Research Awards Recipients

General Audience Summary

Chronic kidney disease (CKD) prevents the body from performing vital functions such as blood filtering, blood pressure maintenance, and waste elimination. With diabetes, obesity, hypertension, and an aging population all increasing, CKD impacts over 1 in 10 adults globally. CKD follows a common histopathological pathway which results in the accumulation of fibrosis within the kidney. Due to deficiencies in how it is identified, fibrosis goes unnoticed until CKD has reached the latter, more severe, stages. Similarly, kidney transplantation has not seen significant improvements in long-term donor organ survival outcomes, driven by a similar sub-clinical accrument of fibrosis. The sole method for evaluating for fibrosis is invasive biopsy, which has limitations including needle phobia, patient pain, risk of hemorrhage, hematuria, and infection, and sampling error. Repeated protocol biopsies, while performed in certain centers, are not routinely performed in adults. This limits the ability to temporally monitor the response to treatment. This work proposes the use of ultrasound (US) and machine learning (ML) to non-invasively assess fibrotic load in a volumetric and real-time manner. US is commonly used to image the kidney but is challenging to use and interpret. ML can assist with this by allowing computer algorithms to automatically learn patterns and perform tasks without being explicitly programmed. It has shown record breaking capabilities in medical imaging. However, ML parameters and how they link to the disease state is difficult to interpret. An US imaging method called S-WAVE may be the missing link. Shear wave absolute vibroelastography (S-WAVE) is a new highly accurate variant of ultrasound elastography that provides volumetric tissue stiffness measurements and at larger imaging depths. S-WAVE has shown that for certain diseases, stiffness correlates with severity. Additionally, changes in glomeruli diameter in the early stages of damage cause fundamental changes to the raw ultrasound signal itself. ML can enable the easy measurement of glomeruli diameter. Using both SWAVE and glomeruli diameter together may yield a long-sought biomarker for renal fibrosis. This work proposes creating the new algorithms using US, ML, and S-WAVE to quantitatively measure fibrosis. In a single-center prospective cohort study on patients receiving all-cause kidney biopsy, patients will undergo a smart kidney ultrasound. Their SWAVE and glomeruli diameter results compared against biochemical and biopsy results. In turn, this may optimize resource usage, improve monitoring capabilities, and improve long-term outcomes.