Nir Pillar, MD, PhD
Using AI to Improve Diagnosis of Systemic Amyloidosis
Systemic amyloidosis is a group of rare disorders characterized by the abnormal buildup of misfolded proteins in various tissues, leading to progressive organ dysfunction and death. The symptoms can vary depending on the organs affected, and may include fatigue, weight loss, and swelling.
Diagnosing amyloidosis is challenging due to the non-specific nature of the symptoms and because it requires the microscopic identification of amyloid deposits in tissue samples. The current diagnostic methods have limitations in terms of sensitivity and specificity and are technically difficult to interpret. Misdiagnosis can lead to the delay of life-saving treatments.
For more accurate detection of systemic amyloidosis, my research will use an innovative approach to conduct virtual histological staining, a process that highlights the microscopic cellular appearance of tissue sections without using any chemical stains or additional laboratory steps. We will combine novel microscopy methodologies and deep learning, a type of artificial intelligence that can learn and improve like the human brain, to generate a virtual stain that highlights amyloidosis in a more precise manner. The proposed methodology has the potential to advance systemic amyloidosis detection and, consequently, patient treatment.
The PhRMA Foundation's Postdoctoral Fellowship in Translational Medicine will enable me to develop a deep learning-based, virtual tissue staining method for faster and more accurate detection and classification of amyloidosis in unstained tissue sections.