Fig 1.
Smartphone-based semen analysis system.
(A) The actual smartphone optical attachment along with a smartphone. (B) The exploded image shows the various components of the smartphone-based semen analysis system. (C) Images acquired with the smartphone imaging platform for the three different assays, (i) HBA, (ii) viability, and (iii) DNA fragmentation.
Fig 2.
Comparison of conventional method of analysis against automated smartphone-based HBA assessment.
(A) A linear regression analysis revealed strong agreement between the two methods (n = 31). The regression line is the solid line and the equation presented, is the line equation. (B) The samples were also compared using the Bland-Altman method of analysis (n = 31). The analysis revealed an absence of proportional and systematic bias for the tested sample set. The solid red line marks the mean bias. The dotted lines are the 95% limits of agreement of the sample set. (C) The scatter plot here represents the device performance in classifying samples as positive (<80%) and negative (= >80%). The system showed a sensitivity of 100% and a specificity of 69.23% (n = 31). The overall accuracy of the system in HBA-score based classification was 87.10% (D). The scatter plot shows the concentration and motility values of semen samples as measured by a CASA system along with its respective HBA score (n = 30).
Fig 3.
Comparison of conventional method of analysis against using smartphone-based optics to measure sperm viability.
(A) Passing-Bablok analysis revealed a strong linear relationship between the smartphone-based analysis and conventional method of sperm viability assessments (n = 102). The regression line is the solid blue line and the equation presented, is the line equation. The red lines represent the 95% confidence intervals of the regression line. The black line represents the line of identity, (B) The Bland-Altman method of analysis (n = 102) was performed on the sample set. The blue line represents the mean bias. The red lines are the 95% limits of agreement of the sample set.
Fig 4.
Comparison of conventional method of analysis against automated smartphone-based sperm DNA fragmentation assessment.
(A) The Bland-Altman method of analysis (n = 20) was performed on the sample set comparing the DNA fragmentation scores obtained through conventional analysis and the scores obtained manually through the use of a smartphone-based approach. (B) The Bland-Altman method of analysis (n = 27) was performed to compare the smartphone-based automated approach and the smartphone-based manual approach. The blue lines represent the mean bias while the red lines are the 95% limits of agreement of the sample set.