2021 Fiscal Year Research-status Report
Development of artificial intelligence assisted Raman microscopy for reliable and automated examination of urine sediment
| Project/Area Number |
21K18081
|
|---|
| Research Institution | Shimane University |
|---|
Principal Investigator |
H Noothalapati 島根大学, 学術研究院農生命科学系, 助教 (30748025)
|
|---|
| Project Period (FY) |
2021-04-01 – 2023-03-31
|
| Keywords | Raman spectroscopy / Urinalysis / Urine sediment analysis / Artificial intelligence / disease diagnosis / machine learning / MCR-ALS / multivariate analysis |
|---|
| Outline of Annual Research Achievements |
Urine sediment microscopy is vital in accurate diagnosis of kidney and urinary tract diseases. However, it is not universally adopted because it expensive, time consuming and required trained personnel. So, we plan to develop artificial intelligence (AI) assisted Raman microscopy for reliable and automated examination of urine sediment. We prepared artificial urine and measured Raman spectra of various components in urine successfully. We then employed multivariate analyses such as principal component analysis (PCA) to differentiate multiple components but it does not provide detailed molecular information. Therefore we further employed multivariate curve resolution analysis (MCR) to extract pure molecular spectra which helped to identify markers for objectively diagnosis.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
Application of multivariate curve resolution analysis to differentiate various cell types as proposed in the application has been successfully tested and the results have already been published as an original peer-reviewed article. Further to expand the ongoing research, we extensively reviewed literature on diagnostic potential of Raman spectroscopy for various diseases especially by liquid biopsy approaches and its advantages of combining microfluidics device for accurate and automated measurements. Our works have been published as two peer-reviewed critical review articles.
|
| Strategy for Future Research Activity |
We plan to measure many cell types in urine sediment this year and accumulate Raman hyperspectral big data. Markers obtained from MCR analysis will be used to build artificial intelligence models for reliable and automated examination of urine. With such encouraging results, we would like to acquire and test the proposed hypothesis in real urine samples. To this end, we plan to initiate an international collaboration with researchers in Advanced Center for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Center, Mumbai, India to measure Raman spectra of clinical urine samples.
|
| Causes of Carryover |
Delivery of the planned purchase has been delayed due to disruption in supply chains during COVID-19 pandemic. Necessary amount has been carried and will be used for the purchase of required microscope stage parts and optics in the next year.
|
