An automatic bacteria classification system
Grant-in-Aid for Developmental Scientific Research.
|Allocation Type||Single-year Grants|
|Research Institution||Dept. of Information Science, Faculty of Engineering, Fukui University|
TANIGUCHI Keiji Fukui University, Faculty of Engineering, Professor, 工学部, 教授 (50029063)
YOKOCHI Takashi Aitchi Medical University, Faculty of Medicine, Professor, 医学部, 教授 (20126915)
|Project Period (FY)
1989 – 1991
Completed(Fiscal Year 1991)
|Budget Amount *help
¥11,500,000 (Direct Cost : ¥11,500,000)
Fiscal Year 1991 : ¥1,700,000 (Direct Cost : ¥1,700,000)
Fiscal Year 1990 : ¥1,800,000 (Direct Cost : ¥1,800,000)
Fiscal Year 1989 : ¥8,000,000 (Direct Cost : ¥8,000,000)
|Keywords||Automatic / Flow-image-analysis / Identification for bacteria / An application for N_2 Laser / Sheath-flow thechiques / Diagnostic procedures / 細菌 / フロ-方式 / 画像処理|
We have developed a prototype acquisition system for automatically imaging bacteria. The hardware of this one consists of a light source (N2 Laser), image acquisition, image processing and control units.
The light beam emitted from the laser controlled by a micro-computer is transmitted to bacteria flowing in the orifice of the chamber.
The light beam modulated by bacteria is magnified by an object lens (X40) and detected by a CCD- camera.
We have also developed softwares for noise reduction, enhancement, binalization, feature extraction and classification of images of bacteria.
Next, we studied to classify the bacteria using the system mentioned above.
First of all, it was tried to discriminate between rods and cocci.
Our system enabled to discriminate them at a high rate.
Further, it was, also possible to distinguish streptococci and staphylococci from the cocci.
Second, we tried to calculate the length of the bacteria.
The length which we calculated was almost consistent with their actual length.
Third, we developed the programs to calculate the number of the bacteria in the given volume. It was found that the contamination of red blood cells and platelets in clinical specimens was negligible for identification of the bacteria. Various kinds of clinical specimens containing much proteins, such as serum, and-pus could be applied to our system.
Furthermore, we succeeded to sterilize clinical specimens by adding formalin up to the final concentration of 10%. This system could be sterilized using 70% ethanol.
Therefore, it was confirmed that this system was well suitable for classification and identification of the bacteria using clinical specimens.
Research Output (16results)