imaging. 2. Develop a spectral post-processing technique for characterizing connective tissue involvement. 3. Demonstrate the feasibility of 6DCT in imaging dynamic joint pathology. This work is the first to develop innovative 6DCT as a robust dynamic imaging tool for both bone and connective tissues, and to provide new anatomical and functional information to facilitate assessing dynamic joint pathologies. These capabilities have clinical significance as they will enable early diagnosis of joint instability, which will subsequently enable early interventions that will improve patient health. This proposal will develop the essential techniques and preliminary data for future large cohort prospective patient studies.