Materials used in this study are shown in Fig. 5. A CT scanner of 96 rows of detectors was used in the conventional scan mode with a tube voltage of 120 kV, pitch factor of 1.2, and effective tube-current time products of 150 mAs were applied. A cylindrical water phantom being 20 cm in a diameter and the chest area of a human-body phantom were used in this study. Dosimetry was performed using OSL dosimeters based on have a construction of Al₂O₃ [6,7,8]. This dosimeter is 1 cm square in size and absorbs a small amount of X-rays, therefore the dosimeter does not affect medical images even when placed within the Field of View (FOV).

Fig 5: Experimental equipment used in this study. A CT scanner having 96 rows of...

<<Novel Analysis Method>>

    A typical example (concept) of analysis flow is presented in Fig. 6. Detailed explanation will be given later (Fig. 7 & Fig. 8). Dose measurement was performed by combining (A) a measured value determined by the OSL dosimeter and (B) the X-ray dose distribution based on X-ray incident direction analysis which was carried out using the CT image.

Fig 6: A typical example (concept) of analysis flow. Dose measurement was performed by...

Fig. 7 shows three principles of the proposed method.

Fig 7: Concept of our analysis procedure. We assumed the surface dose of the human...

    The overall picture of our analysis scheme is shown in Fig. 8, and the mathematical functions used in this figure are summarized in Fig. 9. To analyze the experimental data, the proposed algorithm uses basic experimental data from "helical" and "conventional" scans.

Fig 8: The overall picture of our analysis scheme. To analyze the experimental data,...

Fig 9: Symbol, description and illustration of functions which are used in our...

<<Functions Used in Our Algorithm>>

    Figures 10, 11 and 12 show the mathematical functions used to construct the algorithm. All these functions were obtained experimentally. Analysis was performed using in-house programming software. Fig. 10 shows the dose distribution and SD distribution obtained from helical scanning. OSL dosimeters were placed evenly around the cylindrical water phantom to determine the incident direction of X-rays, and the relationship between the dose distribution and the SD distribution of the corresponding CT image was investigated. As a result, we found that the SD of the area in the X-ray incident direction was small, and the SD in the opposite area was large. The functions shown in Fig. 10 present the dose and SD distributions when X-rays are incident from directly above (Y axis). By rotating these functions, we can estimate the distribution functions of dose and SD for any angle.

Fig 10: Mathematical functions used in the algorithm (part 1/3). Detailed explanations...

    The mathematical functions shown in Fig. 11 were obtained from experiments performed using conventional scans. The graphs on the left and right are the results when using a water phantom and a human body phantom with a scanning bed, respectively. Based on these results, we created correction functions that can be applied to modify the dose and SD distributions defined by the water phantom (Fig. 11) and to those of a human body.

Fig 11: Mathematical functions used in the algorithm (part 2/3). Detailed explanations...

    Fig. 12 shows mathematical functions that can correct the effects of phantom difference and scanning bed on the distributions of dose and SD.

Fig 12: Mathematical functions used in the algorithm (part 3/3). Detailed explanations...

    Fig. 13 shows details of the algorithm for determining the incident angle of X-rays using the measured SD values. We assumed SD distributions by varying incident angles of X-rays from -180 degrees to 180 degrees. In order to estimate the incident direction of X-rays from the experimentally obtained SD values, by comparing the experimental values and simulated values (function), we determined the angle that can minimize the root mean square (RMS) error as shown in the right of Fig. 13.

Fig 13: A detailed explanation of the algorithm to derive the incident angle of X-rays...

<<Verification Experiment Based on Clinical Study>>

    As shown in Fig. 14, dosimetry was performed on 50 patients who underwent chest CT scans. We applied the conditions that we used in normal routine examinations. Our study was approved by the IRB of our hospital (Yamaguchi University Hospital, Japan: 2023-018). Only one OSL dosimeter was placed on the center of the chest. After the examination was finished, by using the CT image of the slice presenting the dosimeter and the measured dose values, we analyzed the average value of the surface dose for each patient. Since this study was conducted using the Automatic Exposure Control (AEC) technique, the Body Mass Index (BMI), which expresses the patient's body size, would affect the exposure dose. Therefore, the correlation between BMI and exposure dose was also investigated.

Fig 14: Clinical examination conditions and patient details. Dosimetry was performed on...