ABSTRACT
Objective:To explore artificial intelligence technology and propose an algorithm for automatic identification of dicentric chromosomes to realize fast and high-throughput biodosimetry. In order to solve the time-consuming and laborious problem of manual analysis of dicentric chromosomes.Methods:Combining artificial intelligence technology and image processing technology, based on MATLAB software, algorithms like image preprocessing, threshold segmentation algorithm, binarization processing, area identification algorithm, convolutional neural network algorithm and double centripetal recognition algorithm were applied. A fuzzy membership function was defined to describe the degree of each chromosome belonging to a dicentric chromosome, and the discrimination threshold was set to realize the automatic identification of dicentric chromosomes.Results:Through the test on 1 471 chromosome images, compared with manual recognition, the detection rate of dicentric chromosomes cells of this algorithm reached 70.7%.Conclusions:This algorithm method carries out a preliminary study on the automatic identification of dicentric chromosomes with good result.
ABSTRACT
In this paper, the research has been conducted by the Microsoft kinect for windows v2 for obtaining the walking trajectory data from hemiplegic patients, based on which we achieved automatic identification of the hemiplegic gait and sorted the significance of identified features. First of all, the experimental group and two control groups were set up in the study. The three groups of subjects respectively completed the prescribed standard movements according to the requirements. The walking track data of the subjects were obtained straightaway by Kinect, from which the gait identification features were extracted: the moving range of pace, stride and center of mass (up and down/left and right). Then, the bayesian classification algorithm was utilized to classify the sample set of these features so as to automatically recognize the hemiplegia gait. Finally, the random forest algorithm was used to identify the significance of each feature, providing references for the diagnose of disease by ranking the importance of each feature. This thesis states that the accuracy of classification approach based on bayesian algorithm reaches 96%; the sequence of significance based on the random forest algorithm is step speed, stride, left-right moving distance of the center of mass, and up-down moving distance of the center of mass. The combination of step speed and stride, and the combination of step speed and center of mass moving distance are important reference for analyzing and diagnosing of the hemiplegia gait. The results may provide creative mind and new references for the intelligent diagnosis of hemiplegia gait.
Subject(s)
Humans , Algorithms , Bayes Theorem , Gait , Gait Analysis , Methods , Gait Disorders, Neurologic , Diagnosis , Hemiplegia , WalkingABSTRACT
Objective To evaluate the value of a dynamic automatic identification system in routine miracidium hatching test with nylon gauzes. Methods Different quantities of fresh Schistosoma japonicum eggs were added to bovine fecal samples and divided into the low-infection group, medium-infection group and high-infection group, while the bovine feces without S. japonicum eggs served as negative controls. The detection efficiency and accuracy were compared between the identification system and manual detection in different groups. Results The identification system can automatically identify S. japonicum miracidium. The detection rate and efficiency of S. japonicum miracidium in bovine fecal samples were both higher by using the identification system than by manual detection. Notably in the low-infection group, the identification system had a significantly higher rate of detection of S. japonicum miracidium than manual detection (χ2 = 10.769, P = 0.002). The identification system completed the detection of bovine fecal samples in the field within 1 min. Conclusions The dynamic automatic identification system may effectively improve the detection efficiency and accuracy of routine miracidium hatching test with nylon gauzes, and it may replace manual detection to be used in the field schisotsomiasis examinations and related researches.
ABSTRACT
Objective To evaluate the value of a dynamic automatic identification system in routine miracidium hatching test with nylon gauzes. Methods Different quantities of fresh Schistosoma japonicum eggs were added to bovine fecal samples and divided into the low-infection group, medium-infection group and high-infection group, while the bovine feces without S. japonicum eggs served as negative controls. The detection efficiency and accuracy were compared between the identification system and manual detection in different groups. Results The identification system can automatically identify S. japonicum miracidium. The detection rate and efficiency of S. japonicum miracidium in bovine fecal samples were both higher by using the identification system than by manual detection. Notably in the low-infection group, the identification system had a significantly higher rate of detection of S. japonicum miracidium than manual detection (χ2 = 10.769, P = 0.002). The identification system completed the detection of bovine fecal samples in the field within 1 min. Conclusions The dynamic automatic identification system may effectively improve the detection efficiency and accuracy of routine miracidium hatching test with nylon gauzes, and it may replace manual detection to be used in the field schisotsomiasis examinations and related researches.
ABSTRACT
Objective To develop a dynamic automatic identification system(device)of Schistosoma japonicum miracidia to achieve the automatic detection and improve the detection rate and efficiency of schistosome miracidia.Methods The dynamic automatic identification system(device)of S.japonicum miracidia was composed of an optical stereoscope,a digital camera,dy-namic automatic tracking recognition software,and a computer.Under different conditions,the detection rates and efficiency of the system were compared with those of five professional persons.Results The basic function of the system was to identify,la-bel and warn the miracidia of S.japonicum,and the records could be saved automatically.The laboratory tests showed that the missing rate of the system was 0.The total consistent rates of the manual detection were 74.67% and 66.67% in the condition with and without water bug,while the total consistent rates of the system were 100.00% and 96.67%,respectively(χ2=9.634, 11.081,both P<0.01).Conclusions The system is much superior to manual detection in the accuracy and speed,and the sys-tem could completely replace the manual detection.Therefore,the system could be used in the field and basic research of schis-tosomiasis.
ABSTRACT
To evaluate the effect of an automatic identification system of Schistosoma japonicum miracidia, and compare it with the traditional eye detection method in the simulation field.A total of 260 fecal samples were collected from schistosomiasis non-endemic areas, and the test sample bottles containing schistosome miracidia were prepared according to different experimental needs. Thirty fecal samples for the sensitivity test were separately added with five fresh miracidia per sample, and then the mixed samples were detected by two experienced technicians (with more than 15 years’ traditional test experience) or the automatic system. The positive detection rates were compared between the two methods. Thirty fecal samples for repetition test were separately added with ten fresh miracidia per sample, and then the mixed samples were detected separately with the automatic identification system by two experienced technicians. The results were compared between two persons. The two methods including the automatic identification system and the traditional eye detection method were carried out blindly with totally 200 samples in the simulation field. There were three groups (each with 30 samples) : Group 1 with more than 21 fresh miracidia, Group 2 with 6 to 20 fresh miracidia, and Group 3 with 1 to 5 fresh miracidia. The other 110 samples were as a negative group. The detection time, accuracy, missed detection rate, and false detection rate of the two methods were statistically compared.The positive detection rates of the 30 positive samples were 43.33% and 33.33% by the two technicians with the traditional eye detection method, respectively, while the detection rate was 80.00% by the automatic identification system, and the difference was statistically significant (χ2 = 7.05, χ2 = 12.97, both P < 0.01). Thirty positive samples were detected by the two technicians using the same automatic identification system, and the positive detection rates of the two were 96.67% and 86.67%, respectively, with no significant difference (χ2 = 0.27, P > 0.05). The experiments showed that the correct detection rate of the positive samples was 98.00% by the automatic identification system, which was higher than 79.75% by the traditional eye detection method. The detection time of the automatic identification system was shortened by half compared with that of the traditional eye detection method. The missed detection rate, and false detection rate of the automatic identification system were 2.22% and 1.82%, respectively, which were much lower than 35.56% and 7.73% of the traditional eye detection method.Compared with the traditional eye detection method, the automatic identification system of S. japonicum miracidia has the advantages of high sensitivity, good repeatability, short detection time, high accuracy, low missed detection rate, and low false detection rate. It can be used in the field and clinical detection in replacement of the traditional eye detection method.