Amyloid deposition in the heart is called cardiac amyloidosis (CA); 95% is immunoglobulin light chain amyloidosis (AL) and transthyretin amyloidosis (ATTR). Hereditary (ATTRm) or wild-type (ATTRwt) depends on whether the ATTRm gene is mutated or not. The most common mutation in Taiwan is A97S, 80% have left ventricular hypertrophy. The good prognosis depends on early diagnosis and correct treatment strategy. Bone-avid tracers such as 99mTc-PYP/DPD/HMDP could detect CA. The mechanism is not clear yet, which may be related to the microcalcification. AL amyloidosis is mostly between visual score grade 0-2, and ATTR-CM is usually grade≥2 on PYP scan, or heart to contralateral (H/CL) ratio, and it might replace invasive myocardial biopsy. However, there are no large-scale clinical studies, lack of standardization data, and limited information in comparison between clinical and imaging parameters. The project will enroll patients with suspected or diagnosed with CA according to CA diagnostic algorithm. Clinical data and image parameters are collected and compared. The project aims to set up prediction models based on the multi-parameters of PYP scan using artificial intelligence technology, including imaging registration and alignment technology, and standardization. The investigators further use the key cardiovascular data elements and imaging-derived database using model training network to extract image features to develop the diagnostic and prognostic prediction models, which are expected to validate the clinical significance and improve patient-centric performance and efficient clinical decision making.